Abstract
Many of the quantum information applications rely on indistinguishable sources of polarizationentangled photons. Semiconductor quantum dots are among the leading candidates for a deterministic entangled photon source; however, due to their random growth nature, it is impossible to find different quantum dots emitting entangled photons with identical wavelengths. The wavelength tunability has therefore become a fundamental requirement for a number of envisioned applications, for example, nesting different dots via the entanglement swapping and interfacing dots with cavities/atoms. Here we report the generation of wavelengthtunable entangled photons from onchip integrated InAs/GaAs quantum dots. With a novel anisotropic strain engineering technique based on PMNPT/silicon microelectromechanical system, we can recover the quantum dot electronic symmetry at different exciton emission wavelengths. Together with a footprint of several hundred microns, our device facilitates the scalable integration of indistinguishable entangled photon sources onchip, and therefore removes a major stumbling block to the quantumdotbased solidstate quantum information platforms.
Introduction
A topical challenge in quantum information processing (QIP) is the generation and manipulation of polarizationentangled photon pairs^{1,2}. Spontaneous parametricdownconversion (SPDC) and fourwavemixing (FWM) have served as the main workhorses for these purposes in the past decade, and the implementation of a fully integrated quantum device is within reach by marrying these sources with chipscale silicon photonics^{3,4,5,6}. However, the generated photons are characterized by Poissonian statistics, that is, one usually does not know when an entangled photon pair is emitted. This fundamentally limits their applications in complex quantum protocols, for example, an eventready test of Bell’s inequality and highefficiency entanglement purifications, where deterministic operations are much favoured^{1}.
The intrinsic limitations of SPDC and FWM processes call for next generation entangled photon sources. III–V semiconductor quantum dots, often referred to as artificial atoms, are among the leading candidates for deterministic quantum light sources. As proposed by Benson et al. single quantum dots (QDs) can generate polarizationentangled photon pairs via its biexciton (XX) cascade decay through the intermediate exciton states X, Fig. 1 (ref. 7). In real III–V QDs the anisotropy in strain, composition and shape reduces the QD symmetry to C_{2v} or the even lower C_{1}, leading to the appearance of an energetic splitting between the two bright X states, the socalled fine structure splitting (FSS)^{8}. High fidelity to the entangled state , with H and V denoting the horizontal and vertical polarizations, can be observed only with a vanishing FSS (typically, smaller than the radiative linewidth of ∼1 μeV). The probability of finding such QDs in an asgrown sample is <10^{−2}. After extensive efforts by many groups, the elimination of FSS can be achieved by applying rapid thermal annealing^{9}, optical Stark effect^{10}, magnetic field^{11,12}, electric field^{13,14}, and more recently, anisotropic strain fields^{15,16} to the QDs. In the past years we have witnessed considerable progress in this field, and entangled photon emissions can be triggered optically^{11,17} from single QDs with high brightness (up to 0.12 pair per excitation pulse)^{18} and high indistinguishability (0.86±0.03 for the XX photons)^{19}. III–V QDs also possess an important advantage of being compatible with mature semiconductor technology, and electrically triggered entangled photon sources have been successfully demonstrated^{16,20}.
Armed with these powerful techniques, III–V QDs have the potential to fulfil the ‘wishlist’ of a perfect entangled photon source^{21}. Among the next goals are the miniaturization and scaling up of the technology. Several important issues need to be considered. First, the FSS of each QD can only be eliminated under particular tuning parameters, and any attempt to manipulate the emission wavelength increases the FSS and spoils the entanglement. This fact undoubtedly restricts the entangled photon emissions at arbitrary wavelengths. The inability to tune the emission wavelength without restoring the FSS, which is unfortunately the common disadvantage associated with all FSS tuning technologies to date, has become a major stumbling block to the QIP applications based on scalable QD sources. Second, as being investigated with SPDC and FWM sources^{3,4,5}, the integration of wavelengthtunable quantum light sources on silicon is arguably one of the most promising choices for onchip QIP applications^{22}.
Here we demonstrate wavelengthtunable entangled photon sources based on III–V QDs integrated on a silicon chip. There are two recent theoretical proposals^{23,24} on generating wavelengthtunable entangled photon from QDs, however, the experimental implementations of their proposals are quite challenging. We design and fabricate a device consisting of QDembedded nanomembranes suspended on a fourlegged thinfilm PMNPT ([Pb(Mg_{1/3}Nb_{2/3})O_{3}]_{0.72}[PbTiO_{3}]_{0.28}) actuator integrated on a silicon substrate. With the combined uniaxial stresses along two orthogonal directions, we are able to keep the FSS strictly below 1 μeV while shifting the exciton wavelength/energy by more than 3,000 times of the QD radiative linewidth. Highfidelity entangled photon emission is demonstrated when the FSS is tuned to below 1 μeV. Therefore wavelengthtunable entangled photons are generated on chip with a singledevice footprint of a few hundred microns.
Results
Concept of device
For the device fabrication we use the industrial transfer printing and die bonding techniques to realize the novel integration of III–V, PMNPT and Si. Unlike the piezo substrate used in all previous works^{15,16,24,25,26,27}, a 15μm PMNPT thinfilm bonded on a silicon substrate is employed here to realize novel microelectromechanical system (MEMS) devices with sophisticated functionalities on chip (Fig. 1, see also Methods section). Arrays of QDcontaining GaAs nanomembranes, each 80 × 80 μm^{2} in size, were then transferred onto the PMNPT MEMS with four actuation legs (Fig. 1a–c). The crystal axes [1–10] and [110] of the GaAs nanomembrane were carefully aligned along the designed stress axes of the actuators. When applying negative (positive) voltages to the electric contacts, the PMNPT legs expand (contract) inplane and therefore exert quasiuniaxial compressive (tensile) stresses to the QDs.
This new device concept has several advantages. First, a controllable anisotropic strain is achieved by the fourlegged configuration, which is not possible with any piezo substrate. Second, the use of piezoelectric film alleviates us from high voltages (typically, up to thousands volts) required for the bulk PMNPT substrate, which is certainly important for onchip integration. And third, theoretically proposed scalable sources of straincontrolled quantum light sources^{23,24} can be realized on chip with such devices.
By sweeping the voltage on only one pair of opposite legs, the exciton emission is shifted over a large range (up to 10 nm in wavelength or 12 meV in energy, Fig. 1d) due to the quasiuniaxial stresses along the legs. For suitable QDs the exciton emission can be tuned across the caesium D1 line at ∼894.7 nm, which is required for realizing a hybrid quantum memory^{28}. To confirm the performance of the device, we have also performed finiteelement modelling of our device and two principle stresses (with a magnitude of ∼2 GPa at voltages of 50 V) along orthogonal directions can be identified (not shown here).
Theory^{26,27,29} predicts that the uniaxial strain tuning behaviour of FSS is determined by the QD principal axis with respect to the uniaxial stress direction. For a QD whose principal axis is closely aligned with the stress direction, the FSS can be effectively eliminated. For typical selfassembled semiconductor QDs the directions of the QD principal axes have a Gaussianlike distribution, and a significant amount^{30} of QDs are closely aligned with the crystallographic directions due to the anisotropic surface adatom diffusion. An example of the FSS tuning behaviour for such an aligned QD is given in Fig. 2a. The uniaxial stress tuning is done by sweeping the voltage V_{AC} from 0 to 100 V on one pair of legs, whereas fixing the voltage V_{BD} at 0 V on the other pair of legs. With increasing V_{AC} the FSS first decreases monotonically to a minimum value and then increases. At V_{AC} of ∼73 V the FSS is completely eliminated. The phase θ, which indicates the angle (see inset of Fig. 2a) between the exciton polarization and the [1–10] crystallographic direction of GaAs, undergoes a sharp phase change of 90°. This result is in agreement with theoretical predictions^{27}.
How does a QD behave when both voltages V_{AC} and V_{BD} are turned on, that is, under the application of a pair of orthogonal uniaxial stresses? This has never been studied due to the lack of realistic experimental techniques. In Fig. 2b we present the FSS tuning result at a different V_{BD} of −25 V. Similar to Fig. 2a we also observe a zero FSS and an abrupt change in θ by exactly 90°. The only difference between the two situations (V_{BD}=0 and −25 V) is the voltage V_{AC} at which the FSS is erased. This is an indication of the recovery of QD electronic symmetry at different experimental conditions. It is then interesting to see how the exciton emission wavelength changes, Fig. 2c. With negative (positive) voltages applied, the legs exert uniaxial compressive (tensile) stresses to the QD which causes a blue (red) shift in exciton emission. This is confirmed by sweeping V_{AC} from 0 to 100 V, from which we observe a red shift of the emission. The device performs remarkably well and we do not observe any hysteresis in the wavelength tuning, see the linear fit. The effect is similar when changing V_{BD} from −25 to 0 V at a fixed V_{AC}. Therefore we have a high degree of control on the exciton wavelength by using two pairs of actuation legs.
Wavelengthtunable entangled photons from QDs
Twodimensional scanning on the two pairs of legs by sweeping both V_{BD} and V_{AC} is then performed. In Fig. 2d we show the results in a threedimensional plot. The astonishing result is that, with this fourlegged device providing orthogonal uniaxial stresses, multiple zero FSS points with different exciton wavelength λ_{X} (energy E_{X}) can be achieved. At different V_{BD}, the electronic symmetry of quantum dot can be always recovered by sweeping V_{AC} and the FSS is erased. The dashed line on the bottom plane of the plot indicates the combinations of (V_{AC} and V_{BD}) at which the FSS reaches its minimum. A linear relationship is found for the ratio of voltage changes ΔV_{AC}/ΔV_{BD}. In terms of the applied stresses (X, Y), indeed, an effective twolevel model (Supplementary Note 1) for the FSS of QDs with exciton polarization closely aligned to principal stress axes predicts a zero FSS with a linear relationship ΔX/ΔY and confirms this experimental finding.
As the independent tunability of exciton wavelength and FSS is a main concern in this work, we plot in Fig. 3a the FSS versus exciton wavelength for different V_{BD}. For clarity, we show only the FSS range from 0 to 5 μeV. It shows exactly a linear behaviour (represented by the solid line fits) in agreement with a k.p analysis, Supplementary Note 1. For selfassembled InAs/GaAs QDs the relationship between the FSS and the attainable entanglement fidelity f^{+} has been well documented and it is commonly accepted that the entanglement persists for a FSS even up to 3–4 μeV (refs 15, 16, 31). As shown in these reports, tuning the FSS to <1 μeV yields a high fidelity f^{+} of >0.7. With this novel device, it is clear that the FSS can be ‘locked’ strictly below 1 μeV, that is, highfidelity entangled photons can be generated from the QDs, for a large range of exciton emissions. Considering the typical lifetime (500 ps–1 ns) of our QDs and therefore a radiative limited linewidth of ∼1 μeV, the tuning range of 3.7 meV (2.3 nm) shown in Fig. 3a corresponds to >3,000 times of the radiative linewidth. This tunability is >1 order of magnitude larger than what has been demonstrated for SPDC sources^{32}.
We have performed the polarization crosscorrelation spectroscopy^{11,14,15,17} on a brighter QDembedded inside another device on the same chip. The FSS is tuned to around zero (0.21±0.20 μeV) to demonstrate the polarization entanglement, and the data are presented in Fig. 3b. A key criterion for entanglement is the presence of a correlation independent of the chosen polarization basis, that is, , with D, A, R, L denoting the diagonal, antidiagonal, righthand circular and lefthand circular polarizations. Clear photon bunching, with a normalized secondorder correlation function g^{(2)}(τ=0)>3, can be observed for the copolarized HH and DD photons, whereas in the circular basis the bunching occurs for the crosspolarized RL photons. The entanglement fidelity f^{+} to the maximally entangled Bell state can be determined from the measurements in Fig. 3b, see Methods section. The peak near the zero time delay yields a fidelity f^{+} of 0.733±0.075 without any background subtraction, which exceeds the threshold of 0.5 for a classically correlated state by >3 s.d. The above results are in line with previous experimental and theoretical works, and verify that highly entangled photons can be generated with our device with large wavelength tunability.
An intuitive understanding of the tuning behaviour can be obtained immediately from a matrix of exciton polarization plots at different voltage combinations, Fig. 4a. The ‘circularity’ of the polarization pedals indicates the relative magnitude of FSS. By sweeping V_{AC} from 0 to 70 V at V_{BD} of 0 V, an increasing tensile stress is applied along legs A&C and the symmetry is gradually recovered. We observe the ‘opening’ of the polarization pedal without any appreciable rotation^{19,20,31}, as predicted by theory^{29}. At the voltage combination (V_{AC}, V_{BD})=(70, 0) V, the exciton emissions become circular polarized with a nearzero FSS and polarizationentangled photons are generated. At V_{BD} of −25 V and −50 V, the symmetry is already partially recovered due to the increased compressive stress along the legs B&D, and therefore the symmetry can be fully recovered with less tensile stress, that is, a smaller V_{AC}, along the legs A&C. Due to the symmetry of orthogonal uniaxial stresses, we can observe the same effect when sweeping V_{AC} at fixed voltages of V_{BD}.
Our experimental findings can also be semiquantitatively understood using the recent theoretical observation, Supplementary Note 1 (ref. 26). The application of two orthogonal stresses of magnitude X and Y can be used as effective knobs to erase the FSS and tune the exciton emission once the stress axis are aligned with the initial exciton polarization. In this case indeed, the FSS scales linearly with the stress anisotropy Δ=X–Y and eventually vanishes at a critical stress anisotropy Δ=Δc. This behaviour is demonstrated in Fig. 4b where we show a density plot of the FSS as a function of the two stress magnitudes X,Y as obtained with the twolevel model Hamiltonian, Supplementary Note 1 (refs 26, 29). Furthermore, the remaining independent hydrostatic part ∝ X+Y can be used to change at will the exciton wavelength λ_{X} (energy E_{X}). This implies a linear interdependence between the exciton emission and the FSS, which is in perfect agreement with the data presented in Fig. 2.
We emphasize that the device presented here holds strong promise for the realization of solidstate scalable QIP platform. First, the wavelength tunability allows the Bell state measurement between two QDs and therefore the swapping of entanglement^{24,33}. Also with suitable QDs (see for example Fig. 1d), it is straightforward to couple these sources with atomic vapours^{34}. Second, the integrated MEMS on silicon have small footprints and low operation voltages, which solves two of the most challenging problems of the strain engineering technique^{15,25}. We can foresee the integration of this device with other photonic structures (for example, with circular ring grating cavities to enhance the photon collection efficiency^{35}) and, most intriguingly, with advanced silicon quantum photonic circuits. The envisioned hybrid can exploit the maturity of CMOS technology, and as well as the waveguiding, processing and detection capabilities associated with silicon photonics.
In summary, we have experimentally realized wavelengthtunable entangled photon sources on a III–V/Si chip, which represents an important step towards scalable entangled photon sources based on III–V QDs. The reported device will play an important role not only in building a solidstate quantum network based on entanglement swapping and quantum memories, but also in building advanced quantum photonic circuits for onchip QIP applications. The MEMS based device features the advantages of sophisticated anisotropic stress control on chip. We envision that it will inspire many other topics in quantum and nanotechnologies, and an interesting perspective is to replace the nanomembranes with the emerging twodimensional materials and to study the straindependent photonic and electronic properties.
Methods
Sample growth
The studied sample was grown on a (001) GaAs substrate by solidsource molecular beam epitaxy. Several samples were used in this work, but their general structures are the same. After the deoxidization and GaAs buffer growth, a thin layer of Al_{0.75}Ga_{0.25}As was grown as the sacrificial layer for the nanomembrane release. The InAs QDs were grown by partial capping and annealing, and embedded in the middle of a GaAs layer with a thickness of a few hundred nanometres. The emission range of the QDs is between 880 and 900 nm and we can easily identify lowdensity regions with <1 QD μm^{−2}.
Device fabrication
As for the device processing, we start with the QDembedded nanomembranes. Standard UV photolithography and wetchemical etching were used to fabricate mesa structures with a size of 80 × 80 μm^{2}. The edge of the nanomembranes was processed along [110] or [1–10] crystal axis of GaAs. The PMNPT film is ∼15 μ thick. The backside is coated with gold as the bottom/ground contact. And the PMNPT film is bonded to a silicon substrate via glue (Ibule photonics). We use focused ion beam to define the trenches in the film, and the depth of the trenches is deep enough to penetrate into the silicon substrate. And then the top contact is deposited by Ebeam sputtering. After this, the device is undercut by wetting chemical etching to form the suspended PMNPT thinfilm legs.
Then we bond the QDembedded nanomembranes to the processed PMNPT/Si substrate using a flipchip bonder. The edge of nanomembrane is carefully aligned along the strain axis of the PMNPT actuation legs. To constrain the displacements of the membrane, the same voltage is applied to opposite legs.
Optical measurement
For optical measurements, the device is loaded into a cryostat chamber which is cooled to ∼5 K. The PL emitted from the QDs is collected by a × 100 microscope objective with numerical aperture of 0.52. The signal is then dispersed by a spectrometer with 1,800 grating and detected by a liquid nitrogen cooled CCD. By inserting a halfwave plate and a linear polarizer directly after the collection lens, polarizationresolved measurements were performed to estimate the FSS. The exciton polarization is determined by aligning the fast optical axis of the polarizer along [1–10] direction of the nanomembrane. With the experimental procedure used in previous works^{13,26}, we can determine the FSS with an accuracy of subμeV. For FSS of below 1 μeV, an error bar of ±0.25 μeV is extracted from the 95% prediction band of the sine fit function. As for the polarization correlation spectroscopy, a nonpolarizing 50:50 beam splitter is placed directly after the objective to divide the optical paths between two spectrometers, which are then used to detect X and XX separately. After each spectrometer, a Hanbury BrownTwiss setup, consisting of a polarizing beam splitter and two highefficiency singlephoton avalanche detectors, is placed. Half and quarterwave were used to select the proper polarization basis. The temporal resolution of the system is ∼450 ps. The entanglement can be quantified by measuring degree of correlation C, which s defined by
where and are normalized secondorder time correlations for copolarized and crosspolarized XX and X photons, respectively. The fidelity f^{+} is calculated by using the formula: f^{+} =, in which C_{HV}, C_{DA} and C_{RL} are degree of correlations in HV, DA and RL bases.
Additional information
How to cite this article: Chen, Y. et al. Wavelengthtunable entangled photons from siliconintegrated III–V quantum dots. Nat. Commun. 7:10387 doi: 10.1038/ncomms10387 (2016).
References
 1.
Pan, J. W. et al. Multiphoton entanglement and interferometry. Rev. Mod. Phys. 84, 777–838 (2012).
 2.
Kwiat, P. G., Waks, E., White, A. G., Appelbaum, I. & Eberhard, P. H. Ultrabright source of polarizationentangled photons. Phys. Rev. A 60, R773–R776 (1999).
 3.
O'Brien, J. L., Furusawa, A. & Vuckovic, J. Photonic quantum technologies. Nat. Photon. 3, 687–695 (2009).
 4.
Politi, A., Cryan, M. J., Rarity, J. G., Yu, S. & O'Brien, J. L. Silicaonsilicon waveguide quantum circuits. Science 320, 646–649 (2008).
 5.
Silverstone, J. W. et al. Onchip quantum interference between silicon photonpair sources. Nat. Photon. 8, 104–108 (2014).
 6.
Najafi, F. et al. Onchip detection of nonclassical light by scalable integration of singlephoton detectors. Nat. Commun. 6, 5873 (2015).
 7.
Benson, O., Santori, C., Pelton, M. & Yamamoto, Y. Regulated and entangled photons from a single quantum dot. Phys. Rev. Lett. 84, 2513–2516 (2000).
 8.
Bayer, M. et al. Fine structure of neutral and charged excitons in selfassembled In(Ga)As/(Al)GaAs quantum dots. Phys. Rev. B 65, 195315 (2002).
 9.
Ellis, D. J. P. et al. Control of finestructure splitting of individual InAs quantum dots by rapid thermal annealing. Appl. Phys. Lett. 90, 011907 (2007).
 10.
Muller, A., Fang, W., Lawall, J. & Solomon, G. S. Creating polarizationentangled photon pairs from a semiconductor quantum dot using the optical stark effect. Phys. Rev. Lett. 103, 217402 (2009).
 11.
Stevenson, R. M. et al. A semiconductor source of triggered entangled photon pairs. Nature 439, 179–182 (2006).
 12.
Pooley, M. A. et al. Energytunable quantum dot with minimal fine structure created by using simultaneous electric and magnetic fields. Phys. Rev. Appl. 1, 024002 (2014).
 13.
Bennett, A. J. et al. Electricfieldinduced coherent coupling of the exciton states in a single quantum dot. Nat. Phys. 6, 947–950 (2010).
 14.
Ghali, M., Ohtani, K., Ohno, Y. & Ohno, H. Generation and control of polarizationentangled photons from GaAs island quantum dots by an electric field. Nat. Commun. 3, 661 (2012).
 15.
Trotta, R., Wildmann, J. S., Zallo, E., Schmidt, O. G. & Rastelli, A. Highly entangled photons from hybrid piezoelectricsemiconductor quantum dot devices. Nano Lett. 14, 3439–3444 (2014).
 16.
Zhang, J. et al. Straintunable entangledlightemitting diodes with high yield and fast operation speed. Nat. Commun. 6, 10067 (2015).
 17.
Akopian, N. et al. Entangled photon pairs from semiconductor quantum dots. Phys. Rev. Lett. 96, 130501 (2006).
 18.
Dousse, A. et al. Ultrabright source of entangled photon pairs. Nature 466, 217–220 (2010).
 19.
Muller, M., Bounouar, S., Jons, K. D., Glassl, M. & Michler, P. Ondemand generation of indistinguishable polarizationentangled photon pairs. Nat. Photon. 8, 224–228 (2014).
 20.
Salter, C. L. et al. An entangledlightemitting diode. Nature 465, 594–597 (2010).
 21.
Lu, C.Y. & Pan, J.W. Quantum optics: pushbutton photon entanglement. Nat. Photon. 8, 174–176 (2014).
 22.
Harris, N. C. et al. Integrated source of spectrally filtered correlated photons for largescale quantum photonic systems. Phys. Rev. X 4, 041047 (2014).
 23.
Wang, J., Gong, M., Guo, G. C. & He, L. Towards scalable entangled photon sources with selfassembled InAs/GaAs quantum dots. Phys. Rev. Lett. 115, 067401 (2015).
 24.
Trotta, R., MartínSánchez, J., Daruka, I., Ortix, C. & Rastelli, A. Energytunable sources of entangled photons: a viable concept for solidstatebased quantum relays. Phys. Rev. Lett. 114, 150502 (2015).
 25.
Ding, F. et al. Tuning the exciton binding energies in single selfassembled InGaAs/GaAs quantum dots by piezoelectricinduced biaxial stress. Phys. Rev. Lett. 104, 067405 (2010).
 26.
Trotta, R. et al. Universal recovery of the energylevel degeneracy of bright excitons in InGaAs quantum dots without a structure symmetry. Phys. Rev. Lett. 109, 147401 (2012).
 27.
Plumhof, J. D. et al. Straininduced anticrossing of bright exciton levels in single selfassembled GaAs/Al_{x}Ga_{1x}As and In_{x}Ga_{1x}As/GaAs quantum dots. Phys. Rev. B 83, 121302 (2011).
 28.
Akopian, N., Wang, L., Rastelli, A., Schmidt, O. G. & Zwiller, V. Hybrid semiconductoratomic interface: slowing down single photons from a quantum dot. Nat. Photon 5, 230–233 (2011).
 29.
Gong, M., Zhang, W., Guo, G.C. & He, L. Exciton polarization, finestructure splitting, and the asymmetry of quantum dots under uniaxial stress. Phys. Rev. Lett. 106, 227401 (2011).
 30.
Gong, M. et al. Statistical properties of exciton fine structure splitting and polarization angles in quantum dot ensembles. Phys. Rev. B 89, 205312 (2014).
 31.
Hudson, A. J. et al. Coherence of an entangled excitonphoton state. Phys. Rev. Lett. 99, 266802 (2007).
 32.
Fedrizzi, A., Herbst, T., Poppe, A., Jennewein, T. & Zeilinger, A. A wavelengthtunable fibercoupled source of narrowband entangled photons. Opt. Express 15, 15377–15386 (2007).
 33.
Troiani, F. Entanglement swapping with energypolarizationentangled photons from quantum dot cascade decay. Phys. Rev. B 90, 245419 (2014).
 34.
Duan, L. M., Lukin, M. D., Cirac, J. I. & Zoller, P. Longdistance quantum communication with atomic ensembles and linear optics. Nature 414, 413–418 (2001).
 35.
Sapienza, L., Davanco, M., Badolato, A. & Srinivasan, K. Nanoscale optical positioning of single quantum dots for bright and pure singlephoton emission. Nat. Commun. 6, 7833 (2015).
Acknowledgements
The work was financially supported by the BMBF Q.ComH (16KIS0106) and the European Union Seventh Framework Programme 209 (FP7/20072013) under Grant Agreement No. 601126 210 (HANAS). The authors acknowledge Carmine Ortix for theoretical supports. Part of the data analysis was done by the XRSP3 software developed by Armando Rastelli. We thank Yongheng Huo for providing one of the test QD sample materials. We also thank B. Höfer, C. Jiang, B. Eichler, R. Engelhard, M. Bauer and S. Harazim for discussions and technical supports. F.D. acknowledges Armando Rastelli for the continuous support and his early contributions to this project.
Author information
Affiliations
Institute for Integrative Nanosciences, IFW Dresden, Helmholtzstraße 20, 01069 Dresden, Germany
 Yan Chen
 , Jiaxiang Zhang
 , Michael Zopf
 , Kyubong Jung
 , Yang Zhang
 , Robert Keil
 , Fei Ding
 & Oliver G. Schmidt
Material Systems for Nanoelectronics, Chemnitz University of Technology, Reichenhainer strasse 70, 09107 Chemnitz, Germany
 Oliver G. Schmidt
Authors
Search for Yan Chen in:
Search for Jiaxiang Zhang in:
Search for Michael Zopf in:
Search for Kyubong Jung in:
Search for Yang Zhang in:
Search for Robert Keil in:
Search for Fei Ding in:
Search for Oliver G. Schmidt in:
Contributions
F.D. conceived the experiment and supervised the project together with O.G.S. who directed the research. The devices were designed by Y.C. and F.D. and fabricated by Y.C. with the help from K.J. and Y.Z. Two samples for the project were designed and grown by R.K. Optical experiments were performed by Y.C. and J.Z. The data were analysed by Y.C., F.D. and J.Z. Finiteelement modelling simulations were done by M.Z. and Y.Z. The manuscript was written by Y.C. and F.D. with the inputs from all the authors.
Competing interests
The authors declare no competing financial interests.
Corresponding author
Correspondence to Fei Ding.
Supplementary information
PDF files
 1.
Supplementary Information
Supplementary Figures 12, Supplementary Note 1 and Supplementary References.
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Further reading

A solidstate source of strongly entangled photon pairs with high brightness and indistinguishability
Nature Nanotechnology (2019)

Piezotronic effect on the luminescence of quantum dots for micro/nanonewton force measurement
Nano Research (2018)

Solidstate ensemble of highly entangled photon sources at rubidium atomic transitions
Nature Communications (2017)

Onchip single photon filtering and multiplexing in hybrid quantum photonic circuits
Nature Communications (2017)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.