DEVICE FOR THE PRODUCTION OF SINGLE PHOTONS AND ANGLED PAIRS OF PHOTONS

Description

Art antérieur

[0001] The present invention relates generally to photon sources and in particular to a device for generating single photons and entangled photon pairs, and to methods implementing the device.

[0002] Deploying a quantum communication network requires a single-photon source capable of generating a train of optical pulses, each containing a single photon. Such sources can provide two distant communicating parties with a secret key, enabling them to encrypt subsequent communications. Imperfections in the propagation channel, such as optical fibers or free space, limit the range of such an application to a few hundred kilometers. The range of quantum communication networks can be extended using quantum repeaters, which rely on the entanglement of a pair of photons. The generation of single-photon pulses or entangled photon pairs can be achieved using nanometer-sized emitters, such that the transition between their electronic states corresponds to the radiation of a single quantum dipole.

[0003] Furthermore, the development of a quantum computer relies on systems that can be used as quantum bits. A quantum bit is a coherent superposition of two basis states, commonly denoted |0> and |1>, meaning it can take on an infinite number of possible values, unlike the binary bits used by classical computers, which are either 0 or 1. Quantum bits can be solid-state or photonic. While the former offers flexibility in terms of storage and processing, photonic quantum bits enable the long-range transmission of quantum information. To enable the processing and transmission of quantum information, it is common practice to use solid-state and photonic quantum bits together, with an interface between the two.

[0004] The use of quantum dots in applications including the deployment of quantum communication networks and the development of quantum computers has seen a major boom in recent years. The interest in such nanometric structures stems from their similarities to the atom in terms of charge carrier confinement and the quantization of their energy levels. These characteristics allow a quantum dot to generate a single photon through a spontaneous emission process in response to a suitable light excitation, enabling an electron to migrate from the valence band to the conduction band. The same mechanism can be adapted to place two electrons in the same energy level of the conduction band and generate a pair of entangled photons. Quantum dots also allow for reliable interaction of light with the characteristics of the confined charge carriers.

[0005] Isolated quantum dots used as sources of single photons or entangled photon pairs are generally limited in terms of brightness and throughput. While brightness quantifies the probability of obtaining a single photon or an entangled photon pair per optical pulse, throughput measures the number of photons emitted per second and is equal to the product of the brightness, as defined above, and the emission clock rate governing the optical pulses that excite the quantum dot. To improve the brightness and throughput performance of quantum dots, it is known to insert the quantum dot into a resonant optical cavity. Such a configuration exploits the Purcell effect, also known as the weak light-matter coupling regime, to increase the spontaneous emission rate of the quantum dot in the cavity mode and allows for more efficient collection of the photons emitted by the quantum dot.Encapsulating the quantum dot in an optical cavity also reduces the sensitivity of the quantum dot to environmental disturbances (mechanical, electrical or other) and allows the generation of indistinguishable single photons, that is to say identical in terms of quantum state (same frequency, same polarization state, same spatial and temporal distribution).

[0006] Quantum dots are generally semiconductor-based, obtained through a molecular beam epitaxy process that involves growing layers of semiconductors with different band gap energies and lattice parameters. These differences create nanometer-sized quantum dots capable of confining charge carriers in three dimensions.

[0007] An electron from the valence band of a quantum dot can transition to the conduction band to form, in association with the hole left in the valence band, an electron-hole pair, also called an exciton. Exciton formation can occur by absorbing a photon with energy greater than or equal to the band gap of the material forming the quantum dot. The exciton state corresponding to exciton formation is an unstable state, and the restoration of the neutral state, also called the ground state, occurs through recombination of the electron with the hole, emitting a photon via a spontaneous emission process characterized by a radiative lifetime on the order of nanoseconds and an emission wavelength (or frequency) corresponding to the band gap of the material forming the quantum dot.

[0008] A quantum dot can confine a biexciton formed by two excitons coupled by the Coulomb interaction, sharing the same ground state. The energy level of the biexciton state is higher than that of a single exciton. The formation of a biexciton can occur by successively absorbing two photons. The relaxation of a biexciton also occurs via a spontaneous emission process that gives rise to two successively emitted photons. The two photons are emitted at different wavelengths because the Coulomb interaction differs depending on whether one electron-hole pair or two electron-hole pairs are excited within the quantum dot. Thus, the photon emitted during the transition of the system from the biexciton state to the exciton state has a different wavelength than the photon emitted during the transition of the system from the exciton state to the neutral state.

[0009] The polarization state of the photon absorbed to form the exciton and that of the photon emitted during electron-hole recombination obey optical selection rules governed by the Pauli exclusion principle and the conservation of angular momentum. Such selection rules depend on the spin states of the electron-hole pair and state that the bright excitonic states emitting a photon during recombination are those corresponding to spin states of opposite signs. In the case where the quantum dot has cylindrical symmetry, the system exhibits two bright excitonic states with degenerate energy. figure 1 represents the relaxation of a 103 biexciton formed in a cylindrical (or "isotropic") semiconductor quantum dot 100 where only bright excitonic states are involved. The relaxation of a 103 biexciton confined in an isotropic quantum dot 100 occurs in a three-energy-level system by successively emitting two photons. Two paths are possible for this: relaxation of biexciton 103 to exciton 1020 with emission of a right-circularly polarized photon, then relaxation of exciton 1020 to ground state 101 with emission of a left-circularly polarized photon; or relaxation of biexciton 103 to exciton 1021 with emission of a left-circularly polarized photon, then relaxation of exciton 1021 to ground state 101 with emission of a right-circularly polarized photon.

[0010] Since the two paths are indistinguishable, because excitons 1020 and 1021 are degenerate, this leads to the emission of a pair of entangled photons in polarization.

[0011] It should be noted that the generation of single photons can be achieved by exploiting the recombination of 1020 or 1021 excitons towards the 101 ground state and that it is unnecessary, in this case, to go through a 103 biexciton state. The creation of a 103 biexciton is required for the generation of entangled photon pairs.

[0012] However, in practice, semiconductor quantum dots are anisotropic, implying a lifting of exciton level degeneracy, and therefore a loss of entanglement of the emitted photons. Such lifting of degeneracy is primarily caused by shape anisotropy effects of the quantum dot, piezoelectric effects, mechanical constraints, etc. In an anisotropic quantum dot, the bright excitonic states are separated by an amount of energy called "fine-structure splitting" and are linear combinations of the ground states as defined in symmetric quantum dots. The bright excitonic states in an anisotropic quantum dot exhibit optical emissions linearly polarized along two eigenaxes "x" and "y" corresponding to the crystallographic directions (the "z" axis being the growth axis). figure 2 illustrates the recombination of a biexciton 203 in an anisotropic semiconductor quantum dot 200. The relaxation of a biexciton 203 to the ground state 201 can only occur in a linear polarization state along one of the eigenaxes of the quantum dot, "x" or "y", leaving an exciton 2020 or 2021 which, in turn, relaxes in the same polarization state as the biexciton 203. The lifting of the degeneracy of the exciton level leads to the total or partial loss of entanglement between the two generated photons.

[0013] It is known to use anisotropic quantum dots to generate single photons of enhanced performance by resonantly exciting a superposition of two bright excitonic states and exploiting the time evolution ΔΦ of the phase between the two excitonic states polarized along "x" and "y", the speed of which (called "transfer velocity", because it governs the transition from a linear polarization state of the emitted photons to an orthogonal state, as will be explained in more detail with reference to the figure 10 ) is proportional to the value of the fine structure splitting according to the following relationship: ΔΦ = exp i Δ FSS ℏ t , , where is the reduced Planck constant. Excitation can be achieved with photons linearly polarized along an axis that is not collinear with the eigenaxes of the quantum dot, ideally with a 45-degree orientation. In order to collect only the photons generated by the quantum dot, only photons linearly polarized in the direction perpendicular to that of the excitation photons are detected. Generally, the fine-structure splitting value determines the performance of the single-photon source in terms of brightness and purity; purity quantifies the ability to generate a single photon during a given excitation period. The fine-structure splitting value is difficult to control during the quantum dot fabrication process. High fine-structure splitting values ​​allow for increased transfer velocity and, consequently, increased brightness of the single-photon source.However, the probability that the quantum dot emits more than one photon during a single excitation period is not negligible, which reduces the performance of the single-photon source in terms of purity. Conversely, low fine-structure splitting values ​​reduce the transfer velocity, thus decreasing the probability of a second photon being emitted during the same excitation period. Such low fine-structure splitting values ​​improve the purity of the single-photon source at the expense of brightness, which degrades because the quantum dot emits more rapidly in the polarization parallel to that of the excitation photons than in the orthogonal polarization. Therefore, there is a need to control the fine-structure splitting value in order to generate single photons at a high rate while reducing the probability of more than one photon being emitted during a single excitation period.

[0014] The use of anisotropic quantum dots as currently fabricated remains unsuitable for generating entangled photon pairs due to the lifting of exciton state degeneracy. Therefore, there is a need to statically reduce the value of fine-structure splitting, specifically to eliminate it, in order to restore the indistinguishability of radiative emission paths from the biexciton and thus obtain entangled photon pairs.

[0015] It is known to reduce and / or control the value of fine-structure splitting in a quantum dot by applying a vertical electric field in conjunction with one or more stress fields, specifically mechanical stress fields, in two or three directions of the environment in which the quantum dot is embedded. Such a solution has been demonstrated in [1] and [4]. Applying such stress fields requires thinning the substrates on which the quantum dot is fabricated in order to place piezoelectric materials as close as possible to the quantum dot and to ensure that the applied mechanical stress is effectively transferred to the implemented quantum dot. The thinning required to implement such an approach poses technological challenges for the fabrication of a three-dimensional (3D) optical cavity incorporating a quantum dot.In addition, the proximity of the quantum dot to a surface makes it sensitive to environmental disturbances, which causes the loss of indistinguishability of the single photons generated.

[0016] It is also known to apply an electric field in two dimensions of the space in which the quantum dot is inserted in order to control the value of the fine-structure splitting. Such a solution is implemented in [2] by inserting several (typically four) electrical contact pads on a plane very close to that of the quantum dot. The proximity of the quantum dot to surfaces where such electrical contact pads are defined is known to increase charge noise and reduce the performance of the emitted photons in terms of indistinguishability.

[0017] There is therefore a need for a source of single photons or pairs of entangled photons at the basis of a semiconductor quantum dot in which fine structure splitting can be controlled statically or dynamically, without the disadvantages of the prior art. Définition générale de l'invention

[0018] To this end, the present invention provides a photon generation device comprising a quantum dot inserted in a micro-pillar type optical cavity having at least one optical mode, the quantum dot having at least one ground state and two states at an elementary excitation, the optical cavity having a lower face and an upper face, the lower face having an electrical contact, characterized in that it comprises at least three electrically isolated contact pads arranged around the upper face of the cavity.

[0019] In one embodiment, the electrical contact pads can be connected to the upper face of the cavity by semiconductor arms oriented radially with respect to the cavity and having, in a tangential direction and, at their end closest to the cavity, a width less than that of the cavity.

[0020] In another embodiment, the electrical contact pads can be connected to semiconductor arms oriented radially towards the cavity and whose ends can be separated from the upper face of the cavity by an empty gap, or one filled with dielectric, of sub-micrometer width.

[0021] Advantageously, the micro-pillar type optical cavity can form a PIN type diode, with the quantum dot located in an intrinsic region of the diode.

[0022] Alternatively, the electrical contact pads can be supported by respective pillars that are optically and electrically decoupled from the cavity.

[0023] In one embodiment, the device may further include at least three adjustable voltage sources to apply respective and variable potential differences between each of the electrical contact pads and the electrical contact carried by the underside of the optical cavity.

[0024] In another embodiment, the micro-pillar type optical cavity may exhibit at least a first and a second pair of modes, each pair of modes being degenerate in polarization, and in which the quantum dot may also exhibit a state with two elementary excitations.

[0025] Advantageously, the device may further comprise a second optical cavity coupled to the optical cavity in which the quantum dot is inserted, the geometry of the first and second cavities and the strength of their coupling being chosen such that the assembly of the two coupled cavities exhibits a first pair of polarization-degenerate modes resonating with the transitions between the two elementary excitation states and the two one elementary excitation states of the quantum dot and a second pair of polarization-degenerate modes resonating with the transitions between the two one elementary excitation states and the ground state of the quantum dot, the modes of each of the pairs exhibiting radiation patterns that overlap by 70% or more.

[0026] In particular, the second optical cavity can also be of the micro-pillar type, with the two cavities arranged side by side.

[0027] A method for generating entangled photon pairs using a photon generation device is proposed; the method may include the following steps: apply a potential difference between each of the electrical contact pads and the electrical contact carried by the lower face of the optical cavity, the potential differences being chosen to: ∘ minimize an energy gap between the two states at one excitation of the quantum dot; and ∘ bring the transitions between the two elementary excitation state and the one elementary excitation states into resonance with the first pair of modes of the cavity, and the transitions between the one elementary excitation states and the ground state into resonance with the second pair of modes of the cavity; and populate the two elementary excitation state of the quantum dot by an energy input.

[0028] Furthermore, a method for generating single photons using a photon generation device is proposed; the method may include the following steps: apply a potential difference between each of the electrical contact pads and the electrical contact carried by the lower face of the optical cavity; and illuminate the quantum dot of the device with a light pulse exhibiting a spectrum and polarization direction corresponding to a mode of the optical cavity; the potential differences being chosen to modify the energies of the two eigenstates of an elementary excitation such that: ∘ the states are resonant with wavelengths of the light pulse spectrum; and ∘ the probability of emission, by the box, of a single photon exhibiting a linear polarization orthogonal to that of illumination is maximized.

[0029] In one embodiment, potential differences can also be chosen to orient eigenaxes of the quantum dot along directions intermediate between those of two eigenaxes of polarization of the optical cavity.

[0030] In another embodiment, the potential differences can vary over time and: during a first period, during at least part of which the quantum dot is illuminated, take first values ​​chosen so that the energies of the two eigenstates of an elementary excitation have a non-minimal difference, the first period having a duration chosen to cause a phase evolution between 45° and 135°, and preferably between 80° and 100°, between the two states; and during a second period, immediately following the first period, take values ​​chosen to minimize the difference between the energies of the two eigenstates of an elementary excitation. Brève description des figures

[0031] Other features and advantages of the invention will become apparent from the following description and the accompanying figures, in which: [ Fig. 1 ] And [ Fig. 2 ] described above, illustrate the relaxation of a biexciton in isotropic and anisotropic quantum dots, respectively, [ Fig. 3A ] And [ Fig. 3B ] represent a device for generating single photons and entangled photon pairs according to an embodiment of the invention, [ Fig. 4 ] illustrates the generation of a three-component electric field in space according to one embodiment of the invention, [ Fig.5 ], [ Fig. 6 ] And [ Fig. 7 ] represent a portion of the top view of a device for generating single photons and entangled photon pairs according to different embodiments of the invention, [ Fig. 8 ] represents a micro-pillar type optical cavity incorporating a quantum dot according to an embodiment of the invention, [ Fig. 9 ] represents a portion of the top view of a device for generating single photons and entangled photon pairs implementing two optically coupled cavities according to another embodiment of the invention, [ Fig. 10 ] represents the evolution over time of the excited states of a quantum dot embedded in a micro-pillar type optical cavity, the quantum dot exhibiting a non-zero exchange anisotropy and the optical cavity exhibiting birefringence, [ Fig. 11 ] represents the use of a single-photon generation device according to an embodiment of the invention, [ Fig. 12 ] represents a calibration method for an anisotropic quantum dot, [ Fig. 13A ] And [ Fig. 13B ] represent two methods of generating single photons according to two embodiments of the invention, [ Fig. 14 ] represents a method for generating entangled photon pairs according to an embodiment of the invention, [ Fig. 15 ] represents experimental measurements of the fine structure splitting of an anisotropic quantum dot subjected to an electric field generated according to embodiments of the invention. Description détaillée

[0032] THE figures 3A et 3B Figures 300 and 301 respectively represent a top view and a cross-sectional view of a single-photon and entangled-photon-pair generation device according to an embodiment of the invention. The device comprises a quantum dot 301 inserted into a micro-pillar-type optical cavity 302. The optical cavity 302 has a cylindrical shape and is connected to three pillars 303 by semiconductor arms 308a, 308b, 308c.

[0033] The 301 quantum dot can be semiconductor-based, obtained through a molecular beam growth process that involves growing semiconductor layers with different band gap energies and lattice parameters. The layer forming the 301 quantum dot can be an indium arsenide (InAs) layer, while the substrate and encapsulating layers can be gallium arsenide (GaAs) layers. The 301 quantum dot can have the shape of a truncated cone with a height of a few nanometers along its growth axis and a base diameter of approximately twenty nanometers. These characteristics give the 301 quantum dot properties similar to those of an atom in terms of charge carrier confinement and energy level quantization.

[0034] The 301 quantum dot can be in a neutral state, also called the ground state, or in an excited state corresponding to the formation of an electron-hole pair, also called an exciton or a state with one elementary excitation. The formation of an exciton in the presence of a charge carrier already confined within the 301 quantum dot corresponds to a trion, which can be quantized as positive or negative depending on the sign of the already confined charge carrier. A biexciton, also called a state with two elementary excitations, is formed in the 301 quantum dot when two excitons are simultaneously confined.

[0035] Charge carriers, electrons and holes, are fermions characterized by a half-integer spin: ±1 / 2 for electrons and ±3 / 2 for heavy holes. The exciton state of a symmetric 301 quantum dot is bright when electron-hole recombination occurs via a spontaneous emission process, resulting in the emission of a photon. Bright exciton states, also called basis exciton states, correspond to a total angular momentum of ±1 and relax by emitting a circularly polarized photon, left- or right-handed depending on the sign of the sum of the spins.

[0036] In practice, the 301 quantum dot is anisotropic, implying a lifting of the degeneracy of the exciton level of the 301 quantum dot. The bright exciton states in an anisotropic 301 quantum dot, also called elementary excitation eigenstates, are linear combinations of the basic exciton states as defined in a symmetric 301 quantum dot and are separated by a non-zero amount of energy called fine-structure splitting.

[0037] The micropillar forming the optical cavity 302 is cylindrical and consists of two stacks of layers 3020, 3022 surrounding a central region 3021. Such layers are typically semiconducting and can be P- or N-doped according to a given configuration. For example, the upper part of the cavity 3022 may be P-type (or N-type) doped, while the lower part 3020 may be doped of the opposite type, so that the cavity forms, from an electrical point of view, a PN diode. Each of the two stacks of layers 3020, 3022 forms a Bragg mirror and is obtained by alternating two layers of two different materials in terms of refractive index. The thickness and refractive index of each of the two layers forming each of the two stacks 3020, 3022 can be chosen so that their product is equal to λ / 4, λ being the operating wavelength.The number of layer pairs forming the bottom stack 3020 in contact with a substrate 305 can be different from the number of layer pairs forming the top stack 3022. Such a difference allows, for example, for the upward emission of photons generated by the quantum dot 301 to increase the brightness of the photon sources 300. Each of the two Bragg mirrors 3020 and 3022 allows incident light to be reflected along the layer stacking axis over a wide range of wavelengths centered around the operating wavelength λ. The central region 3021 placed between the two stacks 3020, 3022 can have a thickness equal to λ and is configured to contain the quantum dot 301. The presence of the central region 3021 allows the reflectivity of the micro-pillar to be modified so that the two Bragg mirrors 3020, 3022 become transmissive at the operating wavelength.The micropillar's radius is on the order of a few micrometers and is chosen to allow at least one optical mode, called the fundamental mode, to propagate through the micropillar with minimal energy. Advantageously, the micropillar's radius can be chosen to allow the propagation of several optical modes. The micropillar can have rotational symmetry, implying two degenerate polarization states for each of the propagation modes. Alternatively, the micropillar is a birefringent medium in which light propagates anisotropically. This birefringence property lifts the polarization degeneracy of the optical propagation modes, which are defined by the minor and major axes of the micropillar, denoted H and V, respectively.

[0038] Advantageously, the optical cavity 302 has at least one first and one second pair of optical modes, each of the mode pairs being degenerate in polarization. An optical cavity 302 having two pairs of degenerate modes in polarization is particularly used in the generation of entangled photon pairs by matching the relaxation wavelength of the biexciton (state with two elementary excitations) to the exciton (state with one elementary excitation) with the energy of the first pair of modes and by matching the relaxation wavelength of the exciton to the neutral state with the energy of the second pair of modes, for example.

[0039] The pillars 303 may have a height equal to, greater than, or less than that of the optical cavity 302. The optical cavity 302 may be positioned at the center of the assembly formed by the three pillars 303, with sufficient radial spacing to optically decouple the optical cavity 302 from the pillars 303. Furthermore, the pillars 303 must be electrically insulated from each other, except at their lower face, corresponding to the electrical contact 305, which will be discussed later. The pillars 303 may be formed by the same layer stacking as the micro-pillar forming the optical cavity 302. Such a structure has the advantage of being compatible with many cleanroom manufacturing processes known to those skilled in the art.The manufacturing steps of such a structure may include a thin film deposition process to achieve layer stacking, a lithography process followed by a dry etching process to define the optical cavity 302 and the three pillars 303.

[0040] The upper face of each of the pillars 303, defined as the face opposite the one in contact with the substrate 305, includes an electrical contact pad 304a, 304b, 304c whose dimensions are chosen so that the electrical contact pad 304a, 304b, 304c is completely supported by the upper face of the corresponding pillar 303. The three electrical contact pads 304a, 304b, 304c may have the same electrical properties and the same geometric shape, which may be triangular, rectangular, or other. The transverse dimensions of such electrical contact pads 304a, 304b, 304c defined in a plane perpendicular to the layer stacking axis (growth axis) are preferably greater by at least a factor 10 compared to the thickness of such pads 304a, 304b, 304c defined along the layer stacking axis.As an example, the thickness of the electrical contact pads 304a, 304b, 304c is chosen to be less than 50 nanometers and their transverse dimensions greater than 50 micrometers, which facilitates their connection to the various voltage sources 306a, 306b, 306c via a wire bonding process. The electrical contact pads 304a, 304b, 304c are arranged symmetrically around the optical cavity 302, which corresponds to the same radial spacing between each of the electrical contact pads 304a, 304b, 304c and the optical cavity 302, and the same angular spacing between each pair of adjacent electrical contact pads.

[0041] The lower faces of the pillars 303 are connected to another electrical contact 305, the electrical contact 305 thus being carried by the lower face of the optical cavity 302. The electrical contact 305 is made on a semiconductor surface whose electrical conductivity is increased by P- or N-doping. The electrical contact 305 can be in the shape of a disk, a rectangle, or another 2D geometric shape that allows it to completely cover the lower face of the assembly formed by the optical cavity 302 and the pillars 303. Such a configuration makes it possible to apply an electrical potential difference between each of the electrical contact pads 304a, 304b, 304c arranged around the upper face of the optical cavity 302 and the electrical contact 305 carried by the lower face of the optical cavity 302 using three independent voltage sources 306a, 306b, 306c. delivering respective tensions Ta, Tb and Tc.The voltage delivered by each of the voltage sources 306a, 306b, 306c can be constant or variable over time.

[0042] Semiconductor arms 308a, 308b, 308c are used to connect each of the electrical contact pads 304a, 304b, 304c to the upper face of the optical cavity 302. The materials forming the semiconductor arms are doped in such a way as to increase their electrical conductivity, without this conductivity becoming of the metallic type (degenerate doping). The semiconductor arms 308a, 308b, 308c can have the same height as the optical cavity 302, their length is long enough to connect the outer face of the optical cavity 302 to the inner face of the pillars 303 above which are arranged the electrical contact pads 304a, 304b, 304c, and their width is small enough so that two adjacent semiconductor arms 308a, 308b, 308c are electrically isolated at the point of their contact with the outer face of the optical cavity 302.Each of the 308a, 308b, and 308c semiconductor arms extends radially towards its corresponding pillar 303, forming a 120-degree angle between each pair of adjacent 308a, 308b, and 308c semiconductor arms. The use of dielectric or semiconductor arms amplifies the horizontal component of the applied electric field by concentrating the field lines.

[0043] Photons generated by the quantum dot 301 according to one of the optical modes and photons exciting the quantum dot 301 leave and enter, respectively, the optical cavity 302 according to the same or two different radiation patterns 307, each of which is characterized by a given opening angle.

[0044] The electrical potential difference between each of the electrical contact pads 304a, 304b, 304c and the electrical contact 305 on the lower face of the optical cavity 302 gives rise to an elementary electric field and controls its characteristics in terms of intensity and orientation. The resulting electric field in the central region 3021, to which the quantum dot 301 is subjected, is the vector sum of three elementary electric fields arising from the electrical potential difference between each of the three electrical contact pads 304a, 304b, 304c arranged around the upper face of the optical cavity 302 and the electrical contact 305 on the lower face of the optical cavity 302.Such a resulting electric field has three components: a vertical component defined in the same direction as the growth axis (layer stacking axis) of the optical cavity 302 and two horizontal components defined in the layer stacking plane forming the two Bragg mirrors 3020, 3022. The characteristics of such a resulting electric field in terms of orientation and intensity are deterministically adjustable by acting on one or more of the elementary electric fields via the associated electric potential difference.

[0045] There figure 4 This illustrates the generation of a three-component electric field defined in an orthonormal trihedron by setting up four electrical contacts according to the configuration described above. Two voltage sources, 306a and 306b, are shown, with the elementary electric fields and the resulting electric field represented by dashed and solid lines, respectively. The different components of the resulting electric field allow us to: adjust the value of the fine structure splitting of the 301 quantum dot; tune the emission wavelength associated with the relaxation of its bright excitonic states; and define the orientation of the 301 quantum dot's eigenaxis along which it can emit or be excited by photons.

[0046] To adjust these three parameters, it is necessary to have at least three degrees of freedom, therefore at least three independent voltage sources connected to three non-collinear electrical contact pads.

[0047] From an experimental perspective, the intrinsic parameters of the 301 quantum dot, such as fine-structure splitting, are not precisely known in advance due to imperfections in the fabrication processes. The parameters of interest for the 301 quantum dot, such as fine-structure splitting, emission wavelengths, and eigenaxis orientations, are generally measured after the fabrication of the single-photon and entangled-photon-pair generation device 300.

[0048] There figure 5 Figure 300 represents a top view of a portion of a single-photon and entangled-photon-pair generation device according to another embodiment of the invention. In such an embodiment, the semiconductor arms 308a, 308b, 308c as described above are not in direct contact with the optical cavity 302. The separation between each semiconductor arm 308a, 308b, 308c and the optical cavity 302 is provided by a layer of dielectric material 309. The dielectric material layer 309 is configured to provide electrical insulation between the optical cavity 302 and the semiconductor arms 308a, 308b, 308c, and its thickness is on the order of tens of nanometers to a few micrometers. The dielectric material forming the dielectric layer 309 may be a vacuum.Electrical isolation of the optical cavity 302 from its adjacent environment prevents an intense electric current from passing through the optical cavity 302 and disrupting the operation of the quantum dot 301.

[0049] There figure 6 represents another embodiment of the invention in which no semiconductor arms are used to connect the optical cavity 302 to the various pillars 303 implemented. In such an embodiment, the pillars 303 are arranged symmetrically around the optical cavity 302, and the radial spacing between the optical cavity 302 and each of the pillars 303 is chosen to be less than 10 micrometers, for example. Such radial spacing allows the quantum dot 301 to be subjected to an electric field sufficiently intense to act on the characteristic parameters of the quantum dot 301, including the value of the fine-structure splitting. The separation between the optical cavity 302 and the pillars 303 can be ensured by a dielectric material or by a vacuum.

[0050] There figure 7 represents another embodiment of the invention in which the electrical contact pads 304a, 304b, 304c are supported by a cylindrical hollow structure 3030 surrounding the optical cavity 302. The hollow structure 3030 may be made of a dielectric material, thus providing electrical insulation between the electrical contact pads 304a, 304b, 304c. The hollow structure 3030 may have the same height as the optical cavity 302, and its upper surface is chosen to be sufficiently large to contain the electrical contact pads 304a, 304b, 304c. Furthermore, semiconductor arms 308a, 308b, 308c associated with a layer of dielectric material 309, as described in the embodiment of the figure 5 are used to connect the electrical contact pads 304a, 304b, 304c to the optical cavity 302. Alternatively, the connection between the electrical contact pads 304a, 304b, 304c and the optical cavity 302 can be achieved by bringing the electrical contact pads 304a, 304b, 304c sufficiently close to the optical cavity 302 as described in the figure 6 or by using 308a, 308b, 308c semiconductor arms as described in the figures 3A et 3B .

[0051] There figure 8 represents the structure of a micro-pillar-type optical cavity 302 according to an embodiment of the invention. In such an embodiment, the two upper Bragg mirrors 3022 and lower Bragg mirrors 3020 forming the micro-pillar are P- and N-doped, respectively. The central region 3021 containing the quantum dot 301 remains intrinsic, so that the micro-pillar forms, from an electrical point of view, a PIN-type diode. The concentration of the impurities used for doping (P or N) can be uniform along the length of each of the two Bragg mirrors 3020, 3022. Alternatively, the concentration of the impurities can decrease so as to be minimal in the regions in contact with the central region 3021. The evolution of the concentration of the impurities can be linear, logarithmic, or otherwise.Advantageously, the different voltages applied to the optical cavity 302, doped according to one of the configurations described above, correspond to a reverse bias of the PIN junction formed by the optical cavity 302. Such a reverse bias ensures that no significant electric current flows through the quantum dot 301 and disrupts its photon emission. Alternatively, the optical cavity 302 is forward biased, which generates a significant electric current flowing through the quantum dot 301 and is likely to disrupt its operation.

[0052] According to embodiments of the invention, each of the electrical contact pads 304a, 304b, 304c surrounding the upper face of the optical cavity 302 is connected to one of the terminals of an adjustable voltage source 306a, 306b, 306c. Each of the adjustable voltage sources 306a, 306b, 306c is configured to apply an electrical potential difference between the electrical contact pad 304a, 304b, 304c to which it is connected and the electrical contact 305 carried by the lower face of the optical cavity 302 which is common to all the adjustable voltage sources 306a, 306b, 306c implemented. The voltage generated by each of the adjustable voltage sources 306a, 306b, 306c measured between the corresponding electrical contact pad 304a, 304b, 304c and the electrical contact 305 can be positive, zero or negative.

[0053] There figure 9 Figure 300 represents a top view of a single-photon and entangled-photon-pair generation device comprising two optical cavities 302, 310 according to another embodiment of the invention. One of the two optical cavities 302 includes the quantum dot 301 and can be configured according to one of the embodiments described above. The other optical cavity 310 does not include quantum dots and can have opto-geometric parameters (size, shape, etc.) different from those of the first optical cavity 302. The two implemented optical cavities 302, 310 are configured so as to be optically coupled with a given coupling strength. The assembly formed by the two optical cavities 302, 310 exhibits two pairs of degenerate optical modes in polarization.The first pair of polarization-degenerate optical modes is configured to resonate with the transitions between the two-elementary-excitation state (biexciton) and the two-elementary-excitation states (exciton). The second pair of polarization-degenerate optical modes is configured to resonate with the transitions between the two-elementary-excitation states (exciton) and the ground state, also called the neutral state, of the quantum dot 301. The modes of each pair of optical modes exhibit radiation patterns 307 that overlap by 70% or more. The coupling between the two optical cavities 302, 310 can also be configured according to the embodiments described in the patent application [3].

[0054] According to embodiments of the invention, the number of electrical contact pads arranged around the upper face of the optical cavity 302 is greater than three. This number of electrical contact pads may be even or odd and may be less than 20. The electrical contact pads may be arranged asymmetrically around the optical cavity 302 such that at least three electrical contact pads, and optionally their associated semiconductor arms, are not oriented in parallel directions, two by two. This implies that the angular spacing between adjacent electrical contact pads and the radial spacing, measured from the optical cavity 302, may not be constant for all the implemented electrical contact pads.In addition, the electrical contact pads can be placed at different distances from the electrical contact 305, i.e. the electrical contact pads can be placed in different planes perpendicular to the layer stacking axis.

[0055] According to embodiments of the invention, the single-photon and entangled-photon-pair generation device 300 may include a cooling unit, for example, a Peltier cooling unit or a nitrogen or helium cryogenic cooling unit, configured to control and maintain a constant operating temperature of the quantum dot 301. Such an operating temperature may be between 4 kelvins and 100 kelvins. Such low operating temperatures limit the interaction between the quantum dot 301 and its adjacent environment and allow the coherence of the excited state with the ground state to be maintained over a sufficiently long time interval to manipulate the state of the quantum dot 301.

[0056] There figure 10 Figure 400 illustrates the evolution over time of the bright excitonic states in a quantum dot 301 inserted into a micropillar-type optical cavity 302. Both the quantum dot 301 and the optical cavity 302 exhibit non-zero anisotropy. The anisotropy of the optical cavity 302 implies a lifting of the degeneracy of the modes of the optical cavity 302, resulting in a low-energy mode and a high-energy mode aligned with the polarization eigenaxes of the micropillar, H and V. The anisotropy of the quantum dot 301 also implies a lifting of the degeneracy between the two excited states, which become separated by a non-zero amount of energy known as fine-structure splitting. Each of the two excited states, also called eigenstates, of the quantum dot 301 is associated with a given linear polarization state, x or y, in which it can either emit or be excited by a photon.Generally, the modes of the optical cavity 302, H and V, do not coincide with the polarization directions of the quantum dot 301, x and y, and an angle, denoted θ, can be defined to quantify this mismatch. Resonant excitation of the exciton level of the quantum dot 301 by photons linearly polarized along one of the modes of the optical cavity 302, H or V, leads to the excitation of both eigenstates with a weighting that depends on the value of the angle θ. Such a resulting excited state, 4021 or 4020, is not stable, and its phase evolves over time with a transfer rate proportional to the value of the fine-structure splitting.This behavior can be exploited to collect only the single photons generated by the quantum dot 301 by exciting it with photons linearly polarized along the H-axis of the optical cavity 302, for example, and collecting only the photons generated with a linear polarization perpendicular to that used during excitation, the V-axis of the optical cavity 302, in this case. In such configurations, increasing the fine-structure splitting value allows photons to be emitted in the detection polarization direction before they are spontaneously emitted in the excitation polarization direction. Subsequently, the fine-structure splitting value should be decreased to limit the probability of having two photons emitted instead of one within the same excitation period.

[0057] There figure 11 represents a single-photon generation device 300 according to one embodiment of the invention. Such an embodiment implements an anisotropic quantum dot 301 incorporated in a birefringent micro-pillar-type optical cavity 302. The quantum dot 301 is subjected to an electric field adjustable in intensity and orientation through electrical contact pads 304a, 304b, 304c, 305 arranged according to one of the embodiments described above, pad 304c and the associated voltage source 306c being hidden in the figure 11 The single-photon generation device 300 includes an excitatory laser source 501 configured to generate light pulses intended to optically pump the quantum dot 301, thereby changing its state from a neutral to an excited state. The polarization state of the light pulses is linear along one of the eigenaxis of the optical cavity 302, and the associated wavelength is chosen so that the energy carried by the photons is sufficient to excite the quantum dot 301. The duration of the light pulses and their repetition rate determine the rate, in number of photons per second, generated by the single-photon generation device 300.The single-photon generation device 300 further includes a polarization splitter cube 502 and a converging lens 503 configured respectively to isolate the photons generated in the optical cavity 302 according to a linear polarization state in a direction perpendicular to that of the excitation photons and to improve the optical coupling between the polarization splitter cube 502 and the optical cavity 302. The eigenaxes of the optical cavity 302, H and V, are configured so as not to be collinear with the eigenaxes of the quantum dot 301, x and y, by an angle θ, ideally equal to 45 degrees.

[0058] There figure 12 represents a calibration method 600 of an anisotropic quantum dot 301 subjected to an electric field as illustrated in the figure 4 The three spatial components of the resulting electric field are independently adjustable by means of at least three adjustable voltage sources. The method is iterative and consists, at each iteration, of modifying the intensity of at least one of the three components of the resulting electric field (step 601) and then measuring the characteristic parameter(s) of interest of the quantum dot 301 associated with the applied resulting electric field (step 602). The method can be stopped when the desired values ​​of the characteristic parameters of interest are obtained and can return the values ​​of the voltages applied to the different electrical contact points. The first iteration of the method can correspond to a resulting electric field of zero intensity.

[0059] The calibration procedure can be used to identify the operating point of a 300 entangled photon pair generation device. Characteristic parameters of interest for such a device include the fine structure splitting, which should be significantly reduced, ideally set to zero, and the emission wavelengths corresponding to biexciton relaxation and exciton relaxation.

[0060] The calibration method can also be used to find an optimal operating point for a single-photon generation device 300 operating in a static mode in which the fine-structure splitting characterizing the implemented anisotropic quantum dot 301 is statically tuned before the creation of any exciton. The fine-structure splitting must allow a transfer velocity sufficient for the excited state to relax to a linear polarization state perpendicular to that of the exciting photons after a transfer time shorter than the duration of the optical excitation pulse. The fine-structure splitting can further be optimized so as to: that it be greater than a lower limit from which the excited state can relax, according to a linear polarization state perpendicular to that of the exciting photons, after a transfer time less than at least a factor of 3 and preferably at least a factor of 5 compared to the lifetime by spontaneous emission of the excitonic state; that it be less than an upper limit from which the 301 quantum dot can emit more than one photon during the duration of the optical excitation pulse.

[0061] There figure 13A This is a flowchart of a method for generating single photons 700a using the photon generation device 300 according to an embodiment of the invention. In such an embodiment, the fine-structure splitting characterizing the implemented anisotropic quantum dot 301 is statically adjusted. The first step of the method 701a consists of applying an electrical potential difference between each of the electrical contact pads 304a, 304b, 304c and the electrical contact 305 on the underside of the optical cavity 302, so as to generate a resultant electric field to which the quantum dot 301 is subjected. The components of such a resultant electric field are adjusted to optimize the value of the fine-structure splitting as described above.In addition to tuning the wavelength of the single photons generated by quantum dot 301, applying the resulting electric field in the three spatial directions allows the orientations of the eigenaxes of quantum dot 301 to be defined so that they are not collinear with the eigenaxes of the optical cavity 302 in which the quantum dot 301 is inserted. Advantageously, the eigenaxes of quantum dot 301 are oriented at 45 degrees to the eigenaxes of the optical cavity 302.

[0062] The second step of process 702a consists of illuminating the quantum dot 301 with a train of light pulses whose wavelength spectrum and polarization state correspond to one of the modes of the optical cavity 302 in which the quantum dot 301 is inserted. The duration of the light pulses is chosen to be long enough to excite the quantum dot 301 and enable it to emit at least one photon through spontaneous emission.

[0063] There figure 13B This is a flowchart of a method for generating single photons 700b using the photon generation device 300 according to another embodiment of the invention. In such an embodiment, the fine-structure splitting characterizing the anisotropic quantum dot 301 is dynamically controlled. This dynamic control is achieved through modulation of the resulting electric field to which the quantum dot 301 is subjected, so as to be able to modify the value of the fine-structure splitting in a time interval shorter than the spontaneous emission time characterizing the quantum dot 301.

[0064] The first step of the 701b process consists of illuminating the quantum dot with a train of light pulses whose wavelength spectrum and polarization state correspond to a mode of the optical cavity 302 incorporating the quantum dot 301. The duration of the light pulses is chosen so as to be long enough to excite the quantum dot and to be able to emit at least one photon according to the process of spontaneous emission.

[0065] The second step of the 702b process consists of applying an electrical potential difference between each of the electrical contact pads 304a, 304b, 304c and the electrical contact 305 on the underside of the optical cavity 302, so as to generate a resultant electric field to which the quantum dot 301 is subjected. The value of the fine-structure splitting increases as a function of the intensity of the resultant electric field, which must remain below a threshold corresponding to damage to the quantum dot 301—that is, a threshold beyond which the electrical and optical characteristics as determined during the calibration of the quantum dot 301 can no longer be restored. The intensity of the resultant electric field is chosen to be high enough so that the time evolution of the phase of the created excited state is between 45 and 135 degrees, and preferably between 80 and 100 degrees.Such a temporal evolution of the phase must also occur in a shorter time (at least by a factor of 3 and preferably by at least a factor of 5) than the lifetime by spontaneous emission of the excitonic state.

[0066] The third step of process 703b can be initiated immediately after the required time evolution of the phase of the initial excitonic state is complete. This step involves modifying the intensity of the resulting electric field to minimize the energy difference between the two eigenstates of an elementary excitation.

[0067] There figure 14 represents a method for generating entangled photon pairs 800 using a photon generation device 300 according to an embodiment of the invention. Such a photon generation device 300 implements an anisotropic quantum dot 301 incorporated in an optical cavity 302 having at least a first and a second pair of optical modes, each pair of optical modes being degenerate into polarization. The entanglement of the generated photons is achieved through static control of the fine-structure splitting characterizing the implemented anisotropic quantum dot 301. The first step of the method 801 consists of applying an electrical potential difference to each of the electrical contact pads 304a, 304b, 304c and to the electrical contact 305 located on the underside of the optical cavity 302.Such electrical potential differences are optimized to reduce the energy gap between the two states to an elementary excitation defined by the fine-structure splitting of the 301 quantum dot. In particular, step 801 may consist of canceling the fine-structure splitting, making the 301 quantum dot a three-energy-level system relative to the relaxation of a biexciton. The electrical potential differences are further tuned such that: the transition between the two elementary excitation state and the one elementary excitation states either in resonance with the first pair of modes of the optical cavity 302; and the transition between the one elementary excitation states and the ground state either in resonance with the second pair of modes of the optical cavity 302.

[0068] The second step of process 802 consists of populating the two-elementary-excitation state of the quantum dot 301 with an energy input. The creation of such a biexcitonic state XX can be achieved by exciting the quantum dot 301 with two successive photons of suitable energies: A first photon of energy ωX sufficient to excite the transition of a first exciton and a second photon of energy ωXX different from ωX to create a second exciton and thus obtain a biexcitonic state XX, or alternatively two photons of energy (ωXX + ωX) / 2.

[0069] There figure 15 This illustrates experimental measurements of the fine-structure splitting characteristic of an anisotropic quantum dot 301 embedded in a micro-pillar-type optical cavity 302, the upper face of the optical cavity 302 being surrounded by two electrical contact pads 304a, 304b according to embodiments of the invention. The experimental measurements are performed for two voltage ranges applied to the electrical contact pads 304a, 304b and illustrate the technical advantages of the embodiments of the invention in terms of deterministic control of the fine-structure splitting value. Such measurements are obtained for the following configuration: the quantum dot 301 is excited by a laser with an energy greater than the optical transition corresponding to the exciton. This non-resonant excitation populates the two excitonic states 2020 and 2021, which then emit photons at two different wavelengths that are collected.The difference in wavelengths between the two photons allows us to extract the value of the fine-structure splitting. figure 15 This shows the result of such a measurement where the color scale reflects the absolute value of the fine structure splitting, as a function of two voltages applied to contacts 304a and 304b. The figure shows that the fine structure splitting changes in absolute value as a function of the applied voltages and changes sign to zero for a whole range of voltage pairs 1 and voltage pair 2.

[0070] According to one embodiment of the invention, the formation of an exciton in an initially neutral quantum dot 301 is carried out by a non-resonant optical pumping consisting of exciting the quantum dot 301 with a photon of energy greater than the band gap of the material forming the quantum dot 301.

[0071] According to another embodiment of the invention, the geometric shape of the micro-pillar forming the optical cavity 302 defined in the plane perpendicular to the growth axis can be elliptical, polygonal or other.

[0072] The invention is not limited to the embodiments described above by way of non-limiting example. It encompasses all alternative embodiments that could be envisaged by a person skilled in the art. [REFERENCES]

[0073] [1] Trotta, Rinaldo et al. "Highly Entangled Photons from Hybrid Piezoelectric-Semiconductor Quantum Dot Devices." Nano Letters 14.6 (2014): 3439-3444. [2] K Kowalik, O Krebs, A Lemaitre, S Laurent, P Senellart, P Voisin, JA Gaj "Influence of an in-plane electric field on exciton fine structure inInAs-GaAs self-assembled quantum dots." Applied Physics Letters 86 (4), 041907| [3] WO / 2011 / 089336 « SOURCE DE PAIRES DE PHOTONS INTRIQUES EN POLARISATION ET SON PROCEDE DE FABRICATION" [4] N. SOMASCHI ET AL: "Near-optimal single-photon sources in the solid state",NATURE PHOTONICS, vol. 10, no. 5, 7 mars 2016 (2016-03-07), pages 340-345

Claims

1. A device for generating photons (300) comprising a quantum box (301) inserted into an optical cavity (302) of micro-pillar type having at least one optical mode, the quantum box (301) having at least one fundamental state and two states with one elementary excitation, the optical cavity (302) having a bottom face and a top face, the bottom face bearing an electrical contact (305), said device comprising at least three electrical bonding pads (304a, 304b, 304c) that are electrically insulated from one another, arranged around the top face of the cavity (302), characterized in that said device comprises also at least three adjustable voltage sources (306a, 306b, 306c) for applying respective and variable potential differences between each of said electrical bonding pads (304a, 304b, 304c) and the electrical contact (305) borne by the bottom face of the optical cavity (302).

2. The device (300) as claimed in claim 1, wherein the electrical bonding pads (304a, 304b, 304c) are linked to the top face of the cavity (302) by semiconductor arms (308a, 308b, 308c) that are oriented radially with respect to the cavity (302) and that have, in a tangential direction and, at their end closest to the cavity, a width less than that of the cavity (302).

3. The device (300) as claimed in claim 1, wherein the electrical bonding pads (304a, 304b, 304c) are linked to semiconductor arms oriented radially toward the cavity (302) and of which the ends are separated from the top face of the cavity (302) by an empty gap (309), or one that is filled with dielectric of sub-micrometric width.

4. The device (300) as claimed in any one of the preceding claims, wherein the optical cavity (302) of micro-pillar type forms a diode of P-I-N type, the quantum box (301) being situated in an intrinsic region (3021) of said diode.

5. The device (300) as claimed in any one of the preceding claims, wherein the electrical bonding pads (304a, 304b, 304c) are borne by respective pillars (303) decoupled optically and electrically from the cavity (302).

6. The device (300) as claimed in any one of the preceding claims, wherein said optical cavity (302) of micro-pillar type has at least one first and one second pair of modes, each said pair of modes being polarization degenerate, and wherein said quantum box (301) has also one state with two elementary excitations.

7. The device (300) as claimed in claim 6, also comprising a second optical cavity (310) coupled to the optical cavity (302) into which the quantum box (301) is inserted, the geometry of the first cavity and of the second cavity (302, 310) and the force of their coupling being chosen such that the assembly composed of the two coupled cavities has a first pair of modes that are polarization degenerate and resonant with transitions between the state with two elementary excitations and the two states with one elementary excitation of the quantum box (301) and a second pair of modes that are polarization degenerate and resonant with the transitions between the two states with one elementary excitation and the fundamental state of the quantum box (301), the modes of each of said pairs having radiation patterns that overlap by 70% or more.

8. The device (300) as claimed in claim 7, wherein the second optical cavity (310) is also of micro-pillar type, the two cavities (302, 310) being arranged side-by-side.

9. A method (800) for generating entangled photon pairs by means of a device (300) as claimed in one of claims 6 to 8, the method comprising the following steps: - applying a potential difference between each of said electrical bonding pads (304a, 304b, 304c) and the electrical contact (305) borne by the bottom face of the optical cavity (302), said potential differences being chosen to: ∘ minimize an energy difference between the two states with one excitation of the quantum box (301); and ∘ bring the transitions between the state with two elementary excitations and the states with one elementary excitation into resonance with the first pair of modes of the cavity (302), and the transitions between the states with one elementary excitation and the fundamental state into resonance with the second pair of modes of the cavity (302), and - populate the state with two elementary excitations of the quantum box (301) with an input of energy.

10. The method for generating single photons (700a) by means of a device (300) as claimed in one of claims 1 to 6 comprising the following steps: - applying a potential difference between each of said electrical bonding pads (304a, 304b, 304c) and the electrical contact (305) borne by the bottom face of the optical cavity (302); and - lighting the quantum box (301) of the device by a light pulse exhibiting a spectrum and a direction of polarization corresponding to a mode of the optical cavity (302); said potential differences being chosen to modify the energies of the two eigenstates with one elementary excitation in such a way that: ∘ said states are resonant with wavelengths of the spectrum of the light pulse; and ∘ the probability of emission, by the quantum box (301), of a single photon exhibiting a linear polarization orthogonal to that of the lighting is maximized.

11. The method as claimed in claim 10, wherein said potential differences are also chosen to orient specific axes of the quantum box (301) according to intermediate directions between those of two specific axes of polarization of the optical cavity (302).

12. The method (700b) as claimed in one of claims 10 and 11, wherein said potential differences are variable in time, and: - during a first period, during at least a part of which the quantum box (301) is lit, take first values chosen for the energies of the two eigenstates with one elementary excitation to exhibit a non-minimal deviation, the first period having a duration chosen to provoke a change of phase of between 45° and 135°, and preferably to between 80° and 100°, between the two said states; and during a second period, immediately succeeding the first period, take values chosen to minimize the deviation between the energies of the two eigenstates with one elementary excitation.