• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    According to one aspect, the invention relates to a quantum photodetection element of incident radiation in a spectral band centered around a central wavelength λ0, having a front face intended to receive said radiation, and comprising: - one or a plurality of layers of semiconductor material comprising at least one layer of absorbing semiconductor material having a cut-off wavelength λc> λ0, and forming a resonant optical cavity; a structure for coupling the radiation incident with the optical cavity to form: a resonance at the central wavelength λ 0 allowing an absorption greater than 80% in the absorbent semiconductor material layer at said central wavelength; ○ an absence of resonance at the radiative wavelength λrad, where the radiative wavelength λrad is the wavelength for which, at operating temperature, the radiative recombination rate is maximum.
    公开号:FR3015113A1
    申请号:FR1362842
    申请日:2013-12-17
    公开日:2015-06-19
    发明作者:Benjamin Portier；Michael Verdun；Riad Haidar；Jean-Luc Pelouard；Fabrice Pardo
    申请人:Office National dEtudes et de Recherches Aerospatiales ONERA；Centre National de la Recherche Scientifique CNRS；
    IPC主号:
    专利说明:

    [0001] STATE OF THE ART Technical Field of the Invention The present invention relates to a very low-noise quantum photodetection element and a method of manufacturing such a photodetection element. The invention applies in particular to the design of photodetectors in the infrared with very good sensitivity. State of the Art An optical detector converts optical radiation into an electrical signal. This photon-electron conversion can be indirect: in the case of thermal detectors, for example, the absorption of light results in a temperature rise of an absorbent material which is then converted into an electrical signal. On the contrary, photonic (or quantum) detectors involve the direct creation of electric charges that can be either ejected (photoemissive effect) or released within the photosensitive material (in the case of semiconductor materials).
    [0002] Among the semiconductor detectors, there are in particular photoresistances and photodiodes. Photoresistors (or LDRs: Light Dependent Resistor) comprise a layer of semiconductor material; they behave like resistors that let a greater or lesser electric current flow depending on the illuminance. The photodiodes are diode junction diodes with electrical polarization given by an electric current depending on the illuminance. They include, for example, a P-type doped semiconductor and an N-type doped semiconductor which form a junction at their interface (PN junction). N-type doping is achieved by introducing into the crystal electron donor impurities that increase the density of negative charges (electrons). P-type doping is achieved by electron-accepting impurities that increase the density of positive charges (holes). Figure 1A shows the energy pattern of electrons in a PN junction. In the vicinity of the junction, there is a depletion zone (absence of free carriers) also called space charge zone (ZCE). Photons of energy greater than the energy of the forbidden band of the semiconductor ("gap"), that is to say of wavelength lower than a so-called cut-off wavelength (X,), are absorbed and induce the passage of an electron from the valence band to the conduction band which results in the appearance of a pair of additional mobile charges: an electron in the conduction band and a hole in the valence band . In the zone of 1 charge of space, under the action of the electric field which reigns in this zone, the hole will be accelerated towards the zone of type P and the electron towards the zone N, each one of them adding to the majority holders of these areas. This results in a separation of the electron-hole pair and an electric current flowing in the device from the N-type zone to the P-type zone, the photocurrent, which increases the reverse current of the diode. This scheme is sometimes improved by the addition between the two zones of a non-intentionally doped layer (called by abuse of language "Intrinsic") making it possible to increase the thickness of the ECA so as to make it closer to the absorption length of photons in the semiconductor ("PIN" junction).
    [0003] FIG. 1B shows the current / voltage characteristic of a photodiode illuminated by a luminous flux E that can take several values E1, E2, E3 such that E1 <E2 <E3. Intensity I corresponds to the intensity of the electric current passing through it as a function of the voltage V at its terminals. As illustrated in FIG. 1B, the particularity of a photodiode is that its current / voltage characteristic depends on the luminous flux that it receives. In the absence of illumination, the characteristic is that of a junction diode with in particular a zero current for a zero applied voltage. As the luminous flux increases, there is an increase of the current towards the negative intensities, in particular for the negative voltages. According to the conventions chosen for the orientation of I and V, if the product P = I * V is positive, the dipole receives energy and functions as a receiver. If the product is negative, the dipole provides energy; he behaves like a generator. The generator quadrant corresponds to the photovoltaic domain. When the dipole operates as a receiver, the device is characterized by its detectivity: 1 2h, riext with rj, '(the quantum efficiency of the device and jobs the dark current density, the dark current being defined as the electric current The main sources of dark current are, on the one hand, intrinsic sources, such as the diffusion current of the minority carriers and the radiative and non-radiative generation-recombination current of the charges, and on the other hand, sources related to the manufacturing process, such as the non-radiative recombination currents on the semiconductor defects (so-called SRH current), which are particularly intense in the vicinity of the surfaces or interfaces. is proportional to the gradient of their density, itself imposed by the boundary conditions at the edge of the ECZ on the one hand and the 2 electrical contact On the other hand, when the electrons or holes are not evenly distributed in the semiconductor, their movement is in a direction which tends to standardize their spatial distribution. In a PN junction, the distribution of the minority carriers in the doped zones is not uniform, hence a diffusion of these carriers. The resulting current is called the diffusion current. It has been highlighted in the article by S. R. Forrest et al. in 1980 (see "In0,53Ga0,47As photodiodes with dark current limited by generation-recombination and tunneling" Appl.Phys.Lett 37 (3), 1980) a large reduction in the diffusion current by means of "barrier layers" in InP in InP / InGaAs type heterojunctions.
    [0004] FIG. 2 thus describes a quantum detector according to the prior art, comprising a PIN type heterojunction 1 composed of three layers 14, 10, 12 of semiconductor material, for example an InP / InGaAs type heterojunction. The layer 10, for example made of InGaAs, situated between the layers 14 and 12, for example made of InP, is a layer of absorbent semiconductor material in the spectral band of interest and forms the zone I of the PIN junction. The thickness of the absorbent layer 10 is about 2 μm or more, to absorb the entire incident radiation. The layers 14 and 12 located on either side of the zone I, have bands of forbidden energy higher than the band of forbidden energy of the zone I, thus forming layers known as "barrier layers" which make it possible to limit strongly the diffusion current in the heterojunction, as has been demonstrated in SRForrest et al. P-type zones 16 are obtained by diffusion or ion implantation in the barrier layer 14. Each P-type zone thus defines a pixel of the detector. Passivation layers 6 on the rear face make it possible to limit the effects of the crystallinity defects at the origin of the SRH (Shockley-Read-Hall) current in the space charge zone. A layer of conductive material 3, for example a metal layer, makes it possible to form an electrical contact at each pixel. Each electrical contact 3 is in contact with a reading circuit 9 via an indium ball 7. The electrical contact 3 represents for each pixel the anode of the electric circuit formed by the heterojunction. The cathode of the circuit is formed by a remote contact 5 in electrical contact with the barrier layer 12 and allowing contact with the read circuit. The barrier layer 12 forming the front face FF intended to receive the incident radiation is covered with an antireflection layer 8 in order to transmit to the heterojunction the maximum of the incident optical flux (represented by arrows in FIG. 2). It can be shown that in the low diffusion current heterojunctions as described in Figure 2, the generation-recombination current becomes predominant. It is said that there is "generation" when an electron occupying a state in the valence band makes a transition to an unoccupied state of the conduction band. It is a process that occurs spontaneously at non-zero temperature. At thermodynamic equilibrium, the thermal generation is exactly compensated by an antagonistic process, called "recombination": an electron of the conduction band transits to the valence band. There are several generation-recombination processes in semiconductors. Radiative and non-radiative processes are usually distinguished according to whether the energy of the electron-hole pair is dissipated by the emission of a photon (radiative) or phonon (non-radiative). All these generation-recombination phenomena induce a parasitic current in the structure, called the generation-recombination current. The problem of reducing the dark current in a heterojunction with a low diffusion current is still relevant, as the article by J. A. Trezza et al. in 2011 (SPIE Proceedings, 8012: 80121Y-80121Y-12, 2011), in which analysis and estimation of dark currents are detailed. In inverse polarization, it is shown that the dark current density is proportional to the generation-recombination rate of the charges and to the thickness of the space charge zone, in which the generation-recombination phenomena take place.
    [0005] It should be noted that the generation-recombination process also exists in the photoresistances, at the level of the semiconductor material layer, creating in the same way a parasitic dark current. To reduce the dark current, it can be sought to reduce the recombination-generation rate while keeping the diffusion current low, for example by cooling the detector, which is a costly solution in energy, or by optimizing the electronic structure . One can also seek to reduce the thickness of the space charge area. However, the decrease in the thickness of the space charge area also results in a decrease in detector absorption. It is known to use optical resonators to concentrate the energy and to compensate for the decrease in the thickness of the space charge zone, in the field of infrared detection (see for example EP patent 2,276,072) or in the field of photovoltaics (see for example Polman, A., & Atwater, HA (2012), Thotonic design principles for ultrahigh-efficiency photovoltaics, Nature materials, 11 (3), 174-177). None of these 4 documents, however, discloses photodetection elements allowing a reduction of the generation-recombination current, especially in junctions with a low diffusion current. An object of the present invention is to provide a detector having optimized detectivity through a reduction of the dark current; this objective is achieved by means of an optical resonator whose particular arrangement allows a drastic reduction of generation-recombination phenomena. SUMMARY OF THE INVENTION According to a first aspect, the invention relates to a quantum photodetection element of incident radiation in a spectral band centered around a central wavelength X0, having a front face intended to receive said radiation. and comprising: - one or more layers of semiconductor material comprising at least one layer of absorbing semiconductor material of cut-off wavelength λ> λ 0, and forming a resonant optical cavity; a structure for coupling the radiation incident with the optical cavity to form: a resonance at the central wavelength λ 0 allowing absorption greater than 80% in the layer of absorbent semiconductor material at said central wavelength; o absence of resonance at the radiative wavelength λ 1 -ad, where the radiative wavelength λ 1 -ad is the wavelength for which, at operating temperature, the radiative recombination rate is maximum.
    [0006] Electrical contacts allow the polarization of the photodetection element. The wavelength Xi-ad, called the radiative wavelength in the present application, is the wavelength for which the radiative recombination rate is maximum. It verifies the inequality X0 <rad <X. Ensuring a lack of resonance at the radiative wavelength, one ensures a limitation of the absorption of the photodetection element to krad On can then radically reduce the generation-recombination current, the emissivity varying as absorption, according to Kirchoff's law. It is also possible to work at a wavelength λ o further from the cutoff wavelength by providing reduced absorption at the radiative wavelength, the reduction to be greater than exp (AEg / kT) where Δg is the variation of the energies corresponding to the wavelengths Xi-ad and, alternatively, the coupling structure of the radiation incident with the optical cavity is adapted to produce a minimum of absorption in the layer of absorbent semiconductor material at the length radiative wave. According to one variant, this minimum absorption results from an antiresonance in the cavity. In the structures of the prior art, the cut-off wavelength is generally chosen very close to the central wavelength of the detection spectral band in order to limit the rate of radiative recombination and hence the current of darkness. According to the present description, the cut-off wavelength is intentionally chosen with a given length difference with the central wavelength of the detection spectral band, the optical cavity having a non-resonance at the radiative wavelength, whose value is strictly between the central wavelength of the detection spectral band and the cut-off wavelength. Advantageously, according to one variant, the coupling structure of the radiation incident on the optical cavity is adapted to reduce the absorption in the absorbing semiconductor material layer at the radiative wavelength λ 1 -ad, the said reduction being greater than exp ( AEg / kT) where AEg is the variation of the energies corresponding to the wavelengths Xi-ad and X. The reduction of the absorption is understood here by comparison with an identical detection element in which there is no structure coupling. Advantageously, the resonance at the central wavelength of the detection spectral band is located in the absorbent semiconductor material layer, making it possible to limit the absorption at the surface and on the interfaces.
    [0007] Advantageously, the photodetection element comprises a metal reflective layer arranged on one side of the optical cavity, opposite the side supporting the front face. According to one variant, the metal reflective layer also serves as an electrical contact for the photodetection element. Advantageously, the resonant optical cavity is a waveguide and the coupling structure comprises a coupling network on the front face and / or a coupling network on the rear face, that is to say on the optical cavity side. opposite the side supporting the front face. The coupling network (s) may be of one or two dimensions, formed by an alternation of dielectric / dielectric materials (dielectric materials having refractive indices with different real parts) or dielectric / metal. The dielectric material may be an insulator or a semiconductor. The coupling network (s) may be limited to the dimensions of a pixel or may be extended over a set of pixels.
    [0008] In the case where the optical cavity is a waveguide, the central wavelength resonance X0 in the layer of absorbing semiconductor material is a guided mode resonance. Advantageously, the waveguide has antiresonance at the radiative wavelength λ-ad in the layer of absorbing semiconductor material, resulting from destructive interference leading to a quasi-zero absorption.
    [0009] The thickness of the waveguide determines a number of resonances in the guide. Advantageously, the thickness of the waveguide is between X0 / 8n and X0 / n, advantageously between X0 / 4n and 3X0 / 4n, where n is the mean value of the real parts of the refractive indices of the layers of semiconducting material. conductor forming the waveguide. By choosing a sufficiently large thickness, it is ensured to have in the absorbent semiconductor material layer at least one resonance at the desired wavelength. By limiting the thickness, the number of resonances in the layer of absorbing semiconductor material is limited, which makes it possible to space the resonant wavelengths and to facilitate the appearance of a "resonance node" at krad. According to one variant, the coupling network (s) comprise a periodic structure having a period between X0 / n and X0 / n1, where n1 is the real part of the refractive index of the radiation incidence medium ( for example air). These conditions over the period of the network (s) (x) can limit the number of diffraction orders in the stack of layers forming the waveguide, specifically to have only the diffraction order 0 in the incident medium and the 0, -1 and +1 diffracted orders in the waveguide, again allowing spacing of the resonant wavelengths. This configuration also allows a very good angular tolerance of the detection element. According to one variant, the coupling structure comprises "metal dielectric metal" structures otherwise called MDM. According to one variant, the photodetection element comprises a stack of semiconductor material layers forming a PN or PIN type heterojunction, said stack comprising the layer (s) of semiconductor material forming the optical cavity. Advantageously, the heterojunction comprises barrier layers, allowing a strong reduction of the diffusion current. According to a variant, at least one of the barrier layers is structured over at least a portion of its thickness to form a coupling network with the optical cavity. Applicants have shown that this configuration can further reduce the dark current due to the reduction of the active surface.
    [0010] According to one variant, the photodetection element according to the first aspect is suitable for detection in one of the bands I, II or III of the infrared. According to a second aspect, the invention relates to an infrared detector comprising an assembly of quantum photodetection elements according to the first aspect. According to a third aspect, the invention relates to methods of manufacturing a quantum detection element according to the first aspect. According to one variant, the method comprises: the formation on a substrate of an epitaxial structure, comprising a stack of layers of semiconductor material, of which at least one layer of absorbing semiconductor material of cut-off wavelength Xc > Xo, and an upper layer of dielectric material, - The structuring of the upper layer of dielectric material to form a coupling network, - The deposition on said coupling network of a metal reflective layer, - The removal of the substrate to form the front face of the detection element.
    [0011] Advantageously, the formation of the epitaxial structure comprises an organometallic epitaxy. BRIEF DESCRIPTION OF THE DRAWINGS Other advantages and characteristics of the invention will appear on reading the description, illustrated by the following figures: FIGS. 1A and 1B (already described), electron energy diagram in a PN junction and characteristics current / voltage of a photodiode as a function of illumination; - Figure 2 (already described), example of detector according to the prior art; FIGS. 3A and 3B, examples of detectors according to the present description; 8 - FIGS. 4A to 4C, examples of coupling structures in detection elements according to the present description; FIGS. 5A to 5E, examples of coupling networks with one or two dimensions adapted to a detection element according to the present description; FIG. 6, diagram illustrating the main elements of the optical resonator implemented in a detection element according to an example embodiment, in which the optical resonator forms a waveguide; FIG. 7, block diagram illustrating the wave propagation paths in a waveguide of the type of that described in FIG. 6 according to the diffracted orders 0 and ± 1 (in normal incidence); FIG. 8, absorption curves as a function of the wavelength in a detection element implementing a resonator of the type of that described in FIG. 6, in the case of an InP / InGaAs type heterojunction; FIGS. 9A, 9B, calculated images representing the distribution of the electromagnetic field at the central wavelength of the polarization detection spectral band TE and TM, respectively, in the layers of FIG. 6 and FIGS. 9C, 9D, calculated images representing the distribution of the electromagnetic field at the radiating wavelength in TE and TM polarization respectively, in the layers of FIG. 6; FIG. 10, diagram illustrating in an exemplary detection element according to the present description, an InP / InGaAs type heterojunction; FIG. 11, curves representing the diffusion current in an InP / InGaAs type heterojunction, compared with the diffusion current in InP and InGaAs type junctions respectively; FIG. 12, curves representing the sum of the diffusion current and the radiative generation-recombination current in an InP / InGaAs type heterojunction, compared with the diffusion current in InP and InGaAs type junctions respectively; 9 - FIGS. 13A and 13B, a diagram showing a mapping of the electrostatic potential in a photodetection element of the type of FIG. 10, and a curve showing the radiative recombination ratio calculated in inverse polarization as a function of the position in the element of photodetection according to a cross section S of Figure 13A; FIG. 14, electroluminescence spectrum determined experimentally in a heterojunction of the type shown in FIG. 10; 15, curves respectively showing the absorption spectrum of the optical cavity formed by the structure shown in FIG. 10, the electroluminescence spectrum of the heterojunction, the luminance of the cavity resulting from the product of absorption spectra and electroluminescence; 16A to 16G, diagrams showing different steps of an exemplary method according to the present description. DETAILED DESCRIPTION In the figures, the identical elements are referenced by the same references. The figures are given for illustrative purposes and the scales are not necessarily respected. FIGS. 3A and 3B show two examples of quantum detectors comprising quantum photodetection elements according to the present description. As in the example illustrated in FIG. 2, the detector comprises a set of photodetection elements (Pi, P2,...), Arranged for example in a matrix manner in a focal plane of an imaging optics (no shown). In FIGS. 3A and 3B, the inter-pixel distance is overestimated for reasons of readability of the figures; The photodetection elements are suitable for the detection of incident radiation in a spectral band centered around a central wavelength λ 0, for example in one of the bands I, II or III of the infrared, corresponding to the spectral bands. of atmospheric transparency. The detector comprises a front face FF intended to receive the light radiation, indicated by arrows in FIGS. 3A and 3B, as well as a reading circuit 9, arranged on the opposite side to the front face. Each photodetection element comprises a first electrical contact 3 connected to the reading circuit for example by an indium ball 7 to form as many "pixels" of the detector. A second electrical contact 5, common to all the pixels, makes it possible to close the electrical circuit of each of the photodetection elements. Each photodetection element comprises, in these examples, a stack of semiconductor material layers forming a junction 1 of the PN or PIN type, with in particular a layer 10 of absorbing semiconductor material of cut-off wavelength λ> λ 0. Advantageously, the stack of layers of semiconductor material forms a PIN type heterojunction, the layer 10 of absorbent semiconductor material forming the so-called intrinsic zone I intentionally undoped. The heterojunction, examples of which will be described in more detail later, comprises in these examples other layers of semiconductor material, including barrier layers 12 and 14, arranged on either side of the layer 10. layer 6 can have two complementary roles. On the one hand, it can serve as a passivation layer when the ECZ reaches the surface of the detector. This is the case, for example, when the P + doped zones are produced by diffusion or local ion implantation of dopants (for example zinc) generating a ZCE at the interface of these zones and of the N type material. case when the pixels are isolated by deep etching of a mesa. On the other hand, it can serve as protection or encapsulation against the aggressions of the external environment. This is particularly useful during the hybridization process but is also useful in reducing the long-term aging effects of the detector. In the present description, the pixels are advantageously electrically isolated from each other, either by etching all or part of the barrier layer 14 between the pixels, or, as will be described in more detail below, by atomic hydrogen diffusion. in the barrier layer between the pixels, which has the effect of making these areas highly resistive electrically. Since the ECZ is not in contact with the surface of the device, the manufacturing method is thus facilitated and the SRH type related peripherals currents are reduced compared to the known methods according to the prior art. The layer 6 may, however, be retained for its role of protection determining the aging of the detector. In the examples shown in FIGS. 3A and 3B, each photodetection element furthermore comprises a metal reflective layer 30 arranged on a face opposite to the front face FF intended to receive the incident radiation, as well as a coupling structure 4, by for example a coupling network formed of a layer of structured dielectric material. The set of layers of unstructured dielectric material, located between the reflective layer 30 and the input face, are advantageously chosen to have refractive indices whose real parts are close. They form a resonant optical cavity 2, for example a waveguide, of given thickness e. The coupling structure 4 of the incident radiation with the optical cavity, formed in the case of Examples 3A and 3B by a coupling network, is adapted to form a resonance at the central wavelength X0 corresponding to a quasi-total absorption ( > 80%) localized in the layer of absorbent semiconductor material 10 and reduced absorption in the layer of absorbent semiconductor material at an Xi-ad wavelength, called the radiative wavelength in the present description, and for which the radiative recombination rate is maximal.
    [0012] The radiative recombination rate R in the layer of semiconductor material is given by the following relation: R (1) where E = hc /), is the energy of the photons emitted at the wavelength λ (h constant of Planck and C speed of light), Eg is the bandgap energy (or gap) of the absorbing semiconductor material (Eg = hc / λt), k is the Boltzmann constant and T is the temperature. Thus, the energy of the incident photons Erad for which the rate of radiative recombination is maximum is the energy for which the derivative of equation (1) vanishes, ie: Kr Erad = 7 Eg This corresponds to a length of X-wave, = -rad, called "radiative wavelength" in the following description, such that: kT 2hc (3) In practice, we choose a semiconductor material absorbing cut-off wavelength X ., strictly greater than the central wavelength of the detection spectral band sought for the photodetection element. At ambient temperature, the radiative wavelength, which is strictly less than the cut-off wavelength while remaining very close, is therefore strictly included between the resonant wavelength of the optical cavity and the length X. cut-off wave, absorbing semiconductor material. The applicants have thus shown, as will be described later, a drastic reduction in the generation-recombination current when the reduction of the absorption (therefore of the emission) at Xi-ad is greater than exp (AEg / kT). where AEg is the energy difference between X, and Xi-ad. Due to the implementation of a resonant optical cavity, the layer of absorbing semiconductor material can be reduced; it may for example be between a few (2) 12 tens and a few hundreds of nanometers for an InGaAs layer, making it possible to gain a significant factor on the reduction of the thickness with respect to the structures of the prior art. This reduction in thickness results in a reduction in the thickness of the ZCE which leads to a reduction in the same ratio of the radiative recombination currents on the one hand and SRH-type non-radiative recombination currents on the other hand . This results in a significant reduction (approximately an order of magnitude) of the dark currents. Furthermore, an antireflection layer is no longer necessary on the front face FF; on the contrary, it is freed to maximize reflection at the interface of the dielectric material layer 12 with the incident medium (for example air), to generate resonances in the optical cavity. Advantageously, when the optical cavity is formed of several layers of dielectric material, the real parts of the refractive indices differ by less than 20% to avoid parasitic reflections at the interfaces.
    [0013] In the examples of FIGS. 3A and 3B, the reflective layer 30 of metallic material also contributes to forming the electrical contact 3 which represents for each pixel the anode of the electric circuit formed by the heterojunction. The electrical contact 3 allows contact with the indium ball. It furthermore optionally comprises a layer 31 for hybridization with the indium ball. As in the example of Figure 2, the cathode of the circuit is formed by a remote contact 5, the electrical contact 5 comprising for example a layer of conductive material 50 in electrical contact with the barrier layer 12 and a layer of conductive material 51 allowing contact with the read circuit 9. As shown in FIG. 3A, the coupling structure may comprise a network on the rear face. The structuring can be carried out in the metal layer to form the coupling network or in a layer of semiconductor material, for example the barrier layer 14. As illustrated in FIG. 3B, the coupling structure can comprise a network of coupling on the front face. The coupling structure can be obtained by deposition of a network on the layer of semiconductor material 12, as illustrated in FIG. 3B, or directly by structuring of the semiconductor material layer 12. The coupling structure may also include a network on the front and a network on the back. Other coupling networks are possible, for example dielectric metal networks. Furthermore, the coupling network (s) may be limited to the dimensions of a pixel (FIG. 3A) or may be extended over a set of pixels (FIG. 3B). FIGS. 4A to 4C thus illustrate different variants of coupling structures suitable for producing a photodetection element according to the present description.
    [0014] In these figures, for the sake of simplification, the dielectric material layer or layers forming the optical cavity 2 are reduced to the single layer 10. Only the reflective layer 30 is represented, which is in all cases arranged on the rear face. The example of FIG. 4A shows a coupling structure with a single coupling network on the front face. The network 4 is transparent or semi-transparent to let at least a portion of the radiation, symbolized in Figure 4A by the wave vector k. This is for example a network formed of alternations of two dielectric materials 41 and 43, semiconductor, or insulator, or air, of different indices. Alternatively, it may be an alternation of a metallic material and a dielectric material, semiconductor, insulator or air.
    [0015] Figure 4C shows a coupling structure with a single backplane coupling network. The network 4 consists for example of a network formed of alternations of two dielectric materials 42, 44, at least one of which is conductive; it can also be an alternation between a metallic or semiconductor material, and a dielectric material, semiconductor or insulating type, or air.
    [0016] FIG. 4B represents an example of a coupling structure comprising two networks 4A and 4B respectively at the front and at the rear. As in the example of FIG. 4A, the network on the front face is transparent or partially transparent. The network on the rear face comprises at least one conductive material. The plurality of networks (2 or more) makes it possible, in particular, to facilitate the generation of antiresonances.
    [0017] FIGS. 5A to 5E illustrate, in the case of coupling networks, several possible forms of networks, with one or two dimensions. As an alternative to the coupling networks illustrated by means of FIGS. 3 to 5, the coupling structures may comprise metal-dielectric-metal (MDM) type structures, as described, for example, in patent application FR 2959352. Note that in FIG. in the latter case, periodic structuring is not necessary but rather motivated for technological reasons. Indeed, MDM-like structures act individually as nano-antennas. FIG. 6 illustrates, in a variant, the optical diagram of a detection element according to the present description implementing a waveguide-type optical cavity with a rear-face coupling network. The optical diagram of the detection element thus illustrated comprises a region A corresponding to the medium in which the light radiation is incident (for example air), a region B corresponding to the waveguide, a region C corresponding to the network. coupling and a region D corresponding to the metal mirror. Structurally, the sensing element comprises one or more layers of unstructured semiconductor material (s) 10, 12, of thicknesses hsi, 11132, forming the waveguide. The set of layers, of near refractive indices, is optically assimilated to a single layer B of given average index. The detection element furthermore comprises a periodic network (region C), with one or two dimensions, of thickness hc, constituted, for example, on the one hand by the same semiconductor material (s) (s). ) 44 that the waveguide, and secondly an electrically insulating dielectric material 42. The network is characterized by a period p and a filling factor w / p. Finally, the detection element 15 comprises a reflective layer 30 (region D), advantageously a layer of metallic material serving both as an ohmic contact and an optical mirror (for example, gold). The resonances used are guided mode resonances. They take place in the waveguide, and are obtained thanks to the coupling between the incident wave and the modes of the guide. The coupling is ensured by the periodic network. A thin layer of semiconductor surrounded by materials of lower optical index behaves indeed as a waveguide: the wave can propagate in guided modes, characterized by a phase variation equal to 0 (modulo 2n) after a round trip in the layer (ie between the AB and B / C interfaces). There is resonance when the incident wave is coupled to one of these modes. The incident energy is then stored in this mode (the photons make several round trips) which increases the components of the electromagnetic field in the guide and consequently the absorption probability (proportional to the square of the electric field) of the photons as well. trapped. To couple the incident wave to the modes of the waveguide, coupling structures are introduced, periodic or not, on one side and / or the other of the guide. In the example of FIG. 6, the coupling is carried out by periodically structuring the rear face of the waveguide so that at least one order diffracted by the grating coincides with at least one mode of the guide. According to one variant, the period is chosen so that only three orders (0, +1 and -1) are diffracted by the network in the waveguide. FIG. 7 illustrates the exchanges between the incident wave and the two diffraction modes (mode associated with the order 0, and mode associated with the symmetrical combination of the orders +1 and -1) in normal incidence in the guide of FIG. wave in the case where the three orders (0, +1 and -1) are diffracted by the grating in the waveguide. In this case, the phase variation associated with a round trip of a diffraction mode in the layer is equal to 0 (modulo 2 n), ie: (p (rA, B (m) exp (2ikz (m) h) rc / B (m) = 0 [2n], where m is the diffraction mode of interest, and the terms rA, B (m) and rc / B (m) are the reflection coefficients at the AB interfaces and B / C of the diffraction mode In Figure 7, the resonance is associated with a coupling of the mode 0 with the modes (+1, -1) of the waveguide.The superposition of the resonances in TE and TM is advantageously obtained by adjusting the phase associated with the reflection coefficient rc / B (m) for the two polarizations with the network parameters (period p, fill factor w / p, and thickness hc).
    [0018] It is thus possible by modifying the optical parameters of the structure to obtain a maximum of absorption (resonance) at the central wavelength of the spectral band of interest and, on the contrary, a low absorption, advantageously a minimum of absorption (antiresonance) at the radiative wavelength. Advantageously, it is sought in the optical resonator of the photodetection element according to the present description, to obtain a Fabry-Perot resonance in the waveguide at the central wavelength of the spectral band of detection and to confine. laterally the electric field at resonance in the waveguide with the orders +1 and -1 diffracted by the network. By definition, the Fabry-Perot resonance occurs when the phase accumulated by the fundamental mode in the waveguide after a round trip between the two interfaces (AB and B / C in FIG. 7) is equal to 0 (modulo 2ft). It is characterized by the appearance of standing waves in the waveguide, with successions of bellies (field maxima) and nodes (minimum of field) in the vertical direction. The objective is to have at least one resonance belly in the absorbing layer in order to have a significant absorption, which corresponds to a Fabry-Perot resonance at the order 1. To the order 0 indeed, there is only one resonance belly, located at the air / waveguide interface; at order 1, there is always this resonance belly, with a resonance belly in addition to X, / 2n of the interface. It is this second belly that one wishes to have at the level of the active layer. To reach this resonator configuration, it is possible to choose a waveguide thickness of the order of 3 X0 / 4n, where n is the index of the medium composing the waveguide. For the network, as we need the orders 0, +1 and -1, we can then set the period halfway between X0 / n and Xo. In this way, the orders 0, +1 and -1 propagate in the waveguide (because p> Xo / n), and the orders +1 and -1 are in total internal reflection at the interface with the air (because p <X0), and may have a significant contribution to the distribution of the electric field. For the other parameters of the network, one can take a filling factor of 50% and a thickness of the order of X0 / 4n. Under these conditions, the Fabry-Perot resonance sought in polarization TE or TM is at a wavelength close to X0. To identify it, we can either check that the resonance is independent of the angle of incidence, or verify that the phase associated with a round-trip mode 0 in the waveguide is equal to 0 (modulo 2pi ) to the resonance. For this, we must study the phase associated with the product (p "rm" p "rA / B"), which corresponds to the gain of order 0 after a round trip in the layer, according to the notations of Figure 7. Once the resonance has been identified for each polarization (TE and TM), it is a question of adjusting the thicknesses of the waveguide and the grating, as well as the period and the filling factor of the grating, in order to superimpose the resonances in TE and TM polarization at X.0. At the end of this step, there is therefore a resonance peak at Xo for each polarization, resulting from the same resonance mechanism. The electric field in the structure is then characterized by a first resonance belly at the air / waveguide interface, and a second resonance belly at X0 / 2n under the first resonance belly. The two bellies are confined laterally in a half-period of the network, and are located above the conductive part of the network (metal or semiconductor). For a classical Fabry-Pérot resonance where the wave propagates in the cavity only in the 0 mode, the electric field is invariant in the horizontal direction (x). The confinement obtained is explained by the presence of an electric field associated with the symmetrical combination of orders +1 and -1 in addition to the electric field associated with mode 0. The sum of these two field terms (constant according to x for mode 0, varying as cos (27cx / p) for mode + -1, where p is the period of the network), gives rise to a maximum of field at x = x0 and a minimum of field at x = x0 + p / 2, where x0 is the middle of the semiconductor part of the network. Lateral confinement of the electric field above the conductive parts of the network further reduces the dark current. Indeed, the applicants have shown that the structuring of the rear face of the heterojunction can render electrically inactive 17 areas located at the engraved portions. As a result, these zones do not generate dark current, ie for a fill factor of 50%, a reduction of the total dark current of 50%. However, photons absorbed in this region do not generate photocurrent. To maintain high quantum efficiency, it is advantageous for all photons to be absorbed into the electrically active portions of the heterojunction. This is the case of the resonance presented here. FIGS. 8, 9A and 9B respectively represent the absorption spectrum and the maps of the electric field 1E12 at the resonance wavelength in polarization TE (FIG. 9A) and TM (FIG. 9B), in the different regions A, B , C, D of a photodetection element whose simplified optical scheme is shown in FIGS. 6 and 7. The parameters are optimized for operation of the photodetection element in the infrared band I, around 1, 5 i_tm, the set of semiconductor material layers forming an InGaAs / InP type heterojunction. More precisely, the numerical simulations are obtained with an electromagnetic simulation code, based for example on the modal B-Spline (BMM) method, or the Fourier modal method (FMM). The simulations were carried out taking as indices n (InP) = 3.2, n (dielectric) = 2, n (air) = 1. The InGaAs and gold indices are calculated respectively at using the theoretical formulas given in the papers (Sadao Adachi, Optical Dispersal Relationships for GaP, GaAs, GaSb, InP, InAs, InSb, AlxGal-xAs, and Inl-xGaxAsyP1 and Journal of Applied Physics, 66 (12): 6030- 6040, 1989) and Aleksandar D. Rakic, Aleksandra B. Djurigic, Jovan M. Elazar, and Marian L. Majewski. Optical properties of metallic films for vertical-cavity optoelectronic devices. Applied Optics, 37 (22): 5271-5283, 1998.). The other optical parameters of the structure are given in Table 1 below. Parameters Values (nm) hBl 61 11132. 315 hc 200 w 400 P 800 Table 1: Dimensions of the optical structure layers 18 The numerical simulations are carried out with the BMM code, based on the modal B-Spline method (see P.Bouchon et al., "Fast modal method for subwavelength gratings based on B-spline formulation", Journal of the Optical Society of America A, 27 (4): 696702, 2010).
    [0019] Numerical simulations made it possible to show that this type of structure has, for the TE and TM polarizations, an almost perfect absorption at X = 1.5 μm and a reduced absorption at kmd = 1.62 FIG. 8 represents the absorption spectra 110, 114 total calculated respectively in TM and TE mode for a device having the dimensions as described in Table 1 and in normal incidence. An almost perfect absorption, ie a resonance, is observed in TE and TM at X = 1.5 μm. In this same figure are represented absorption spectra 112, 116 calculated solely in the layer 10 of semiconductor material. The absorption spectra in the optical cavity as a whole and the absorption spectra only in layer 10 have similar behavior for wavelengths below 1.4m. The layer 10 in semiconductor absorbs and thus generates the photocurrent. The few differences are due to absorption losses in metal parts. 80% of useful absorption is thus observed in 315 nm of InGaAs at X0 = 1.5 [lm, corresponding to a gain of 2.5 on the detectivity compared to a non-resonant absorption. In the state of the art, the radiation is absorbed in a single path, with 80% absorption for a semiconductor layer of 2 μm thickness. In the device described in the present application, the absorption is the same as that obtained in the state of the art but the dark current is 6 times lower; there is a gain of a factor of 2.5 on the detectivity. As previously described, guided mode resonances are associated with the appearance of standing waves, resulting from the combination of counter-propagating waves (horizontal and vertical) in the waveguide. These interferences are characterized by local extrema of the electric field. FIGS. 9A and 9B illustrate, for the TE and TM polarizations, the maps of the electric field 1E12 for the resonance observed at 1.5 μm from FIG. 8. The absorption, proportional to 1E12, is effective only if the semiconductor conductor has a sufficiently low bandwidth (E g) (), <, where)., = 1.24 / Eg). FIGS. 9A and 9B show that the two resonance bellies (in white in the figures) are confined laterally in a half-period of the grating, and are located above the conducting part of the network (metal or semiconductor) in the layer of absorbent semiconductor material. FIGS. 9C and 9D illustrate, for the TE and TM polarizations, the maps of the electric field 1E12 for the radiative wavelength 21-ad = 1.62 i_tm. A contrario is observed an absence of the electric field in the absorbent semiconductor material layer at the radiative wavelength, resulting from the non-resonance at this wavelength.
    [0020] FIG. 10 represents a diagram illustrating a particular example of a heterojunction, for example an InP / InGaAs type heterojunction adapted for infrared band I detection, structured in such a way that one of the barrier layers of the heterojunction has the function of coupling network in the simplified optical scheme shown for example in Figure 6.
    [0021] In this example, more specifically, the heterojunction comprises a P-type zone (set of layers 44, 17, 15, 13), an intrinsic zone I (10) and an N-type zone (set of layers 11 and 12) to form a PIN junction. The intrinsic zone is an absorbing layer made with a semiconductor whose bandgap is smaller than the energy of the incident photons. Preferentially, the incident wavelength λ 0 is smaller than the radiative wavelength λ rad (see equation (3) above) in order to have better absorption but also to be able to differentiate spectrally absorption and radiative emission. The intrinsic area 10 is the active area of the detection element. For example, the intrinsic area 10 may be formed of an InGaAs layer having a band gap energy of 0.74eV for 0.8eV detection. The P-type and N-type zones, called quasi-neutrality zones (ZQN), are composed of highly doped semiconductor layers P and N, respectively. These zones may in particular be formed of several layers, one of which is thicker, forming the body layer, respectively 15 and 44 for the N type zone and 12 for the P type zone. The body layer of the ZQN preferably has a greater forbidden bandwidth than that of the intrinsic zone 10, for example a forbidden bandwidth greater than 10 kBT with kB the Boltzmann constant and T the temperature of the device, advantageously InP. The other layers of the doped zones of smaller thicknesses 11, 13 and 17 than the body layers of the N type and P type ZQNs are formed of the same material as that of the layer of the intrinsic zone, for example layers of InGaAs. The fine InGaAs layers make it possible, in particular, to obtain good electronic transport conditions with low contact resistances and a better collection of the majority carriers. The doped body layers with a much larger forbidden bandwidth than that of the intrinsic zone make it possible to reduce the diffusion current bound to the minority carriers. For example, in the N-type zone, the InP layer will have the effect of introducing potential barriers to the InGaAs / InP interface in both the valence band and the conduction band. The potential barrier of the valence band is large enough to reduce the density of electrons injected into the ZQN. The gradient of the minority carriers in this area being reduced in the same ratio, it results in a reduction in the same ratio of the diffusion current. The barrier of the conduction band is thin enough for the electrons to pass through the tunnel effect, ensuring a thermodynamic equilibrium of the electron populations on either side of this barrier. The majority carriers are thus efficiently collected while the diffusion of the minority carriers is greatly reduced (by a factor close to: exp (A Eglk, T)), the dark current is reduced without hindering the collection of the current photo. Two electrical contact layers 18 and 19 are present in the structure. These electrical contacts are made with a semiconductor of low bandwidth, for example InGaAs, to ensure a low contact resistance. On the input face of the light, the contact 19 is offset at the edge of the detector or detector matrix so as not to create shading for the incident flux of photons and also to facilitate manufacture. On the upper side the contact is structured as is the barrier on this side. The P-doped upper layers are structured by etching on almost all their thicknesses. A layer 17 placed in the upper barrier acts as a stop layer for etching. It allows the control of the manufacturing process. It has no effect on the dark current. Table 2 shows an example of a composition of a semiconductor hetero structure for a low-noise quantum detection element as described in the present application. Zone Reference Material Doping (cm-3) Thickness (nm) P 18 InGaAs: C 1'19 20 P 44 InP: Be 3e18 170 P 17 InGaAs: C 1'19 10 P 15 InP: Be 3e18 93 P 13 InGaAs: C InGaAs - 70 N 11 InGaAs: Si 5e18 10 N 12 InP: Si 3e18 204 N 19 InGaAs: Si 5e18 21 Table 2: Example of composition of a semiconductor hetero structure for a detection element low-noise quantum Another source of dark current exists: the surface current mainly due to the non-radiative recombination processes located at the edge of the device. They are greatly reduced in this structure by eliminating any contact of the space charge area (mainly InGaAs) with the free surface of the device. In fact, the structuring of the upper part of the device leaves a semiconductor layer with a large weakly doped gap (here hydrogenated InP) with a low surface recombination rate (- 5000 cm / s).
    [0022] In addition, the pixels are separated from each other by the same method: etching of the P-type zone to the InGaAs barrier layer. In a variant, it will be possible not to engrave the P-type zone and to practice atomic hydrogen diffusion from the surface. To validate this approach, FIG. 11 represents the electrical current-voltage characteristic III (IV1) calculated for the structure as represented in FIG. 10, compared with that of an InGaAs junction and an equivalent InP junction (doping levels). and thicknesses preserved). In all cases, the radiative recombinations were artificially suppressed. The curves 101, 102 thus represent the current-voltage characteristics of the InGaAs junction, respectively in forward and reverse bias. The curves 103, 104 represent the current-voltage characteristics of the InP junction, respectively in forward and reverse bias. The curves 105, 106 represent the current-voltage characteristics of the InGaAs / InP heterojunction, respectively in forward and reverse polarization. The insertion of the InP barriers "sandwiching" the InGaAs PIN junction effectively reduces the diffusion current by nearly 10 orders of magnitude (here zlEg = 23 kBT). Thus, the proposed structure (FIG. 10) has a diffusion current comparable to that of an InP junction while exhibiting the absorption of an InGaAs junction. FIG. 12 represents current-voltage electrical characteristics calculated for the same structures but this time, taking into account the phenomena of radiative recombination. It is observed that the current-voltage characteristics of the InP (curves 103, 104) and InGaAs (curves 101, 102) junctions are similar to those calculated without taking into account the radiative recombinations. In these structures, the diffusion current is dominant. On the other hand, the very strong reduction of the diffusion current in the InGaAs / InP heterojunction dominates the contribution of the radiative recombination phenomena (curves 105, 106). The dark current of the InP / InGaAs heterojunction can thus be reduced by about 3 orders of magnitude compared to that of an InGaAs junction. Figures 13A and 13B illustrate the effect of structuring on radiative generation-recombination phenomena. FIG. 13A shows a mapping of the electrostatic potential in reverse bias (detector operation) in a device similar to that described in FIG. 10, with the parameters described in table 2. The N-type zones (layer 12) have a negative potential relatively constant and the P-type zones (layer 44) have a relatively constant positive potential. In the space charge area (layer 10), the potential varies. The arrow lines represent the electric field lines. Under the effect of the strong field that exists at the level of the structured part of the layer 44, the electron pairs holes generated via the intrinsic density are harvested by the semiconductor pads. FIG. 13B shows the modeling of the inverse polarization radiative recombination rate R as a function of the position in the device according to the cross section S of FIG. 13A. In reverse bias, in the situation where there is generation of electrons and holes, the recombination rate is negative. There is a decrease in the radiative recombination rate centered around 0.2 pm which corresponds to the region between the semiconductor pads. The effect of structuring on the recombination rate is thus observed. By adjusting the structuring of the network, it is possible to modify the profile of the recombination rate with respect to a flat profile of an unstructured device. FIG. 14 illustrates the electroluminescence spectrum calculated for an InGaAs / InP heterojunction photodiode as represented in FIG. 10, with the parameters given in table 2 and a direct polarization of 0.57 V. This spectrum was measured under strong direct voltage to facilitate the experiment. It has been shown experimentally that it depends very little on the applied voltage, in particular its maximum remains fixed at the wavelength rad. Figure 14 thus shows the radiative processes at the origin of the dark current as a function of the wavelength (also denoted "luminance of the heterostructure"). The function plotted on this figure is expressed as: - f: E-E, 9) / kr, Eg = 0.75 where E = hc / X ,, R the radiative recombination rate. The implementation of an optical resonator as described above makes it possible to very significantly reduce the generation-recombination current, related to the radiative recombination effects as they were demonstrated in FIGS. 11 to 14.
    [0023] It is possible firstly, thanks to the optical resonator according to the present description, to reduce the thickness of the zone I (active layer of absorbent semiconductor material). The corresponding component of the dark current density is then reduced in the thickness reduction ratio. In doing so, the quantum efficiency of the detector is also reduced. The thickness of zone I is advantageously reduced to a compromise value between these two antagonistic effects. On the other hand, there is a limit value, about 50nm for InGaAs, below which the contribution of the band-to-band tunnel current component generates an increase in the dark current. By reducing the thickness of the InGaAs layer at 50 nm instead of the 300 nm of the previously described structure, while keeping the thickness of the epitaxial layers constant by increasing the thickness of the InP barrier layers, a reduction in the thickness of the InGaAs layer is obtained. a factor of 6 on the dark current density. Moreover, the dimensioning of the optical resonator (parameters of the cavity and the coupling structure) is adapted to obtain, at the Xi-ad wavelength for which the radiative recombination rate is maximal (see equation 2 above), a reduced coupling with the free space so as to increase the lifetime of the carriers in this zone. It is therefore advantageous to place the detection device in an optical resonator having a resonance at the detection wavelength λ 0 (maximum coupling with the free space) and an antiresonance (minimal coupling with the space free) at the transmission length% i-ad. The reduction of the absorption at -rad thus produced is in particular, to be useful, greater than exp (AEg / kT) where 3Eg is the energy difference between krad and kc. FIG. 15 represents, as a function of the wavelength, the absorption of the averaged cavity over all angles of incidence (curve 122), the luminance of the heterojunction (curve 120) and the luminance resulting from the cavity, obtained by multiplication of the two spectra. The effect sought by the optical resonator used in the present description is clearly observed, namely an absorption maximum at the central wavelength of the spectral band of interest (here X0 = 1.5 mm) and reduced absorption at the radiative wavelength.
    [0024] FIGS. 16A to 16G illustrate an exemplary embodiment for the manufacture of a detection element according to the present description, adapted for example to layers of monocrystalline semiconductor material. In a first step (FIG. 16A), a stack of layers 21, 12, 10, 15, 17 and 44 of semiconductor material (for example alternately GaInAs) is produced by epitaxial growth on a suitable substrate 20 (for example made of InP). and InP). A first barrier layer 21 (for example InGaAs) is epitaxially grown on the InP substrate. The layers 15, 10 and 19 are suitable for producing the desired detector element. They are respectively an N-doped InP layer 15, an intrinsic layer and a P-doped layer 19 to form a PIN junction. The Intrinsic zone can have dimensions typically of 70 nm. The N-doping layer 19 may have dimensions of typically 200 nm with a carbon doping of typically 3 × 10 -8 cm-3. The doping layer n may have dimensions typically of 200 nm with a silicon doping of typically 3 × 10 18 cm -3.
    [0025] In a second step (FIG. 16B), a mask 60 is created by lithography on the InP 44 layer. This mask makes it possible in particular to define the shape of the coupling networks. In a third step (FIG. 16C), the InP layer is etched by reactive ion etching and by wet etching. The InGaAs layer then serves as a stop layer for etching. Then, the mask is removed. In a fourth step (FIG. 16D), the spinning planarizing resin 42 (dielectric) is coated. In a fifth step (FIG. 16E), a reactive ionic etching of the dielectric is effected until it is touched. In a sixth step (FIG. 16F), an Au metal layer 30 is deposited. In a seventh step, the sample is transferred to a host substrate by a bonding method (epoxy, anodic bonding, etc.). Finally, in an eighth step, the substrate 2 is removed to the stop layer 21 and then removed by the wet process. It is possible, to protect the device against external aggressions (hybridization, long-term atmosphere) and to reduce the effects of aging, to proceed to a final step of encapsulation by the deposition of a resin outside the metallized zones .
    [0026] The monocrystalline semiconductor material layers epitaxial as described in the above process may comprise, for example, gallium arsenide (GaAs) layers and the associated alloys (for example AlGaAs and aluminum arsenide), indium phosphide layers InP and the associated alloys (for example gallium arsenide and InGaAs indium), gallium antimonide (GaSb) and indium arsenide (InAs) layers, in particular a super InAs / GaSb network and the associated alloys (for example gallium aluminum antimonide A1GaSb), silicon layers (Si). The same method can be applied for the realization of other heterojunctions suitable for detection in other spectral bands. For example: - InAlAs / InGaAs at X1.5 iam; A1GaAs / GaAs at X0.9 iam; GaInP / GaAs at X0.9 i.tm; - [SR InAs / AlSb] / [SR InAs / GaSb] in the band 3-5 iam; - HgCdTe / HgCdTe in bands 3-5 i.tm and 8-12 iam; InP and InAlAs / GaAsSb at X1.5 μm. Although described mainly in the context of detectors in the infrared, the invention applies in the same way to the detection in the visible, especially for photovoltaic applications. The photodetection elements suitable for producing a device are substantially the same as those described for the detection, but they are directly polarized. Indeed, the conversion efficiency of solar energy into electrical energy is limited by several factors, including radiative recombinations of photocarriers. According to the article (Polman, A., & Atwater, HA (2012), Photonic design principles for ultrahigh-efficiency photovoltaics, Nature rnaterials, 11 (3), 174-177), one would gain about 7% on the output of solar cells by avoiding these recombinations. Although described through a number of detailed exemplary embodiments, the quantum sensing element and the method of manufacturing said element include various alternatives, modifications, and enhancements that will be apparent to those skilled in the art. 26
    权利要求:
    Claims (13)
    [0001]
    REVENDICATIONS1. Quantum photodetection element for incident radiation in a spectral band centered around a central wavelength Xo, having a front face intended to receive said radiation, and comprising: - one or more layers of semiconductor material comprising at least one minus one layer of absorbing semiconductor material of cut-off wavelength λ '> Xo, and forming a resonant optical cavity; a coupling structure of the radiation incident on the optical cavity to form: a resonance at the central wavelength λ 0 allowing absorption greater than 80% in the absorbent semiconductor material layer at said central wavelength; o absence of resonance at the radiative wavelength λ 1 -ad, where the radiative wavelength λ 1 -ad is the wavelength for which, at operating temperature, the radiative recombination rate is maximum.
    [0002]
    The photodetection element of claim 1, wherein the central wavelength resonance X0 is located in the layer of absorbent semiconductor material.
    [0003]
    A photodetection element according to any one of the preceding claims, wherein the coupling structure of radiation incident with the optical cavity is adapted to reduce absorption in the layer of absorbing semiconductor material at the radiative wavelength. Xi-ad, the said reduction being greater than exp (3Eg / kT) where 3Eg is the variation of the energies corresponding to the wavelengths Xi-ad and X.
    [0004]
    4. A photodetection element according to any one of the preceding claims, wherein the coupling structure of the radiation incident with the optical cavity makes it possible to form an antiresonance at the radiative wavelength Xi-ad in the semiconductor material layer. absorbent.
    [0005]
    A photodetection element according to any one of the preceding claims, wherein the optical cavity is a waveguide and the coupling structure comprises a front-face coupling network and / or a back-side coupling network, that is to say on one side of the optical cavity, opposite the side supporting the front face.
    [0006]
    6. The photodetection element according to claim 5, wherein the thickness of the waveguide is between X0 / 8n and Xo / n, advantageously between X0 / 4n and 3X0 / 4n, where n is the average value of the real parts. refractive indices of the layer (s) of semiconductor material forming the waveguide.
    [0007]
    7. photodetection element according to any one of claims 5 or 6, wherein the network or networks (x) coupling comprises a periodic structure, period between Xo / n and Xo / ni, where n is the average value real portions of the refractive indices of the layer (s) of semiconductor material forming the waveguide and ni is the real part of the refractive index of the radiation incident medium.
    [0008]
    8. photodetection element according to any one of the preceding claims, comprising a metal reflective layer arranged on one side of the optical cavity, opposite to the side supporting the front face.
    [0009]
    9. photodetection element according to any one of the preceding claims, comprising a stack of layers of semiconductor material forming a heterojunction PN or PIN type, said stack comprising the layer (s) of semiconductor material forming the optical cavity.
    [0010]
    The photodetection element of claim 9, wherein the heterojunction comprises barrier layers.
    [0011]
    The photodetection element of claim 10, wherein at least one of the barrier layers is structured over at least a portion of its thickness to form a coupling network with the optical cavity.
    [0012]
    12. photodetection element according to any one of the preceding claims, adapted for detection in one of the bands I, II or III of the infrared.
    [0013]
    An infrared detector comprising an assembly of quantum photodetection elements according to any of the preceding claims. 28. A method of manufacturing a quantum detection element according to any one of claims 1 to 12, comprising: the formation on a substrate of an epitaxial structure, comprising a stack of layers of semiconductor material, including minus one layer of absorbing semiconductor material of cut-off wavelength λc> Xo, and an upper layer of dielectric material, - structuring of the upper layer of dielectric material to form a coupling network, - the deposit on said coupling network of a metal reflective layer, - removal of the substrate to form the front face of the detection element. The method of manufacture of claim 14 wherein the formation of the epitaxial structure comprises organometallic epitaxy. 29
    类似技术:
    公开号 | 公开日 | 专利标题
    EP3084843B1|2021-01-27|Quantum detection element with low noise and method for manufacturing such a photodetection element
    EP2752893B1|2015-09-16|Avalanche-photodiode semiconductor structure and method for manufacturing such a structure
    EP2786426B1|2015-12-30|Semiconductor structure able to receive electromagnetic radiation, semiconductor component and process for fabricating such a semiconductor structure
    FR2970599A1|2012-07-20|PHOTODETECTOR OPTIMIZED BY METALLIC TEXTURATION AGENCED IN REAR FACING
    EP1903612A1|2008-03-26|Avalanche photodiode
    EP0849809A1|1998-06-24|Uncooled quantum well infrared detector
    EP2476143B1|2015-11-11|Light mixer for generating terahertz radiation
    EP2865017B1|2019-11-13|Semiconductor structure with absorbing region in a focusing cavity
    WO2012080989A2|2012-06-21|Photodetection device
    EP2184788A1|2010-05-12|Photodetector with internal gain and detector comprising an array of such photodetectors
    EP1672696B1|2011-09-14|Semiconductor device to detect multi-spectral radiaton and method of using the device
    US20140217540A1|2014-08-07|Fully depleted diode passivation active passivation architecture
    EP2801115B1|2017-02-15|Photodetection device
    EP3350842B1|2020-03-18|Photodetector with reduced dark current
    EP2382672B1|2018-11-07|Method of making a photodiode and corresponding photodiode and electromagnetic radiation detector
    FR2721439A1|1995-12-22|Optical semiconductor elements and methods of manufacturing thereof
    FR3092700A1|2020-08-14|BARRIER PHOTO-DETECTOR MATRIX WITH LOCAL DEPLOYMENT PIXELLISATION
    Lo et al.2005|Suppressing dark-current in planar Si-based MSM photodetector with alternated ia-Si: H/ia-SiGe: H grade superlattice-like layers
    FR3041815A1|2017-03-31|PHOTODETECTOR COMPRISING A STACK OF SUPERIMPOSED LAYERS
    Tseng et al.2012|Low breakdown voltage silicon avalanche photodetector implemented by interdigitated pin junctions
    同族专利:
    公开号 | 公开日
    FR3015113B1|2018-06-01|
    EP3084843B1|2021-01-27|
    US20160322516A1|2016-11-03|
    US9755090B2|2017-09-05|
    CN105981179B|2017-11-24|
    WO2015091709A1|2015-06-25|
    CN105981179A|2016-09-28|
    JP2017500743A|2017-01-05|
    EP3084843A1|2016-10-26|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US5455421A|1985-08-13|1995-10-03|Massachusetts Institute Of Technology|Infrared detector using a resonant optical cavity for enhanced absorption|
    US5818066A|1995-11-21|1998-10-06|Thomson-Csf|Optoelectronic quantum well device having an optical resonant cavity and sustaining inter subband transitions|EP3261130A1|2016-06-20|2017-12-27|ams AG|Photodetector device with integrated high-contrast grating polarizer|FR2748604B1|1996-05-13|1998-08-07|Commissariat Energie Atomique|PHOTODETECTOR WITH RESONANT OPTICAL STRUCTURE WITH A NETWORK|
    US6055262A|1997-06-11|2000-04-25|Honeywell Inc.|Resonant reflector for improved optoelectronic device performance and enhanced applicability|
    US6423980B1|1999-06-25|2002-07-23|California Institute Of Technology|Multi-directional radiation coupling in quantum-well infrared photodetectors|
    JP2008288293A|2007-05-16|2008-11-27|Nippon Telegr & Teleph Corp <Ntt>|Semiconductor photodetector|
    FR2970599B1|2011-01-17|2012-12-28|Commissariat Energie Atomique|PHOTODETECTOR OPTIMIZED BY METALLIC TEXTURATION AGENCED IN REAR FACING|
    US8941203B2|2012-03-01|2015-01-27|Raytheon Company|Photodetector with surface plasmon resonance|
    FR3015113B1|2013-12-17|2018-06-01|Centre National De La Recherche Scientifique- Cnrs|LOW NOISE QUANTUM DETECTION ELEMENT AND METHOD FOR MANUFACTURING SUCH PHOTODETECTION ELEMENT|FR3015113B1|2013-12-17|2018-06-01|Centre National De La Recherche Scientifique- Cnrs|LOW NOISE QUANTUM DETECTION ELEMENT AND METHOD FOR MANUFACTURING SUCH PHOTODETECTION ELEMENT|
    FR3069956B1|2017-08-03|2019-09-06|Thales|IMPROVED INFRARED DETECTOR|
    JP2019125699A|2018-01-16|2019-07-25|日本電気株式会社|Infrared detection element, infrared detector and method of manufacturing infrared detection element|
    FR3083645B1|2018-07-05|2020-07-31|Thales Sa|IMPROVED BI-SPECTRAL DETECTOR|
    CN109411500B|2018-10-31|2021-01-22|京东方科技集团股份有限公司|Detection panel and manufacturing method thereof|
    法律状态:
    2015-11-23| PLFP| Fee payment|Year of fee payment: 3 |
    2016-11-21| PLFP| Fee payment|Year of fee payment: 4 |
    2018-01-02| PLFP| Fee payment|Year of fee payment: 5 |
    2018-04-13| AV| Other action affecting the existence or the validity of an industrial property right|Effective date: 20180314 |
    2019-12-31| PLFP| Fee payment|Year of fee payment: 7 |
    2020-12-28| PLFP| Fee payment|Year of fee payment: 8 |
    2021-12-31| PLFP| Fee payment|Year of fee payment: 9 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1362842A|FR3015113B1|2013-12-17|2013-12-17|LOW NOISE QUANTUM DETECTION ELEMENT AND METHOD FOR MANUFACTURING SUCH PHOTODETECTION ELEMENT|
    FR1362842|2013-12-17|FR1362842A| FR3015113B1|2013-12-17|2013-12-17|LOW NOISE QUANTUM DETECTION ELEMENT AND METHOD FOR MANUFACTURING SUCH PHOTODETECTION ELEMENT|
    US15/105,753| US9755090B2|2013-12-17|2014-12-17|Quantum detection element with low noise and method for manufacturing such a photodetection element|
    EP14825129.1A| EP3084843B1|2013-12-17|2014-12-17|Quantum detection element with low noise and method for manufacturing such a photodetection element|
    JP2016541002A| JP2017500743A|2013-12-17|2014-12-17|Low noise quantum detector and method for manufacturing such photon detector|
    PCT/EP2014/078329| WO2015091709A1|2013-12-17|2014-12-17|Quantum detection element with low noise and method for manufacturing such a photodetection element|
    CN201480075386.0A| CN105981179B|2013-12-17|2014-12-17|Low noise quantum detecting element and the method for making the optical detection device|
    [返回顶部]