Patent Application
20210193888
The field of the invention is that of methods for producing light-emitting diodes comprising a semiconductor stack of inorganic or organic layers, including an emissive active layer and an extraction layer.
The light-emitting diode 1 rests here on a substrate 2 comprising a conductive layer 3, in contact with which is an advantageously reflective electrode 4 forming the anode. A preferably transparent electrode 5 forming a cathode rests on the N-doped layer 11. The reflective electrode 4 is preferably made of a metal material.
The active layer 13 may be formed, in the case of a PN junction, by the interface of the two doped layers 11, 12 in contact with one another, or in the case of a PIN junction, by at least one intrinsic layer 13 (i.e. a layer that is not intentionally doped) that may comprise one or more quantum wells. Each quantum well may be formed by a layer having a bandgap energy lower than that of the two barrier layers between which it is located, or may be formed by quantum dots.
The charge carriers (electrons and holes) are introduced into the semiconductor stack by the electrodes 4, 5, then diffuse to the active layer 13 where they recombine radiatively. The luminous radiation associated with the radiative recombination of the electron-hole pairs in the active layer 13 corresponds to the electric dipole radiation emitted by what is referred to as an “emitting” dipole which oscillates harmonically along the axis of its dipole moment μ (also called TDMV, for transition dipole moment vector).
There is thus a need for a method for producing at least one light-emitting diode whose luminous efficacy is improved.
The objective of the invention is to overcome, at least in part, the drawbacks of the prior art, and more particularly to propose a method for producing at least one light-emitting diode exhibiting improved luminous efficacy.
To that end, the subject of the invention is a method for producing at least one light-emitting diode comprising: a semiconductor stack formed of a first semiconductor layer, of a second semiconductor layer, and of an active layer located between the two semiconductor layers; and an extraction layer, made of a transparent dielectric material comprising nanoscale particles, extending in contact with the first semiconductor layer. The method comprises the following steps:
Some preferred but non-limiting aspects of this method are the following.
The nanoscale particles may be quantum dots made of semiconductor nanocrystals, and/or be made of at least one metal material.
The nanoscale particles may have a mean diameter of between 0.2 nm and 500 nm.
The nanoscale particles may extend in a plane parallel to the active layer, and be arranged at the interface with the first semiconductor layer.
The first determined distance may be smaller than or equal to 50 nm, thus optimizing the non-radiative coupling of dipole-dipole type between the emitting dipoles and the optical dipoles.
The first determined distance may be defined along an axis orthogonal to the plane of the active layer, from an interface between the extraction layer and the first semiconductor layer to a plane passing through halfway through a thickness of the active layer.
The active layer may comprise at least one quantum well emissive layer, and a barrier layer located between the first semiconductor layer and the emissive layer, the determined distance being defined along an axis orthogonal to the plane of the active layer, between, on the one hand, the interface between the extraction layer and the first semiconductor layer, and, on the other hand, a plane passing through halfway through the thickness of the emissive layer.
The method may comprise a step of determining a second non-zero distance between the optical dipoles associated with the nanoscale particles and a surrounding medium extending in contact with an upper face of the extraction layer opposite the first semiconductor layer, comprising the following sub-steps:
The step of determining the second distance may comprise the following sub-steps:
The method may comprise:
The reflective electrode may be an anode capable of injecting holes into the semiconductor stack, and the second semiconductor layer may be made of a P-doped semiconductor crystalline material, or be made of a hole-conducting organic semiconductor material.
The first semiconductor layer may be made of an N-doped semiconductor crystalline material, or be made of an electron-conducting organic semiconductor material.
Other aspects, aims, advantages and features of the invention will become more clearly apparent from the following detailed description of preferred embodiments thereof, which description is given by way of non-limiting example and with reference to the appended drawings, in which:
In the figures and in the remainder of the description, the same references have been used to designate identical or similar elements. In addition, the various elements are not shown to scale for the sake of clarity of the figures. Moreover, the various embodiments and variants are not mutually exclusive and may be combined with one another. Unless indicated otherwise, the terms “substantially”, “about” and “of the order of” mean to within 10%, and preferably to within 5%. Moreover, the terms “comprised between . . . and . . . ” and equivalents mean that the bounds are included, unless indicated otherwise.
The invention relates to a method for producing at least one light-emitting diode exhibiting improved luminous efficacy. The luminous efficacy is defined here as the ratio of the luminous flux emitted by the light-emitting diode to the electrical power injected, and corresponds to the external quantum efficiency (EQE). It is equal to the product of the internal quantum efficiency (IQE) and of the light extraction efficiency. The internal quantum efficiency is the ratio of the number of photons generated by radiative recombination to the number of electrons injected by the cathode, and the light extraction efficiency is the ratio of the number of photons emitted out of the diode to the number of photons generated.
For that, a first layer of the semiconductor stack of the light-emitting diode is coated at least partly with an extraction layer notably intended to increase the light extraction efficiency. This extraction layer is formed of at least one dielectric material that is transparent to the luminous radiation emitted by the active layer of the light-emitting diode. It also comprises, located in the dielectric material, nanoscale particles such as quantum dots and/or metal particles, such that the light extraction induces a near-field coupling of dipole-dipole type between the optical dipoles referred to as “donors” associated with the active layer and the optical dipoles referred to as “acceptors” associated with the nanoscale particles.
Additionally, as explained in detail below, the first semiconductor layer is dimensioned such that the donor optical dipoles associated with the radiative recombination of the electron-hole pairs in the active layer have a vertical orientation that predominates over a horizontal orientation. The horizontal or vertical character of the orientation of the optical dipole is in relation to a plane along which the active layer of the light-emitting diode extends. In addition, the orientation of an emitting dipole corresponds to the angle of inclination θ formed by the dipole moment μ of the optical dipole with respect to an axis orthogonal to the plane of the active layer.
An orthogonal direct coordinate system XYZ, in which the axes X and Y form a plane parallel to the main plane of a substrate (not shown) on which the light-emitting diode 1 rests, and in which the axis Z is oriented orthogonally to the plane XY and in the direction of the extraction layer 6, is defined here and will be referred to in the rest of the description. Here, the active layer 13 extends along the plane XY.
The light-emitting diode 1 thus comprises a semiconductor stack formed of a first layer 11, of an active layer 13, and of a second layer 12. In this example, the semiconductor stack is produced on the basis of an inorganic crystalline semiconductor material, but, as a variant, it may be produced on the basis of an organic semiconductor material (OLED). The light-emitting diode 1 may comprise additional layers which are not shown, for example an electron-blocking layer, a buffer layer for matching the lattice parameter, etc.
What is meant by “produced on the basis of semiconductor material” is that the semiconductor stack is made of this semiconductor material or of one or more compounds comprising this semiconductor material. By way of example, the semiconductor stack is produced on the basis of a III-V compound, for example on the basis of GaN, and may thus comprise at least one semiconductor layer made of doped or undoped GaN, and at least one semiconductor layer made of a compound comprising GaN, for example of InGaN, AlGaN, InAlGaN, or is even produced on the basis of InP, for example of the type AlInGaP, etc.
It is noted, as indicated above, that the light-emitting diode 1 may be an organic diode, in which case it is formed of two, anode and cathode, electrodes between which organic semiconductor layers are stacked, including an electron transport layer (ETL), an active layer (EML) and a hole transport layer (HTL).
In this example where the light-emitting diode 1 is produced on the basis of an inorganic material, the semiconductor stack is produced on the basis of GaN, the first layer 11 is made of N-doped GaN, the second layer 12 is made of P-doped GaN. The active layer 13 comprises here a quantum well layer 13.1 made of intrinsic InGaN located between two barrier layers 13.2 made of intrinsic GaN. Of course, the active layer 13 may comprise a plurality of quantum well layers located between two barrier layers. The reflective electrode 4 is made of one or more metal materials chosen from among silver, aluminum, copper, titanium, gold, nickel, iridium, tungsten, indium, inter alia. It extends in contact with the second layer 12. It is referred to as reflective insofar as its reflection coefficient is at least equal to 75%, at least equal to 85%, or even to 95%, or even more, at the emission wavelength of the light-emitting diode 1. The active layer 13 extends along a plane XY parallel to that of the extraction layer 6. A cathode electrode (not shown) may extend in contact with the first layer 11, for example coplanar with the extraction layer 6.
The extraction layer 6 is suitable for allowing a near-field non-radiative coupling of dipole-dipole type with the active layer 13, thus improving the light extraction efficiency and therefore the performances of the light-emitting diode 1. It extends in contact with the first layer 11 (here made of N-doped GaN) and at least partially coats it (notably when another portion of the surface of the first layer 11 is coated with the cathode electrode).
The extraction layer 6 is formed of a dielectric material 6.2 that is transparent to the emission wavelength of the active layer 13 which forms a binding matrix in which nanoscale particles 6.1 are located. The transparent dielectric material 6.2 has a coefficient of transmission of the incident radiation that is at least equal to 50%, preferably at least equal to 75%, or even to 90%, or even more. It may be chosen from among silicone, polysiloxane, PDMS, PM MA, PVA, oxide-based mineral layers of sol-gel type such as SiO2, Al2O3, ZnO, TiO2, inter alia.
The nanoscale particles 6.1 have an average diameter of the order of a few nanometers to a few hundreds of nanometers, for example are between 0.2 nm and 500 nm, and preferably between 2 nm and 150 nm. They may notably be metal particles and/or quantum dots. The particles are elements that are distinct from one another, and may be of any shape, for example spherical, angular, flattened or elongated, or any other shape. The size of the particles is here the smallest dimension of the particles and the average diameter is the arithmetic mean of the size of the particles.
In the case where the nanoscale particles 6.1 are made of a metal material, this may be chosen from among Ag, Cu, Au, Pt, Pd, Ni, Co, Rh, In, Ru, Fe, CuNi, inter alia, and from among the compounds formed from a mixture of at least two or more of these materials. The surface density of metal particles is determined by Mie scattering so as to optimize the absorption efficiency of a metal particle according to the refractive index of the dielectric medium. By way of example, for a dielectric medium with a refractive index of 1.5, an average diameter of 55 nm for nanoscale particles of silver makes it possible to optimize the absorption efficiency at 460 nm. The absorption cross section of these particles is then 0.5×10−14 m2 and the surface density may then be of the order of 2×1014 m−2.
In the case where the nanoscale particles 6.1 are quantum dots, they may be produced in the form of semiconductor nanocrystals, the average size of which may be between 0.2 nm and 500 nm, for example between 1 nm and 100 nm, and notably between 2 nm and 30 nm. The semiconductor material of the nanocrystals may notably be chosen from among cadmium selenide (CdSe), indium phosphide (InP), cadmium sulfide (CdS), zinc sulfide (ZnS), cadmium oxide (CdO), cadmium zinc selenide (CdZnSe), or from among other potentially suitable semiconductor materials. The surface density is here notably dependent on the desired light conversion efficiency, i.e. on the ratio of the intensity of the photoluminescence radiation emitted by the quantum dots 6.1 to the intensity of the electroluminescence radiation emitted by the active layer 13.
The nanoscale particles 6.1 may also be plasmon-effect quantum dots, produced for example from a metal core surrounded by a dielectric sheath, itself surrounded by a layer forming a quantum dot. The metal core, for example made of silver, forms an absorbent antenna for the dipole-dipole coupling, and couples this absorbed energy with the material forming the quantum dot, which will subsequently emit a photoluminescence radiation.
The nanoscale particles 6.1 are arranged in the extraction layer 6 preferably in proximity to or at the interface with the first layer 11, so as to allow the near-field non-radiative coupling between the optical dipoles μ1 associated with the active layer 13 and the optical dipoles μ2 associated with the nanoscale particles 6.1. They are preferably arranged along a plane parallel to the plane XY. Preferably, the distance along the axis Z separating the nanoscale particles 6.1 from the active layer 13 is smaller than or equal to 50 nm. This distance may be defined between the plane in which the nanoscale particles 6.1 mainly extend and a plane passing through halfway through the thickness of the active layer 13.
Since the active layer 13 is the main location of the radiative recombination of the electron-hole pairs, the optical dipoles μ1 referred to as emitters (also called donors μ1(D) in the context of dipole-dipole coupling) are located in the active layer 13, and are therefore spaced by a distance h1 with respect to the extraction layer 6. It is considered that they are located in a plane parallel to the plane XY and spaced by the distance h1 from the extraction layer 6. Let h1=0 denote the interface between the extraction layer 6 and the first layer 11.
The distance h1 separating the emitting dipoles μ1 with respect to the extraction layer 6 is defined for the rest of the description as being the distance along the axis Z between, on the one hand, a plane passing through halfway through the thickness of the active layer 13, and, on the other hand, the extraction layer 6. However, as a variant, and notably when the active layer 13 comprises a plurality of quantum well emissive layers 13.1, the distance h1 may be defined as being the distance along the axis Z between, on the one hand, a plane passing through halfway through the thickness of the quantum well layer 13.1 located as close as possible to the layer 12 made of P—GaN, and, on the other hand, the extraction layer 6. A prior study may be carried out in order to determine where exactly in the active layer 13 most of the radiative recombination of the electron-hole pairs is located. In this example, the distance h1 is therefore equal to the sum of the thickness of the first layer 11, of the thickness of the barrier layer 13.2, and of halfway through the thickness of the emissive layer 13.1.
The orientation of an emitting dipole μ1 corresponds to the angle θ1 between its dipole moment and the axis Z orthogonal to the plane XY of the active layer 13. Thus, an angle θ1 equal to 0° corresponds to a vertical orientation of the emitting dipoles μ1 with respect to the plane of the active layer 13 and here in the direction of the extraction layer 6, and an angle θ1 equal to 90° corresponds to a horizontal orientation.
It is considered here that the emitting dipoles μ1 are located in a first optically linear, uniform and isotropic medium of dielectric constant ε1 (relative permittivity) and of refractive index n1, this first medium being formed of the active layer 13 and of the first layer 11. What is meant by optically uniform is that the dielectric constant ε1 is substantially constant at any point in this first medium, to within 10%, or even to within 5%, or even less.
The dielectric material 6.2 of the extraction layer 6 forms a second medium that is considered, to a first approximation, as being optically linear, uniform and isotropic, of dielectric constant ε2 and of refractive index n2. By extension, it is considered that the extraction layer 6 forms this same second uniform medium. The refractive index may be the optical index having a real part (refractive index per se) and a non-zero imaginary part (extinction index).
The method for producing a light-emitting diode 1 comprises a phase of determining a distance denoted by h1s between the emitting dipoles μ1 associated with the radiative recombination in the active layer 13 and the extraction layer 6, such that the emitting dipoles μ1 have a vertical orientation (θ1=0°). The thickness of the first layer 11 may then be determined by taking the value h1s into account.
For that, a function g is predetermined which expresses a change, according to the distance h1, in a parameter representative of a lifetime of an emitting dipole μ1 having one or the other of said orientations, according to the optical properties n1, n2 of the first and second uniform media. Thus, the function g is denoted by gh or gv when it is in relation to a horizontal (θ1=90°) or vertical (θ1=0°) orientation.
The lifetime τ is defined as being that of an emitting dipole located in an optically linear, uniform and isotropic medium, of dielectric constant ε. It corresponds to the lifetime of the spontaneous emission of a two-level system, for which it is considered, to a first approximation, that it is identical to the lifetime of the emitting dipoles in the context of electroluminescence. Let τn denote the normalized lifetime equal to the ratio of the lifetime τ/τ0, where the lifetime τ0 is defined as being that of an emitting dipole located in the same medium but of infinite dimensions, hence at an infinite distance from the second uniform medium.
Now, the normalized lifetime τn is equal to the inverse of the normalized radiated power P/P0 of the emitting dipole, in other words: τn=τ/τ0=P0/P. Here, P is the energy dissipation rate of the emitting dipole, in other words the optical power radiated by the emitting dipole in the first uniform medium and located at a distance h from the second uniform medium, and P0 is the optical power radiated by the same emitting dipole in the first uniform medium of infinite dimensions (far from the second uniform medium, for example more than 500 nm away from the second uniform medium).
The work by Novotny & Hecht entitled Principles of Nano-Optics, Cambridge University Press, 2006, indicates on page 344 (equation no.10.26) an expression for the normalized radiated optical power P/P0 of an emitting dipole located in a first uniform medium at a distance h from a substrate formed of a second uniform medium (thin layer) and of a third uniform medium.
This equation may be adapted for the present configuration of a light-emitting diode 1 where the emitting dipole of the active layer 13 is located in the first uniform medium formed of the active layer 13 and of the first layer 11, and is arranged at a distance h from the second uniform medium formed of the extraction layer 6. Thus, the normalized optical power radiated by such an emitting dipole may be written as:
where:
However, in the configuration according to the invention, the reflection coefficient rp at the interface with the extraction layer 6 for the luminous radiation of p polarization may be written, according to the variable of integration s, in the following manner:
Furthermore, the reflection coefficient rs at the interface with the extraction layer 6 for the luminous radiation of s polarization may be written, according to the variable of integration s, in the following manner:
Additionally, the term μz2/μ2 corresponds to cos θ, and the term (μx2+μy2)/μ2 is equal to sin θ. The wave vector k has, in the first uniform medium, a norm denoted by k1 equal to 2π/(n1×λ). As mentioned above, n1 and n2 are the refractive indices of the first and second uniform media, which are deduced from the dielectric constants ε1, ε2.
Thus, the radiated power P radiated by the emitting dipole is formed of three main terms, namely the intrinsic radiated power P0 (apart from any effect related to the environment of the first uniform medium), a term corresponding to the radiated power associated with the dipole moment μx and μy in the plane XY and a term corresponding to the radiated power associated with the dipole moment μz along the axis Z.
This equation is obtained on the basis of Maxwell's equations, using dyadic Green's functions to describe a single oscillating point dipole, and the angular spectrum method for the field of the dipole extends the Green's functions to plane and evanescent waves, respectively, in a cylindrical system. The emitting dipole therefore interacts with its own reflected plane and evanescent waves.
Thus, a function g is obtained that expresses the change in a parameter representative of the lifetime of an emitting dipole μ1 of predefined orientation θ1 according to the distance h1 from the extraction layer 6, taking into account the optical properties n1 of the first uniform medium (formed of the first layer 11 and active layer 13) and of those n2 of the second uniform medium (extraction layer 6). The representative parameter is preferably the normalized lifetime τn=τ/τ0.
More precisely, the light-emitting diode 1 comprises an emissive layer 13.1 made of InGaN (no barrier layer made of intrinsic GaN) and a first layer 11 made of N-doped GaN which together form the first uniform medium. The refractive index n1 is equal to about 2.4764 for an emission wavelength of λ=460 nm. The first layer 11 is in contact with the extraction layer 6 made of a dielectric material which forms the second uniform medium and has a refractive index n2 equal here to 1.5.
It is seen that the normalized lifetime τn1,h of the emitting dipoles μ1 of horizontal orientation remains substantially constant and equal to 1.0, whatever the value of the distance h1 over the range Δh1ref. However, the normalized lifetime τn1,v of the emitting dipoles μ1 of vertical orientation decreases as the distance h1 increases, going here from about 4.5 for h1=0 nm to 1.0 around 500 nm. Thus, over the entire range Δh1ref, the vertical orientation of the emitting dipoles μ1 remains predominant over the horizontal orientation. The thickness of the first layer 11 may therefore be dimensioned such that the distance h1s retained between the emitting dipoles from the active layer 13 to the extraction layer 6 is between about 1 nm and 500 nm.
However, it is advantageous for the distance h1s to be determined such that a difference between the normalized lifetime τn1,v of the emitting dipoles μ1 of vertical orientation and the normalized lifetime τn1,h of the emitting dipoles μ1 of horizontal orientation is large in terms of absolute value. Thus, it is possible to define a parameter referred to as the “difference” parameter S1, called selectivity, representative of a difference between the normalized lifetime τn1,h of an emitting dipole μ1 having the vertical orientation and the normalized lifetime τn1,h of an emitting dipole μ1 having the horizontal orientation. This selectivity S1 may be defined as being the difference or the ratio, in terms of absolute value or otherwise, between the normalized lifetime τn1,v of an emitting dipole μ1 having the vertical orientation and the normalized lifetime τn1,h of an emitting dipole μ1 having the horizontal orientation. In other words, S1==τn1,v−τn1,h or, as a variant, S1=τn1,v/τn1,h. Hereinafter, S1==τn1,v−τn1,h is used.
It is seen that the parameter S1 has a value that is higher than or equal to 2.0 for h1 shorter than or equal to 100 nm, and a value that is higher than or equal to 2.5 for h1 shorter than or equal to about 50 nm. The thickness of the first layer 11 may therefore be dimensioned such that the distance h1s between the emitting dipoles from the active layer 13 to the extraction layer 6 is between about 1 nm and 50 nm. Preferably, the thickness of the first layer 11 could be chosen between 1 nm and 100 nm, and preferably between about 5 nm and 10 nm so as to maximize the near-field effect. What is meant by “near field” is that the distance between the acceptor and donor dipoles is shorter than or equal to λ/5, or even shorter than or equal to λ/10, where λ is the emission wavelength of the radiation emitted by the active layer, by way of the refractive index of the medium, in which the electric field radiated by the dipole decreases by 1/r3.
Thus, it is possible to determine a distance h1s for which the normalized lifetime τn1,v associated with the vertical orientation of the emitting dipoles μ1 is longer than the normalized lifetime τn1,h associated with the horizontal orientation. In this case, the emitting dipoles μ1 having the vertical orientation will predominate over those having the horizontal orientation. The light-emitting diode 1 may then be dimensioned, in particular the thickness of the first layer 11, such that the emitting dipoles μ1 are located at the distance hs1 from the extraction layer 6.
Thus, a near-field non-radiative coupling of dipole-dipole type is present between the emitting dipoles denoted here by μ1(D) (for Donors) that are associated with the radiative recombination in the active layer 13 and the dipoles denoted here by μ2(A) (for Acceptors) that are associated with the nanoscale particles 6.1. These dipoles μ1(D) and μ2(A) are the vectors of the dipole moments of the optical dipoles. The intensity of this non-radiative coupling is characterized by the angular coupling factor K2.
In a known manner, the angular coupling factor K2 between the emitting dipoles μ1(D) and μ2(A) is defined by the relationship:
K
2=(nA·nD−3(nA·nD)(nr·nA))2
This expression is notably described in the work by Novotny & Hecht 2006 on page 290 (equation 8.169). The factor K2 is dependent on the orientation of the unit vectors nA and nD that are associated with the acceptor μ2(A) and donor μ1(D) dipoles, and on the unit vector nr linking the acceptor μ2(A) and donor μ1(D) dipoles in question.
The dipole-dipole interaction modifies the absorption properties of the acceptor dipoles μ2(A). There is a non-radiative transfer of energy between the donor dipoles μ1(D) and the acceptor dipoles μ2(A), called FRET (Förster resonance energy transfer) coupling, which manifests as an increase in the luminous efficacy of the light-emitting diode 1.
It is seen that the angular coupling factor K2 is maximum when the donor μ1(D) and acceptor μ2(A) dipoles are collinear, in which case the factor K2 is equal to 4. Thus, dimensioning the first layer 11 in order to obtain a predominant vertical orientation of the donor dipoles μ1(D) allows the acceptor dipoles μ2(A) to be made collinear with the donor dipoles μ1(D), and thus the luminous efficacy of the light-emitting diode 1 to be improved further. Specifically, in the context of dipole-dipole non-radiative coupling, the acceptor dipoles μ2(A) are oriented according to the orientation of the donor dipoles μ1(D).
In the case where the extraction layer 6 is formed of metal particles 6.1, the luminous efficacy of the light-emitting diode 1 is further increased by an additional luminous emission by plasmonic effect. The metal particles 6.1 may thus emit a luminous radiation at a resonance wavelength that is substantially identical to the electroluminescence wavelength λe of the light-emitting diode 1. For that, plasmonic modes of the metal particles 6.1 are excited, a resonant mode of which may result in the emission of the additional luminous radiation. Here, the plasmonic resonant mode of the metal particles 6.1 corresponds to acceptor dipoles μ2(A) that are located in the near field of the donor dipoles μ1(D). Additionally, in the case where the nanoscale particles 6.1 are quantum dots, the extraction layer 6 provides an addition color conversion function, by converting a portion of the luminous radiation emitted by the active layer 13 into a luminous radiation of longer wavelength.
Additionally, it may be advantageous to dimension the thickness of the extraction layer 6 such that the optical dipoles μ2 associated with the nanoscale particles 6.1 are located at a determined distance h2s with respect to the interface between its upper face 6a and the surrounding medium (e.g. the air). Specifically, it may be advantageous to make the vertical orientation of these optical dipoles μ2 predominate so as for example to further optimize the near-field non-radiative coupling of dipole-dipole type with the optical dipoles μ1 of the active layer 13. Conversely, it may be advantageous to make the horizontal orientation of these optical dipoles μ2 predominate when the nanoscale particles 6.1 form quantum dots, so as to optimize the intensity of the far-field photoluminescence luminous radiation. The nanoscale particles 6.1 are still located at the predetermined distance h1s with respect to the active layer 13.
It is seen that the normalized lifetimes τn2,v, τn2,h exhibit an increase from h2=0 nm to about 50 nm for τn2,h and to about 150 nm for τn2,v. Next, they remain substantially constant around 1.0 as h2 increases, while exhibiting damped oscillations.
However, it is seen that the lifetime associated with one of the orientations predominates over the lifetime associated with the other orientation for a plurality of domains Δhi=1,2 . . . of distance h2 in the range Δh2ref, and that this predominance alternates according to the distance h2.
Thus, it is possible to determine a distance h2s for which the normalized lifetime τn2,s associated with the chosen orientation (from among the vertical and horizontal orientations) of the emitting dipoles μ2 is longer than the normalized lifetime τn,ns associated with the other, non-chosen orientation. In this case, the emitting dipoles μ2 having the chosen orientation will predominate over those having the non-chosen orientation. The light-emitting diode 1 may then be dimensioned, in particular the thickness of the dielectric material 6.2 of the extraction layer 6, such that the emitting dipoles μ2 are located at the distance h2s from the upper face 6a, and therefore have the chosen orientation. The emitting dipoles μ2 may thus, preferably, have a horizontal orientation when the nanoscale particles 6.1 are quantum dots, or have a vertical orientation when the nanoscale particles 6.1 are metal for a plasmonic effect.
In a step 100, the materials of the first layer 11 and of the active layer 13, which together form the first uniform medium of refractive index n1, are chosen. Here they are N-doped GaN and InGaN, the refractive index of which is substantially equal for these two materials and corresponds to 2.4764 at the emission wavelength of 460 nm. In the case where the refractive indices are not identical, it is possible to define an average refractive index on the basis for example of a volume weighting of the refractive indices. The material of the extraction layer 6 is also chosen, which layer forms the second uniform medium of refractive index n2 substantially equal to that of the dielectric material 6.2.
As mentioned above, the distance h1 corresponds here to the distance along the axis Z between, on the one hand, the extraction layer 6/first layer 11 interface, and, on the other hand, halfway through the thickness of the active layer 13. In this example, the active layer 13 is a quantum well emissive layer 13.1 with a thickness of 3 nm, and it is desired to determine the thickness of the first layer 11 such that the emitting dipoles μ1 (located, to a first approximation, at the center of the active layer 13) are located at the determined distance h1s in order to obtain a vertical dipole orientation.
In a step 200, a value h1s of the distance h1 is determined such that a lifetime of the emitting dipoles μ1 having the vertical orientation θv is longer than that of the emitting dipoles μ1 having the horizontal orientation θh. For that, the predetermined function g expressing a relationship between a lifetime of an optical dipole having a predefined orientation θ and the distance h is used. This function g is that described above, which expresses the change in the normalized lifetime τn of an emitting dipole according to the distance h, by means of the normalized radiated optical power P/P0.
In a sub-step 210, the change gh according to h1 in the normalized lifetime τn1,h of an emitting dipole μ1 of horizontal orientation θh and the change gv in the normalized lifetime τn1,v of an emitting dipole μ1 of vertical orientation θv are determined over a distance range Δh1ref ranging for example from 0 nm to 500 nm. These changes gh and gv are determined on the basis of the relationships τn=P0/P and P/P0=f(h) indicated above. Thus obtained is τn1,v=gv(h1) and tn1,h=gh(h1) for any h1 within the range Δh1ref.
In a sub-step 220, a value h1s is determined such that the normalized lifetime τn1,v of the emitting dipole μ1 of vertical orientation θv is longer than the normalized lifetime τn1,h of the emitting dipole μ1 of horizontal orientation θh, i.e. here such that τn1,v(h1s)>τn1,h(h1s). Additionally, the value h1s is advantageously determined such that the selectivity |S1(h1s)| is maximum over at least one domain in the range Δh1ref, and preferably over the entire range Δh1ref, in other words |S1(h1s)|=maxΔh1ref(|S1(h1)|).
Other conditions may also be taken into account, such as for example the fact that the value h1s is higher than or equal to a predefined non-zero minimum value hth, for example in order to optimize the diffusion of the charge carriers in the plane XY within the first layer 11 from the cathode electrode 5 (illustrated in
In an optional but advantageous step 300, a value h2s of the distance h2 is also determined such that a normalized lifetime τn2,s of the optical dipoles μ2, having a chosen orientation θ2s (from among the horizontal orientation θh and the vertical orientation θv), is longer than the normalized lifetime τn2,ns of the optical dipoles μ2 having the non-chosen orientation θns.
For that, in a sub-step 310, the desired orientation θ2s of the optical dipoles μ2 that are associated with the nanoscale particles 6.1, i.e. here the angle θ2 formed by the dipole moment μ2 with respect to the orthogonal axis Z, is chosen from among the vertical orientation θv (θ2=0°) and the horizontal orientation θh (θ2s=90°). In this example, the chosen orientation θ2s is the horizontal orientation θh (θ2s=90°).
In a sub-step 320, the change gh according to h2 in the normalized lifetime τn2,h of an optical dipole μ2 of horizontal orientation θh and the change gv in the normalized lifetime τn2,v of an optical dipole μ2 of vertical orientation θv are determined over a distance range Δh2ref ranging for example from 0 nm to 300 nm.
In a sub-step 330, the normalized lifetime τn2,h of the optical dipole μ2 of chosen orientation is compared with the normalized lifetime τn2,v of the optical dipole μ2 of non-chosen orientation τn,v for all h2 within Δh2ref. The selectivity parameter S2 is advantageously determined such that, for all h2 within Δh2ref, the absolute value |S2| of the selectivity S2 is equal to the absolute value of the difference between the normalized lifetime τn2,s of the optical dipole μ2 of chosen orientation and the normalized lifetime τn2,ns of the optical dipole μ2 of non-chosen orientation, i.e. here: |S2(h2)|=|τn2,s(h2)−τn2,ns(h2)|.
In a sub-step 340, a value h2s is then determined such that the normalized lifetime τn2,s of the optical dipole μ2 of chosen orientation θs is longer than the normalized lifetime τn2,ns of the optical dipole μ2 of non-chosen orientation θns, i.e. here such that τn2,s(h2s)>τn2,ns(h2s). Additionally, the value h2s is advantageously determined such that the selectivity |S2(h2s)| is maximum over at least one domain in the range Δh2ref, and preferably over the entire range Δh2ref, in other words |S2(h2s)|=maxΔh2ref(|S2(h2)|).
In a step 400, the light-emitting diode 1 is produced such that the emitting dipoles μ1 are located at the distance h1s from the extraction layer 6. In the case where the active layer 13 has fixed dimensions, for example 3 nm for a quantum well layer and 12 nm for a barrier layer, the thickness of the first layer 11 is determined such that the sum of this thickness and of half of the active layer 13 is equal to the determined value h1s, for example to within 10 nm, or even to within 5 nm.
Thus, the emitting dipoles μ1 associated with the radiative recombination of the electron-hole pairs in the active layer 13 essentially have the vertical orientation θv. The luminous efficacy of the light-emitting diode 1 is thus improved by a near-field non-radiative coupling of dipole-dipole type between the emitting dipoles μ1 and the optical dipoles μ2.
Additionally, the thickness of the extraction layer 6, in particular the thickness of the dielectric material 6.2, the nanoscale particles 6.1 advantageously remaining at the interface with the first layer 11, is determined such that the optical dipoles μ2 are located at the distance h2s with respect to the interface with the air (face 6a), and thus have the chosen orientation θ2s. This also contributes to improving the luminous efficacy of the light-emitting diode 1.
Additionally, the production method may comprise a phase of dimensioning the second layer 12 arranged between the active layer 13 and the reflective electrode 4 (anode). Thus, it is also advantageous to dimension this layer 12 such that the emitting dipoles μ1 have the vertical orientation θ1v.
In this example, the first uniform medium of refractive index n1 is formed of the second layer 12 made of P-doped GaN and of an emissive layer made of InGaN (no barrier layer made of intrinsic GaN). The fourth uniform medium is formed by the reflective electrode 4 made here of gold, the optical index n4 of which is equal to 1.3489+i×1.8851 at 460 nm.
The normalized lifetime τn1′ of the emitting dipoles μ1 having a horizontal or vertical orientation increases from a zero value for h1′ equal to zero to a first peak, then tends toward one and the same constant value while exhibiting damped oscillations. The constant values are substantially equal for both orientations. However, since the oscillations are not in phase with one another, there are domains Δhi=1,2 . . . in the range Δh1ref for which the normalized lifetime associated with one of the orientations predominates over the normalized lifetime associated with the other orientation. It is then possible to dimension the light-emitting diode 1, in particular to choose a value of the distance h1s′, such that the emitting dipoles μ1 predominantly have the vertical orientation.
In this example, the emitting dipoles μ1 of vertical orientation predominate in the domains Δh2, Δh4, Δh6 in the range Δhref going from zero to 300 nm. Furthermore, the emitting dipoles μ1 of horizontal orientation predominate in the domains Δh1, Δh3, Δh5, Δh7. Thus, a distance h1s′ in the domain Δh2 or Δh4, for example, will make it possible to obtain radiative recombination of the electron-hole pairs, the emitted radiation of which corresponds to that of emitting dipoles μ1 having an essentially vertical orientation.
As mentioned above, to obtain a preponderance of the emitting dipoles μ1 having the chosen orientation over those having the non-chosen orientation, it is important for the lifetime of the emitting dipoles μ1 having the chosen orientation to be longer than that of the emitting dipoles having the other orientation.
The selectivity S1′ thus cancels out between each domain Δhi and has a maximum value for each of them. By way of example, the selectivity has a value of 0.18 for 18 nm in Δh1 (horizontal orientation), a value of 0.36 for 46 nm in Δh2 (vertical orientation), a value of 0.15 for 90 nm in Δh3 (horizontal orientation), and a value of 0.08 for 140 nm in Δh4 (vertical orientation). It is therefore seen that the various domains do not have a selectivity S1′ of the same intensity, indicating that the normalized lifetimes change according to the distance h1′ in the form of damped oscillations.
Thus, to improve the luminous efficacy of a light-emitting diode 1, it is possible to dimension it such that the emitting dipoles μ1 have a vertical orientation. For that, the distance h1s′ is chosen in one of the domains Δhi in the range Δh1ref for which the lifetime of the emitting dipoles μ1 having the vertical orientation is longer than that of the emitting dipoles μ1 having the horizontal orientation. In addition, the distance h1s′ is advantageously determined such that the selectivity S1′ has a maximum in the determined distance range Δh1ref. In this example, to obtain a predominance of the emitting dipoles μ1 of vertical orientation, it is advantageous for the distance h1s′ to be equal to 54 nm given that the selectivity S1′ has a maximum equal to 0.38.
Particular embodiments have just been described. Various modifications and variants will be apparent to a person skilled in the art.
| Number | Date | Country | Kind |
|---|---|---|---|
| 19 15503 | Dec 2019 | FR | national |
