FR2985374A1 - PHOTOVOLTAIC PANEL WITH DIODES MOUNTED IN PARALLEL WITH DIFFUSING CENTRAL STRUCTURE AND RE-REFLECTIVE REAR STRUCTURE - Google Patents

Abstract

The invention applies to a photovoltaic cell. According to the invention, two diodes, one comprising a semi-conductor layer preferably in pmSi and the other comprising another semiconductor layer, preferably SiGe, the two layers being of suitable dimensions, are connected in parallel between them, the cells being mounted in series to form a photovoltaic panel. The diodes connected in parallel are deposited in two modules comprising a different substrate and separated by a diffusion encapsulant comprising fine particles of index and dimension adapted to maximize forward diffusion and minimize backward diffusion, said parameters being optimized by applying the laws of Mie. The rear module advantageously comprises a mirror reflector for maximizing optical trapping.

Description

The present invention applies to the field of the production of electrical energy from photovoltaic panels. BACKGROUND OF THE DISCLOSURE Despite its high cost of production, this area has experienced a great expansion, especially in the production of cells based on amorphous silicon wafers, mono or poly-crystalline. Among the cost reduction pathways, thin-film deposition techniques are promising as they benefit from improvements introduced in processes and equipment for the production of semiconductor electronic components. However, the development of this sector remains hampered by a lower electrical conversion efficiency achieved so far (less than 10% for the amorphous silicon / microcrystalline silicon thin-film system the most efficient - technology called micro-morph, against more than 15 % for traditional crystalline silicon channels A promising way to significantly increase the efficiency of electrical conversion is opened by the so-called "split" technology which combines two series of diodes each having different light absorption spectra in parallel assembly. This principle is disclosed in the French patent application published under No. FR 2948498 belonging to the applicant In a split diode cell, the conversion efficiency is increased by optical trapping by incorporation of diffusing particles between the two series of diodes each different light absorption spectra. There are options for improving the performance and efficiency of this optical trapping. Incorporation of fine particles into one or both sets of diodes and alternately or complementarily in a polymer layer between these two electrodes. The realization of these solutions is possible according to the teachings of the patent application. quoted above. However, among all possible combinations of materials and particle sizes, it is necessary to make a choice that is optimal for a targeted application. The present invention makes it possible to choose an optimal combination and to obtain a split cell with a greatly improved yield (of 20-30% relative to the micro-morphine cells). For this purpose, the invention provides a photovoltaic cell comprising: a first module comprising, on a first transparent substrate with incident light, a first and a second electrode enclosing a first material chosen for photoconverting light radiation of length wave less than a threshold wavelength and substantially transparent to the light radiation of wavelength greater than said threshold, the first and second electrodes and the first material deposited in thin layers on the first substrate, A second module comprising on a second substrate, a third and a fourth electrode enclosing a second material chosen to photo-convert a light radiation of wavelength greater than said wavelength threshold, the third and fourth electrodes and the second material deposited in layers on the second substrate, said first and second modules being integrally together, the second and fourth electrodes being separated by an encapsulant of refractive index n1 and being electrically connected by two parallel electrical connections of the 1st and 3rd electrodes and the 2nd and 4th electrodes said photovoltaic cell being characterized in that at least one of the elements selected from the group of the encapsulant and the second and fourth electrodes of the scattering particles have a refractive index n2 greater than n1 and a characteristic dimension d selected from maximize the scattering angle and minimize backscattering at wavelengths above said threshold. Advantageously, said scattering particles are inserted into the diffuser encapsulant.

Advantageously, said scattering particles are inserted into at least one of the second and fourth electrodes. Advantageously, the third electrode is constituted by a mirror reflector with controlled roughness. Advantageously, the encapsulant consists of a polymer. Advantageously, the 1st, 2nd and 4th electrodes are in Transparent Conductive Oxide. Advantageously, the first material is chosen for an optimal photoconversion of the visible light. Advantageously, the thickness of the first material is chosen for an optimal photo-conversion of the visible light at the end of the life of the cell. Advantageously, the first material is a large gap amorphous material. Advantageously, the second material is a small gap material. Advantageously, the second material is chosen from the group comprising SiGe, pc-Si, poly-Si and CIGS. Advantageously, the refractive index n2 and the characteristic dimension d are chosen by applying the laws of Mie. Advantageously, the application of the Mie law is made at a wavelength substantially equal to the average of the absorption range of the second module. Advantageously, the refractive index n2 is at least 2.5 and the characteristic dimension d is substantially equal to 0.3 microns.

The invention also discloses a solar panel comprising at least two photovoltaic assemblies according to one of claims 1 to 14, the first set comprising first cells C1 connected in series by their long side of dimension L, and the second set comprising second C2 cells connected in series by their long side of the same dimension L and each having a dimension L x 11 and L x 12 respectively, said panel being characterized in that the ratio of dimensions 11 and 12 of said first and second cells of said sets is equal to the ratio of open circuit voltages of C1 and C2.

The invention has the additional advantage of allowing fine tuning of the optimization of optical trapping as a function of the wavelength of the light processed by the rear diode. In addition, in a split cell according to the invention, the presence of a distinct rear substrate on which the rear diode is deposited makes it possible to achieve a state of controlled roughness surface on which a mirror-controlled, mirror-controlled, reflective layer will advantageously be deposited. which will increase the effect of optical trapping. Finally, structurally, a split cell makes it easier to benefit from any improvement in the materials that can be used to increase the efficiency of each of the diodes of the cell without having to worry about the constraints on these materials that would make it impossible to use them in a single cell. successive deposit structure. The invention will be better understood, its various features and advantages will emerge from the following description of several embodiments and its accompanying figures, of which: FIGS. 1 a and 1b respectively illustrate a photovoltaic cell with a micro-morph structure and split structure; FIGS. 2a, 2b and 2c show detailed views of the structure of the mirror reflector of FIG. 1 in two different embodiments of the invention; FIGS. 3a and 3b illustrate the scattering effect of the light rays at the layer interfaces respectively in a photovoltaic cell with a micro-morph structure of the prior art and in a split structure in an embodiment of the invention; FIGS. 4a, 4b, 4c and 4d show detailed views of the structure of the central part of the split structure according to three different embodiments of the invention; FIGS. 5a, 5b and 5c respectively represent the angular distribution curves of the diffusions due to the encapsulation layer in several embodiments of the invention; FIGS. 6a and 6b show the reflection effect at the rear respectively in a micromorphic structure photovoltaic cell of the prior art and in a split structure in one embodiment of the invention; FIGS. 7a and 7b respectively represent the current / voltage curves of the front and rear diodes in one embodiment of the invention.

FIGS. 1 a and 1 b respectively illustrate a photovoltaic cell with a micro-morph structure and a split structure. One type of photovoltaic cells of the prior art is shown in FIG. Such cells are described in particular in US Pat. No. 5,085,711. They are called "micro-morphs" in that they combine an amorphous silicon layer (aSi) 130a and a microcrystalline silicon layer (pSi) 140a. Amorphous silicon does not convert long wavelength (infrared) light rays; said rays are converted by the microcrystalline silicon layer. The assembly of Figure 1 has an electrical module whose connections are located at 1210a and 1510a and connect the external circuits two electrodes 120a and 150a, the first zinc oxide (ZnO), the second ZnO or reflective material. The incident light enters the module through a first substrate-forming glass layer 110a. The module is closed by an encapsulant layer 160a and a second glass layer 170a. A type of split-structure photovoltaic cells is shown in Figure 1b. The principle of such cells has been described in the patent application FR No. 2948498. The structure is improved in the context of the present invention. A cell of this type consists of two modules 2 9853 74 6 11 b and 12b. A first module 11b is constituted by depositing on a first glass substrate 110b a first conductive layer 120b and a second conductive layer 140b, said two layers being for example made of zinc oxide and enclosing a layer 130b constituted by a 5 material converting light rays of wavelength lower than a threshold. This threshold depends on the gap of the material and its thickness. For this first module, a so-called "large gap" material will be used which may for example be pmSi (polymorphous silicon), or amorphous silicon or any other material having a gap greater than 1.6 eV. A second module 12b 10 is constituted by a second glass substrate 190b on which a third conductive layer (a multilayer metal / TCO (Transparent Conducting Oxide or Conductive Transparent Oxide), for example Ag / ZnO) forming a 180b mirror reflector is deposited. , a layer of material called "small gap" 170b and a fourth conductive layer 160b, for example also ZnO. The small gap material is chosen to convert light rays of longer wavelength to the threshold of the first module. For example, an alloy of silicon germanium (SiGe), microcrystalline silicon (p-cSi), CIGS (Copper-Indium-GaliumSelenium) or any other material having a gap of less than 1.5 eV which can be used can be chosen. adjust the thickness and the optical characteristics so as to increase the generated current, without being limited by the current of the first module, as will be explained later in the description. The semiconductor layers 130b and 170b of each of the two modules 11b and 12b are therefore deposited on different substrates, which makes it possible to optimize both the choice of materials, the thickness of the layers and the parameters of temperature and deposition pressure. In particular, it is thus possible to optimize the diffusion of the front electrode for an optimal diffusion of the visible light, for a maximum efficiency of the front module, 11b. It is also possible to provide a thickness of this module before which allow to preserve maximum efficiency at the end of life, despite degradation under light. Taking into account the aging under light is, however, impossible in the micro-morphic structures of the prior art because the modification of the active layer of the first module would have a direct impact on the optical performance of the second module. The micro-morphic structures use an insertion of high index layers, for example TCO (ZnO) or SiO 2 between the diode a-Si and the diode pcSi to overcome this problem. But this method is constraining vis-à-vis the interfaces between the layers, on the one hand, and on the other hand does not fully satisfy the requirements of diffusion up short wavelengths and at the same time the downward diffusion of long wavelengths. The two modules 11b and 12b are assembled by bonding the layers 140b and 160b via a layer 150b of encapsulant. . The glass substrate of the first module 11b is exposed to light rays, said first module constituting the front face of the photovoltaic cell. The substrate of the second module constitutes the rear face of the cell. The technical function of the layers 140b and 160b is both electrical (charge collection) and optical in the context of the present invention by ensuring the maximum diffusion of the long wavelength light having passed through the first module. In the same way, the technical function of the diffuser encapsulant is both to achieve the mechanical connection between the two modules and the maximum diffusion of the long wavelength light having passed through the first module. The means used in the context of the present invention to perform this second function will be detailed below in relation to FIGS. 2a and 2b, then 5a, 5b and 5c. The module 11b constitutes a current generation module whose first conductive layer 120b has a connection with the outside via a 1st electrode 1210b and the second conductive layer 140b also has a connection with the outside via a 2nd electrode 1410b . The 3rd and 4th conductive layers (180b and 160b) included in the second module 12b respectively have connections with the 3rd and 4th electrodes 1810b and 1610b. The two modules 11b and 12b thus constitute two diodes. Advantageously, they are mounted in parallel, connecting the 1st and 3rd electrodes between them and the 2nd and 4th electrodes between them. The advantages of parallel mounting include that the current of the first module is not limiting for the second module and the possibilities of microcrystalline silicon (or SiGe) can be exploited to the maximum to maximize the current of the second module. The respective widths 11 and 12 of the cells of the first and second modules must be chosen such that the voltages at the 1st and 3rd electrodes and 2nd and 4th electrodes are equal.

Figures 2a, 2b and 2c show detail views of the structure of the mirror reflector of Figure 1 in two different embodiments of the invention. As illustrated in FIG. 2a and indicated in comment on FIG. 1, the material 182b can be a TCO. The mirror reflector 181b may be a metal mirror and the substrate 190b may be made of glass. The assembly 181b + 182b represents the mirror reflector 180b of FIG. 1b. The interfaces between the metal mirror and the glass are advantageously of controlled roughness. Two modes of roughness control of this interface are illustrated in Figures 2b and 2c. FIG. 2b schematically shows the result of an interface layer deposition on the glass before deposition of the metal mirror, said layer consisting of particles typically having a characteristic dimension of the grating in the plane of 0, 5 to 2 microns and a thickness of between 0.1 and 1 micron. This deposit can be achieved either by so-called "bottom-up" techniques by manufacturing these particles directly by in-situ methods such as vacuum deposition (PVD, PECVD, etc.), or by techniques known as "top-down" techniques. starting from particles already manufactured elsewhere and deposited either directly by Spray methods for example, or mixed in a liquid and deposited on the substrate following the evaporation of the liquid (SOLGEL, etc.). In FIG. 2c, the roughness is achieved by etching either in the metal or on the substrate before metal mirror deposition. The characteristic dimensions are equivalent to those made by deposit. The etching may be of the wet or dry type. FIGS. 3a and 3b illustrate the scattering effect of the light rays at the layer interfaces respectively in a photovoltaic cell with a micro-morph structure of the prior art and in a split structure in an embodiment of the invention.

In a micro-morph structure as shown in Figure 2a, no scattering structure is specifically provided. The diffusion effect therefore results from the roughness of the interfaces between the layers. The same interface performs the diffusion function at the wavelengths converted by the first semiconductor layer (310a), and the wavelength diffusion function converted by the second semiconductor layer (320a), which does not will not maximize optical trapping in this second layer. Indeed, optical trapping may possibly be adapted for the wavelengths converted by the first semiconductor layer but not for those converted by the second layer. Indeed, insofar as the deposits of the layers are in conformity, the states of roughness at the interfaces will be substantially identical and the diffusion at longer wavelengths will not be optimal, if it is at shorter wavelengths. .

On the other hand, in a split diode, the optimizations at the two wavelength bands, 310b and 320b, are carried out separately, the first at the interface between the conductive layer 120b and the semiconductor layer 130b, the second at the interface between the conductive layer 120b and the semiconductor layer 130b. the electrodes 140b and 160b, either in the diffuser encapsulant 150b or in any combination of these three layers. The constitution of the diffusing layer may be chosen to optimize the diffusion at the wavelengths converted by the second diode. The different scattering layers (140b, 150b, 160b) can be charged by particles to optimize infrared scattering in the direction of light propagation. This will increase the optical path in the rear sub-module so its current. In the case of the invention, the charges used in the different layers of the two panels are intended to diffuse incident light transmitted through the first diode with an angle as large as possible relative to the axis of propagation of light by maximizing backscatter. The fact that the two modules 11b and 12b are separated makes it possible to have access to the central part of the device and thus to improve the optical structure of the device. Thus, the introduction into the central layers of fine particles makes it possible to diffuse the light that has passed through the first module. According to the constitution of the semiconductor layer of this first module, the wavelength threshold of the unabsorbed light beams and diffusing towards the second module will be chosen around 850 nm and the majority of this light will be located in the infrared or near infrared. The purpose of this diffusion is not, as in the case of conventional modules, to return light to the front diode but to diffuse it with the widest possible diffusion angle (between 0 and 90 ° with respect to the axis). incident light) in the second module (rear module). The laws of physics allow us to calculate the size of particles that need to be incorporated. This must be between 0.1 and 1.5p, the material must be non-absorbing with a refractive index n2 as high as possible (n1 being the refractive index of the material in which the particles are incorporated). The concentration and size of the particles must be defined according to the thickness of the layer to have a maximum angle of diffusion and no or very little backscattering.

Depending on the size of the particles, the laws of diffusion of the light will be different: If the product n2 x d is very much smaller than the wavelength of the light, one is in a Rayleigh scattering model; the light is diffused at 360 °; this configuration is disclosed in US20100059101 is not that retained in the context of the present invention; If the product n2 x d is of the order of magnitude of the wavelength of the light, by selecting a pair (n2, d) the diffusion angle is modified with respect to the axis of propagation of the light; we will also try to minimize the backscattering of light. Mie's theory makes it possible to calculate the particle size best suited for maximum diffusion with maximum angle while minimizing backscattering. The theory of Mie is based on the resolution of Maxwell's equations in the case of a sphere of radius R consisting of a material of index np dipped in a medium of index ri. Let us give here as an illustration the Mie equations giving the scattering cross sections: L) 2 Chf = (2L + 1) (aL 12 + 11 / L12) With x = kR, where k = 27c IX, X being the length wave. 2 9853 74 11 With L being the orde of multipolar field development as a function of spherical symmetry. If we take into account only the first order, that is to say L = 1, in this case we have: a = 1111P 1, (111x) VL (x) -IFL (x) '' L (rnx ) n / YL (nix); (x) (x) VL (lm) 5 With m = nlin, and IPLetLand the cylindrical functions of Riccati-Bessel. One skilled in the art will have the means to solve these equations in the case of application of the present invention. For further details reference may be made, for example, to the publication of Bohren CF, Huffman D. R., "Absorption and Scattering of Light by Small Perfides", John Wiley and Sons, Inc., New York, NY, first edition (1983). Given the different materials that can be used for the rear cell (the second module), the calculation of the particle size depends on the wavelength that will be preferred. In the case of a material with a gap of 1.1eV we will try to maximize the diffusion for a wavelength of the order of 850nm (near IR). This value represents the median wavelength of the light received by the second module after it has passed through the first module. For example, for a module before 20 with a large gap material of 1.7 eV and a small gap rear module of 1.1 eV this median will be in the range 700-1000 nm which gives a median at 850 nm which has served as an example in the diffusion calculations. The modeling results are given later in the description. In the case of the incorporation of the particles into the polymer, at the interface of the two sub-modules, the latter must be transparent and provide two main functions, serve as a dispersing matrix for the particles and encapsulant to ensure the panel's protection against climatic elements. This polymer may for example be EVA (Ethylene Vynil Acetate), PVB (Polyvinyl Butyral) or silicone or any other insulating and transparent elastomer for encapsulating solar panels with a service life greater than 20 years.

Figures 4a, 4b, 4c and 4d show detail views of the structure of the central portion of a diode according to three different embodiments of the invention. Figure 4a shows Figure 3b and shows three possible embodiments of the invention shown in Figures 4b, 4c and 4d. In FIG. 4b, the scattering particles whose index and size are calculated according to the method discussed above in relation with FIG. 3b are incorporated in the electrode 140b. In FIG. 4c, said scattering particles are incorporated in the diffusion encapsulant layer 150b. In FIG. 4d, the scattering particles are incorporated in the electrode 160b. In all three cases it is the index of the medium which changes (nm), so that the ratio m of indices between the particle np and the medium m = np / nm (as defined above) will change and have an influence on the diffusion of light. The size of the particle should in principle be adapted for maximization of diffusion. Nevertheless, in our practical case, the difference in index between the encapsulating material (generally around 1.5) and the TCO (generally 1.9-2 at 600 nm but which is even smaller for long wavelengths). ) is not excessively large to significantly influence the outcome of the broadcast. Indeed, as mentioned above, the size of the particle is chosen for maximum diffusion for an average wavelength on the spectrum. This means that a variation of 10 to 20% of the diameter of the particle can be tolerated. Otherwise, if necessary, a slight adaptation of the particle size can be made when the particles are inserted into the TCO. In the case of integration into the TCO, the incorporation of the particles into the TCO in contact with the small gap absorber will be preferred. Figures 5a, 5b and 5c respectively show the angular distribution curves of the diffusions due to the encapsulation layer in several embodiments of the invention. These figures are only examples of the incorporation of the particles into the encapsulant. Likewise, identical curves could have been made to establish the same angular distribution curves in the cases of the embodiment of the invention by incorporation of scattering particles in the electrodes 140b and / or 160b. The following hypotheses are used as illustrative examples: n2 (index of fine particles implanted in the polymer matrix) 2.61 - n1 (Index the polymer matrix of the encapsulant): 1.5 - Wavelength in vacuum: 850 nm - d = diameter of the fine particle in p The calculations according to the modelization of Mie then give the following values: D 0.04 0.1 0.2 0.3 0.4 0.5 0.6 Efficiency of 0.001 1 0.043 0.657 2.773 3.75 4.09 4.712 forward diffusion Efficiency of the 0.001 5 0.055 0.413 0.215 1 0.652 1.655 1.983 diffusion to the back D 0.6 0.7 0.8 0.9 1 1.5 5 Effectiveness of the diffusion 4.712 3.57 2.759 1.58 1.92 2.44 2.17 forward Effectiveness of the broadcast 1.983 7.716 12.06 5.528 8.98 5.94 24.45 backward Forwarding is often referred to as "forward scattering". Backcasting is often referred to as "backward scattering". The best solution in the preferred embodiment is then obtained for 0.3 p particles where the forward scattering report on back scattering is the strongest. Moreover this information is confirmed by the angular diffusion distribution: FIG. 5a: angular distribution for d = 0.2p and wavelength in vacuum 850 nm; Figure 5b: angular distribution for d = 0.3p and wavelength in vacuum 850 nm; Figure 5c: Angular distribution for d = 0.4p and wavelength in vacuum 850 nm.

The different graphs show that for particles of 0.2 p there is a bit of back scattering, that for d = 0.3 p the diffusion angle is well open with no back scattering and that for d = 0.4 p we see reappear a little back scattering with a decrease in the scattering angle. A particle size of 0.3p is therefore the best suited for the calculation taking the following hypotheses, wavelength 850 nm, polymer matrix having an index of 1.5 and particles having an index of 2.61. This calculation can be repeated as a function of the various parameters, the index of the medium receiving the particles, the index of the charge particles and the wavelength. Most polymeric matrices having an index of 1.5, the optimum size of the particles allowing diffusion with the greatest possible angle and with the lowest possible backscattering will be of the order of 0.1 to 2p maximum. The diameter range of the particles is included in the very wide envelope given for other applications, in particular in the patent US20100059101. On the other hand, this document clearly explains (paragraph [0090]) that the purpose of the load is to diffuse light in both the aSi diode and the pSi. This leaves no ambiguity that, according to the disclosure of this invention, what is sought is forward scatter and backscatter. This document determines a first range between 0.1 and 10 p and then a smaller optimum range between 0.5 and 2 p, higher than the values obtained in the case of application to the device of the invention. FIGS. 6a and 6b show the rear reflection effect respectively in a micro-morph structure photovoltaic cell of the prior art and in a split structure in an embodiment of the invention. The structure of the device according to the invention makes it possible to deposit the reflective layer with controlled roughness on the glass. This roughness makes it possible to have optimal diffusive reflection with a low absorption of the mirror. . This mirror will reflect the light scattered by the particles included in the layers 140b, 150b and 160b and thus allow the latter to increase its number of paths through the absorbent layers.

As described in FIGS. 2b and 2c, the achievement of the controlled roughness can be obtained either by deposition of aggregates of the order of one micrometer arranged regularly on the glass, or by micrometric etching of the glass. The diffusive mirror will be obtained by conformal deposition of a metal on this structure. FIGS. 7a and 7b respectively represent the current / voltage curves of the front and rear diodes in one embodiment of the invention. The electrical connection of the diodes of each of the modules (front or front and rear) manufactured and optimized separately is performed without loss of efficiency of each of the modules. The conversion efficiencies of the two diodes, the values of which are indicated in FIGS. 7a and 7b, are preserved by the parallel connection of the two diodes. The acronyms in the figures have the following meanings: Jsc: current density (mA / cm2) of the photovoltaic cell; Voc: Open circuit voltage (V) of the cell; FF: Shape factor (%) of the cell = VMax x JMax / Jsc x VOC Efficiency: Jsc x VOC x FF / incident power (a value of 1 kW / m2 corresponds to the Spectra AM standard 1 .5) We will assemble the photovoltaic cells according to the invention in solar panels for themselves to be grouped in electricity production units and to be mounted on the site of production of electrical energy and to be connected to the local network (house, building .. .) or extended, in the case of assemblies of a number of cells. Advantageously, a first set of first large-dimension C1 cells of dimension L and small size side 11 are put in series by their long side and a second set of second cells C2 of dimension L × 12 also mounted in series are combined. The dimensions 11 and 12 are chosen so that their ratio is equal to the ratio of the open circuit voltages across C1 and C2. The examples described above are therefore given by way of illustration of some of the embodiments of the invention. They in no way limit the scope of the invention which is defined by the following claims.

Claims (15)

REVENDICATIONS1. Photovoltaic cell comprising: a first module (11b) comprising, on a first substrate (110b) transparent to an incident light, a first and a second electrode (120b, 140b) enclosing a first material (130b) selected for photoconverting radiation light of wavelength less than a threshold wavelength and substantially transparent to the light radiation of wavelength greater than said threshold, the first and second electrodes and the first material deposited in thin layers on the first substrate, A second module (12b) comprising on a second substrate (190b), a third and a fourth electrode (160b, 180b) enclosing a second material (170b) selected for photoconverting light radiation of wavelength greater than said threshold length of d wave, the third and fourth electrodes and the second material deposited in thin layers on the second substrate, said first and second modules being integrally mounted together, the second and fourth electrodes being separated by an encapsulant (150b) of refractive index n1 and being electrically connected by two parallel electrical connections of the 1st and 3rd electrodes (1210b, 1810b) and of the 2nd and 4th electrodes (1410b, 1610b), said photovoltaic cell being characterized in that are inserted in at least one of the elements selected from the group of the encapsulant (150b) and the second and fourth electrodes (140b, 160b) diffusing particles having a refractive index n2 greater than n1 and having a characteristic dimension d chosen to maximize the scattering angle and to minimize backscattering at wavelengths above said threshold. 2. Photovoltaic cell according to claim 1, characterized in that said scattering particles are inserted into the diffuser encapsulant (150b). Photovoltaic cell according to claim 1, characterized in that said diffusing particles are inserted in at least one of the second and fourth electrodes (140b, 160b). 4. Photovoltaic cell according to one of claims 1 to 3, characterized in that the 3rd electrode is constituted by a mirror reflector with controlled roughness. 5. Photovoltaic cell according to one of claims 1 to 4, characterized in that the encapsulant (150b) consists of a polymer. 6. Photovoltaic cell according to one of claims 1 to 5, characterized in that the 1st the 2nd and 4th electrodes are Transparent Conductive Oxide. 20 7. Photovoltaic cell according to one of claims 1 to 6, characterized in that the first material is chosen for an optimal photoconversion of visible light. 25 8. Photovoltaic cell according to one of claims 1 to 7, characterized in that the thickness of the first material is chosen for optimal photoconversion of visible light at the end of life of the cell. 9. Photovoltaic cell according to one of claims 1 to 8, characterized in that the first material is a large gap amorphous material. 10. Photovoltaic cell according to one of claims 1 to 9, characterized in that the second material is a small gap material. 35 2 9853 74 19 Photovoltaic cell according to claim 10, characterized in that the second material is selected from the group consisting of SiGe, pc-Si, poly-Si and CIGS. 5 12. Photovoltaic cell according to one of claims 1 to 11, characterized in that the refractive index n2 and the characteristic dimension d are chosen by applying the laws of Mie. 13. Photovoltaic cell according to claim 12, characterized in that the application of the Mie law is made at a wavelength substantially equal to the average of the absorption range of the second module. 14. Photovoltaic cell according to claim 11, characterized in that the refractive index n2 is at least 2.5 and the characteristic dimension d is substantially equal to 0.3 microns. 15. Solar panel comprising at least two photovoltaic assemblies according to one of claims 1 to 14, the first set 20 comprising first cells C1 in series by their long side of dimension L, and the second set comprising second cells C2 put in series by their long side of the same dimension L and each having a dimension L x 11 and L x 12 respectively, said panel being characterized in that the ratio of the dimensions 11 and 12 of said first 25 and second cells of said sets is equal to the ratio open circuit voltages of C1 and C2.