FR2917898A1 - Solid photovoltaic device for e.g. roof of building, has absorber component composition disposed in continuous absorption layer, where absorption layer is formed of stibnite and argentite-based/stibnite and copper sulphide-based component - Google Patents

Abstract

The device has an absorber component composition disposed in a fine continuous absorption layer (2) between transparent n-type and p-type semi-conductor components. The components are presented in a form of a porous substrate (1) with pores (1-1). The substrate is made of crystals. An internal surface of the pores is covered with the layer. The pores are filled in volumic proportion of 15 percentages by a filling layer (3) constituted of the n-type or p-type semi-conductor component. The absorption layer is formed of stibnite and argentite-based or stibnite and copper sulphide-based component.

Description

The present invention relates to all-solid photovoltaic devices

  of the type comprising 3 inorganic and solid materials comprising respectively: - an n-type and transparent semiconductor compound (1), - a p-type and transparent semiconductor compound (3), and - a composition of at least one an absorber compound arranged in a continuous layer (2) between said n-type semiconductor compounds and which are not in contact, and - said n-type or p-type semiconductor compound in the form of a porous substrate (1 ), with pores (11) of size 10 to 100 nm, the inner surface of said pores is fully coated with a thin continuous layer of absorption (2) of said absorber compound, and of which said pores are then filled in a a volume proportion of at least 10%, preferably more than 15%, by a filler layer (3) consisting of the other p-type semiconductor solid compound or respectively n, The conventional photovoltaic devices consist of a junction two layers of semi- conductors, one of type p (where the majority charge carriers are holes, or positive charges) and the other of type n (where the majority carriers are electrons, or negative charges). The most widely used photovoltaic devices are those based on silicon (Si), which consist of a homojunction, in which the two layers are made of Si, one of the layers being doped n and the other p. Other known photovoltaic devices are based on heterojunctions, in which the two layers consist of two different semiconductor materials, such as devices comprising CdTe / CdS or CuInSe2 / CdS layers. Both types of devices are manufactured in a configuration with two stacked flat layers. Apart from the two semiconductor layers, other layers may be present in the final device, such as front and back contact layers, buffer layers, etc., but still keeping the configuration flat and stacked. The principle of operation of this type of device is as follows: when the solar radiation enters the junction, the photons are absorbed in the semiconductor layers, they are converted into electron pairs / holes, and a separation of charges takes place ; Then the electrons move to the n-side contact, and the holes to the p-side contact.

  One of the problems of this type of junction is that of having to use materials of high purity and with a very controlled doping, this to avoid the losses due to the recombination of the charges. Indeed, the two opposite charge carriers move within the same material and, in the presence of certain recombination centers such as impurities, lattice defects, grain boundaries, etc., they can recombine and are not not converted into electricity. This recombination phenomenon is a major factor of decrease in photovoltaic efficiency. The precise purification and doping operations involve the use of sophisticated techniques and considerably increase the manufacturing cost. However, cost is a limiting factor when this type of device wants to be used for large-area panels in the application as a domestic photovoltaic generator in the roofs of homes. The typical power rating of these systems is in the order of 1-5 kilowatts (at the point of maximum power), which represents a panel area of the order of 10 to 30 m2 for a nominal photovoltaic efficiency typical of order of 10%. Another problem with devices based on pn heterojunctions is that of using certain toxic compounds, such as those based on Cd, Se or Te, or that of using elements whose availability is very low, such as linen. Examples of interpenetrated cells that are made with quantum dots based on (inter alia) antimony sulphide are described in Vogel et al., J. Phys. Chem., 93, 3183 (1994), but have the disadvantage of use a liquid electrolyte, which is unstable. Other examples that use quantum dots are described in US2002 / 017656, 2001, which claim a lead sulfide composition, PbS, and cite other possible compositions such as Sb2S3 antimony sulfide, among others, but have the disadvantage of filling the pores with a compound hole conductor consisting of an organic molecule, which makes them unstable. In all cases the results in terms of photovoltaic efficiency are not satisfactory. Other technologies for converting radiation into electricity have emerged claiming as a benefit a low manufacturing cost. In particular, a new type of photovoltaic configuration has been proposed in publications [Siebentritt, Koenenkamp et al., 14th European Photovoltaic Solar Energy Conference & Exhibition Proceedings, p. 1823, Barcelona 1997); Rost, Koenenkamp et al., 2nd World Conference & Solar Photovoltaic Solar Energy Conversion Proceedings, p 212, Vienna 1998]. It consists mainly of three inorganic and solid materials arranged in an interpenetrating configuration, different from that of stacked flat layers. The device consists of a "transparent" and n-type semiconductor substrate present in the form of a porous layer, which is covered in its inner surface by a thin layer of an absorber and semiconductor material, finally the layer being filled in its pores with a "transparent" material and semiconductor p-type. Their operating principle is as follows: the radiation (photons) is only absorbed by the absorber material; then an electron is transferred to the n-type "transparent" (or non-absorber) semiconductor, while a positive charge is transferred to the semiconductor p. According to this principle, the transport of opposite charge carriers is carried out in two differentiated phases, which theoretically considerably reduces the drop in performance by recombination of charges. The use of the nanocrystalline or microcrystalline porous layers of the semiconductor substrate is necessary to provide a sufficient amount of absorber present in the layer per unit area projected. The main advantage of this type of interpenetrated configuration compared to known devices with stacked flat layer configuration is the possibility of producing low cost manufacturing photovoltaic devices because cheap materials are used which do not have need to be very pure and manufacturing techniques are not expensive (in particular, we do not use energy-intensive techniques, unlike technologies involving the implementation of silicon). This configuration also has other advantages over others that use interpenetrated configurations. First, having the absorber in the form of a thin layer, rather than in the form of molecules (as in so-called dye-sensitized solar cells, such as those described in Desilvestro et al., J. Amer. Chem 107, 2988 (1985) and in US Pat. by the greater amount of absorbing material, and the absence of inter-facial recombination due to the absence of contact between the n and p semiconductors, and then the use of solid and inorganic materials makes it possible to obtain more stability a long time, compared to other technologies using liquid electrolytes or, in their solid version, hole-conductive organic materials, which are sensitive to high temperatures and high solar irradiation (as is the case case for outdoor applications) and degrade. The devices manufactured so far with the interpenetrated all solid and inorganic configuration have shown a very low photovoltaic efficiency. Examples based on this concept have been presented in some publications [Tennakone et al., J. Phys. D Appl. Phys. 31, 2326 (1998) and Kaiser, Konenkamp et al., Solar Energy Materials & Solar Cells 67, 89 (2001)], in which porous layers of TiO2 (based on crystals or grains), absorber coverings based on Se or CuInS2 made respectively by electro-deposition or by the ILGAR technique (impregnation of Cu and In precursors followed by treatment with hot H25 gas), and a filling or covering of a semiconductor " transparent "p-type based CuSCN. A spectral response (quantum efficiency) has been demonstrated, but the photovoltaic efficiency was very low, resulting from a very short photocurrent short circuit (less than 10-3 or 10-4 mA / cm2).

  A similar type of interpenetrating configuration device has been described [Lévy-Clément et al., 205th Electrochemical Society Meeting, San Antonio, USA (2004) Abstract 402; Adv Mater. 17, 1512 (2005)] made with ZnO / CdSe / CuSCN materials, whose n-type porous semiconductor is ZnO in the form of prism-shaped crystals, more than 100 nm in width and 1 to 2 micrometers length. They claimed a photovoltaic yield of about 2% at an irradiation of 360 W / m2, and attributed this improvement to having used a porous layer with a very open (and very wide) pore size. However, this structure has the problem of providing a low roughness factor, less than 10 (the roughness factor being the ratio between the inner surface of the porous layer and the projected surface), which limits the maximum value of the photocurrent that we can hope to obtain since this is related to the amount of absorber in the layer and, consequently, limits the maximum value expected for the photovoltaic efficiency. The Applicant has proposed in the patent application FR 2,881,881 interpenetrated three-component photovoltaic devices all solid and inorganic comprising (a) an n-type compound, (b) an absorber compound, and (c) a semi compound p-type conductor, with a porous layer of semiconductor compound consisting of particles whose particles are coated with a layer of said absorber compound consisting of a compound of Bi sulphide, Sn sulphide, Mo sulphide, Cu sulphide, Although the sulfide compounds of the above mentioned patent have a relatively high absorption threshold, the spectral response thresholds of the best photovoltaic devices (those of Bi sulfide) have been found to be very low, i.e. of the order of 550 nm. Absorption threshold (in English onset) is the value of the wavelength below which the material absorbs the radiation significantly. Other photovoltaic devices with interpenetrating configuration with three components, all solid and inorganic, have been described by the applicant in the French patent application No. 06/02791 filed March 31, 2006 unpublished. These devices made with an absorber layer based on antimony sulphide (Sb), a n-based semiconductor TiO2 and a p-based semiconductor CuSCN had an improved spectral response with a spectral response threshold at 750 nm and a higher photovoltaic efficiency 3.4% However, an absorber layer with an absorption threshold at 750 nm always limits the theoretical maximum photocurrent to -24 mA / cm 2, which can drop to 16-20 in practice. mA / cm 2, because of radiation absorption losses near the threshold and losses due to absorption by a conductive glass substrate combined with the semiconductor layers mentioned above as front and / or back contact. An absorption of solar radiation in a wider solar spectrum, especially with a spectral response threshold greater than 800 nm and less than 1200 nm, would give these photovoltaic devices higher theoretical yields, and a theoretical photocurrent expected higher. Absorption above 1200 nm is disadvantageous because the value of the open-circuit voltage, Voc, which enters into the calculation of the photovoltaic efficiency (see below) decreases by increasing the wavelength of the threshold. Indeed, the photovoltaic efficiency results from a compromise between the photocurrent (which increases with the absorption threshold) and the open circuit voltage (which decreases with this threshold), which gives a theoretical maximum yield for thresholds between 800 and 1200 nm. Finally, photovoltaic devices based on antimony sulphide and silver or based on antimony sulphide and copper of another type have been described, but their photovoltaic performance was very low.

  Thus, a photovoltaic device based on silver sulphide and antimony has been published [Ch. Laubis, PhD Thesis, 2001, Berlin], consisting of a stacked flat layer configuration of the Ag-Sb-S absorber type (2 to 3 microns thick) / CdS. Photovoltaic efficiency was very low, of the order of 0.1%, with photocurrents of 0.5 - 1.7 mA / cm2. In the publication [Nûnez Rodriguez et al., Semicond. Sci. Technol. 20 (2005), 576] the compound AgSbS2 has been mentioned as a possible candidate for photovoltaic applications, but the authors have limited themselves to making thin flat layers without arriving at the manufacture of any photovoltaic device.

  Similarly, a photovoltaic device based on copper sulphide and antimony has been published [Rodriguez-Lazcano et al., J. Electrochem. Soc. 152 (2005) G635] consisting of a stacked plane layer configuration of the absorber CuSbS2 type (100 to 300 nm thick) / Sb2S3 / CdS. Photovoltaic efficiency is very low, less than 0.05%, with photocurrents less than 0.2 mA / cm 2 and phototensions less than 0.35 V. Another photovoltaic device based on copper sulphide and antimony has been published [Manolache et al. Proceedings of the 20th European Photovoltaic Solar Energy Conference (2005), Barcelona, Spain, p. 1906], consisting of a two-component interpenetrated configuration different from the three-component one mentioned above, manufactured with a porous layer based on TiO 2 nanocrystals whose pores are filled with the absorber compound CuSbS 2, which has the dual role of absorber and positive charge carrier, due to its p-type semiconductor nature and the fact of filling the pores sufficiently to allow this transport. The difference with the interpenetrated three-component configuration mentioned above is that of being based on two main components (instead of three), the transparent p-semiconductor being absent. This device does not show significant photovoltaic performance, and only a diode behavior (black current). In total, the interpenetrated configuration photovoltaic devices based on three solid inorganic components (two transparent semiconductors, n and p, and an absorber) have the advantage of being of low manufacturing cost, while presenting a satisfactory photovoltaic efficiency, see higher in some embodiments. Another advantage of all-solid 3-component photovoltaic devices is the expectation of a longer service life for outdoor applications, and more particularly for the application for house roofs, compared to other devices. known interpenetrated configuration, which use liquid and / or organic materials, which have short lifetimes, caused by loss of liquid sealing and / or problems of photodegradation of organic compounds. The object of the present invention is therefore to provide novel interpenetrated 3-component solid and inorganic photovoltaic devices as described above of the type having a porous semiconductor layer covered with an absorption layer which have improved photovoltaic properties and performances, with short-circuit photocurrents greater than 1 mA / cm 2, preferably greater than 3 mA / cm 2, under irradiation of 1000 W / m2 and photovoltaic conversion efficiencies greater than 0.4%, and in particular with extended spectral response in the near-infrared, in particular with a spectral response threshold greater than 800 nm and less than 1200 nm.

  Another object of the present invention is to provide novel photovoltaic devices that use inexpensive materials, sufficiently available in natural resources, stable and not inducing any serious problem of toxicity or negative impact on the environment. To this end, the present invention provides an all-solid photovoltaic device of the type comprising 3 inorganic and solid materials comprising respectively: an n-type and transparent semiconductor compound (1), a p-type and transparent semiconductor compound ( 3), and - a composition of at least one absorber compound arranged as a continuous absorber layer (2) between said n-type semiconductor compounds and which are not in contact, and - said semiconductor compound of n or p type being in the form of a porous substrate (1), with pores (11), in particular of size from 10 to 100 nm, whose inner surface of said pores is entirely coated with a continuous thin layer of absorption (2) of said absorber compound, and of which said pores are then filled in a volume proportion of at least 10%, preferably more than 15%, by a filling layer (3) consisting of the other said semi solid compound -conductor type p or res n, characterized in that said absorber layer comprises at least one antimony and silver sulfide compound or antimony and copper sulfide compound. Transparency means that said compound does not absorb significantly, that is to say with an absorption percentage of less than 35%, the radiation at wavelengths between 400 and 1200 nm and preferably lower at 10% between 450 and 1200 nm. Absorber is understood to absorb with an absorption percentage of at least 50%, radiation in a wavelength range from 400 nm to a wavelength of 600 to 1200 nm. By at least one compound based on antimony sulphide and silver or antimony sulphide and copper, said compound (s) have a composition predominantly of said sulphide compound. More particularly, said compound has a composition of the MXSbYSZ type, with M = Ag or Cu, in which the ratio x / y is from 0.1 to 1.2 and the ratio (x + y) / z is from 0.7 to 3. It is understood that x, y and z are real numbers, and not necessarily integers, which would represent only a defined stoichiometric composition. This compound (s) may, however, contain a certain percentage of the oxide and / or hydroxide ions in place of the sulphides. This is why, more particularly, the rate (M + Sb) / S, that is to say (x + y) / z, is 0.70 to 3.0. This absorption layer may or may not be crystalline or the crystalline domains may not be large enough to be detectable by the X-ray diffraction (XRD) technique. Different absorption layers were obtained with varying composition of antimony and silver sulfide. The composition of the layers may be given by the general formula AgxSbYSZ in that it may contain a mixture of several phases, in particular, one or more Ag-Sb-S ternary phases, more particularly AgSbS2, and other phases. binaries such as Sb2S3. The Ag / Sb ratio, that is to say the ratio x / y when M = Ag, can thus vary by a value of approximately 0.1 (little AgSbS2 with respect to Sb2S3), up to Ag / Sb = 1 (pure composition of AgSbS2). However, the presence of a small amount of other Ag-Sb-S ternary phases, such as Ag3SbS3 or Ag5SbS5, or the Ag2S binary, can cause the Ag / Sb level to rise slightly above 1. And, because of the possible presence of O ions instead of S, in particular, the presence of Sb2O3, the rate (Ag + Sb) / S can vary by a value slightly greater than 0.67 (zero Sb2O3, many of Sb2S3 and little AgSbS2) to high levels (if there is a large excess of Sb2O3), but less than 3.

  The layers which have too much silver (typically with an Ag / Sb ratio higher than 1.2), in particular those which contain the Ag2S compound or the Ag3SbS3 compound in large quantities (detected by X-ray diffraction, XRD) have a spectrum of too much absorption (wavelength threshold> 1200 nm, spectrum similar to that corresponding to the Ag2S absorber or too narrow (threshold of around 550 nm, which seems to correspond to the Ag3SbS3 compound), respectively.) Photovoltaic devices made with these layers do not have a significant photovoltaic response.The layers that have little silver (Ag / Sb ratio <0.1), which contain the compounds AgSbS2 and Sb2S3 (according to the X-ray diffraction), show a spectrum of satisfactory absorption, with an absorption threshold of 880 nm.In addition, the photovoltaic devices made with these layers have a significant photovoltaic response.However, the spectral response has presented by the moment a threshold of 750 nm, similar to that of the antimony sulphide absorber alone, and therefore less than a desired value which should be equal to the absorption threshold of 880 nm. Absorption layers that have a silver content with an Ag / Sb level between 0.2 and 1.0, contain the AgSbS2 compound and may contain Sb2S3 (based on X-ray diffraction). These layers have an absorption spectrum that has a higher absorption threshold than that of the antimony sulfide absorber alone, namely at a wavelength of about 880 nm. And, the photovoltaic devices according to the invention made with these layers have a significant photovoltaic efficiency, with spectral response thresholds of about 880 nm, similar to the values of the absorption threshold. These absorption layers may also contain Sb (Sb 2 O 3) oxide and even small amounts of other different Ag-Sb-S ternary phases of the AgSbS2 compound, such as Ag3SbS3 or Ag5SbS5, or the Ag2S binary. Other compounds could be present in the amorphous state or with crystalline domains too small to be identifiable by X-ray diffraction, DRX (no large enough DRX signal). As a result, only the atomic ratios measured by EDX are available to give a more accurate layer composition.

  More particularly, a device according to the invention comprises a composition of antimony sulphide compound (s) and silver of the AgXSbYSZ type, in which the Ag / Sb content (that is to say, x / y). is greater than 0.1 and less than 1.2. Preferably, the composition of antimony and silver sulfide compound (s) is of the AgXSbYSZ type, in which the Ag / Sb ratio is greater than or equal to 0.2 and less than or equal to 1. More particularly, the composition of Antimony and silver sulfide compound (s) comprises AgSbS2 and optionally Sb2S3. Various absorption layers with varying composition of antimony and copper sulfide have also been obtained with Cu / Sb ratios. variables between about 0.1 and less than 1.2. The composition of the layers may be given by the general formula CuxSbYSZ, in that it may contain a mixture of several phases, in particular, one or more Cu-Sb-S ternary phases, more particularly CuSbS2, and others binary phases such as Sb2S3.

  The photovoltaic devices according to the invention which have layers which contain the compound CuSbS2, and optionally Sb2S3 (based on X-ray diffraction) have a spectrum of spectral response (quantum yield threshold) at a wavelength of about 830 nm greater than that of the antimony sulfide absorber layer alone (ù750 nm), larger than the Cu (I) sulfide absorber layer alone (ù650 nm) , and below 1200nm (unlike layers based on Cu (II) sulphide alone). Photovoltaic devices made with these layers also have a significant photovoltaic efficiency. The absorber layer may also contain Sb oxide (Sb2O3) and even small amounts of different Cu-Sb-S ternary phases of the CuSbS2 compound, such as Cu3SbS3, Cu3SbS3, or Cu12Sb4S13, or binary phases such as CuS. The rate (Cu + Sb) / S remains in general less than 3.0. Preferably, in the device according to the invention, the composition of antimony sulphide compound (s) and copper is of the CuxSbYSZ type, in which the Cu / Sb ratio is greater than 0.1 and less than 1.2.

  More preferably, the composition of the antimony and copper sulphide compound (s) is of the CuxSbYSZ type, in which the Cu / Sb ratio is greater than or equal to 0.2 and less than or equal to 1.

  More particularly, the composition of antimony and copper sulfide compound (s) comprises CuSbS2 compounds and optionally. In the various embodiments with compounds (s), when the absorption layer comprises antimony compound (s) based on antimony sulphide and silver or compound (s) based on sulfide d antimony and copper, the absorption layer may further comprise the compound Sb2O3. This compound results from the substitution of a part of the sulphide ions by oxide and / or hydroxide ions in the general composition of the compound (s). According to other advantageous characteristics: said n-type transparent semiconductor material is a metal oxide such as TiO 2, ZnO and SnO 2, more preferably TiO 2; and said transparent p-type semiconductor material is chosen from Cu (I), preferably CuSCN, CuI or CuAlO 2 based materials, and metal oxides, preferably a nickel oxide.

  Preferably, according to other particular characteristics of the device: the roughness factor of said porous substrate before deposition of the absorber layer is greater than 50, preferably greater than 100. Here, the term "roughness factor" refers to the ratio between the surface area internal of the porous layer (which follows the inner surface of the pores of which it is composed) and the projected area of the layer on its substrate. This porosity may be due to the fact that the layer consists of particles or crystals whose aggregation has created pores in the interstitial space between the particles or crystals during its manufacture and / or the creation of pores directly in a material monolithic with no grain boundary. The roughness thus defined can be estimated approximately (for pores greater than 2 nm and pores approaching the spherical shape) by measuring the porosity (and the pore size distribution) by the adsorption / desorption technique. nitrogen followed by calculation of BET active area; the porous substrate consists of particles or crystals. It is understood that the inner surface of the pores is constituted by the surface of said particles or crystals; the average size of the particles or crystals is from 30 to 50 nm, the porosity of the porous layer before deposition of the absorption layer is from 30 to 70%, more preferably from 40 to 60%; the average pore size is from 20 to 50 nm, the pore size here being the pore opening size, that is to say the average dimension of their cross section and not their length or depth in the thickness of the layer.

  More particularly, said porous substrate consists of a porous n-type metal oxide semiconductor layer and said filler layer is composed of a p-type semiconductor compound. Preferably, said n-type semiconductor metal oxide is TiO 2 having an average particle size of 30 to 50 nm. For TiO 2, these particle sizes correspond to a BET surface of said porous substrate before absorber deposition, greater than 25 m 2 / g, more preferably 25 to 50 m 2 / g. More preferably, said filler layer also constitutes an overlayer layer at least 10 nm thick over one of the surfaces of the porous substrate, especially over the surface of the porous substrate which is not in contact with the substrate. contact with a conductive substrate where appropriate. More preferably, said filler layer consists of CuSCN or nickel oxide. More particularly, in the device according to the invention comprises: the said porous substrate constitutes a layer with a thickness greater than 1 μm, preferably from 2 to 10 μm, and the said absorption layer has a thickness of 1 to 25 nm, preferably from 1 to 10 nm, and said filler layer fills at least 10% of the pore volume, preferably more than 15%, and preferably said filler layer also constitutes a layer of at least 10 nm thick overlap between said porous substrate, and the conductive layer of said second back contact conductive substrate, and - said porous substrate and its absorption and filler layers are confined between two front contact conductive substrates and back contact. Said first front contact substrate being conductive and transparent, the second rear contact conductive substrate may be transparent or not, in particular made of metal or coal.

  Deposition of the porous nanocrystalline layer of metal oxide semiconductor such as TiO2 can be achieved using low cost manufacturing techniques such as that known from doctor blading (or tape casting), using colloidal dispersions of said oxide. metal such as TiO2, or such as screen printing with a paste based on said TiO2 metal oxide. Deposition of the absorber layer can be achieved using a low cost manufacturing technique such as chemical bath deposition or electro-deposition. The filling of the pore free volume with the second transparent semiconductor can be done using a low cost manufacturing technique such as impregnation with a solution of the dissolved material followed by evaporation of the solvent, or by a technique of impregnation of a precursor solution followed by the spin coating method, or by electro-deposition. In addition to this interpenetrated configuration, the three components must satisfy certain conditions, including compatibility of the respective energy levels.

  Taking the semiconductor band model as a first approximation, the absorber conduction band must be less negative than the n-type semiconductor conduction band (in the most commonly used negative values that start at zero, the vacuum level), and the valence band of the absorber must be more negative than the valence band of the p-type semiconductor, all this to allow the injection of the electrons and holes, respectively. Advantageously, the device according to the invention further comprises a transparent nonporous thin layer of a semiconductor component which is interposed between said porous layer of this semiconductor and the transparent conductive substrate layer known as forward contact. This thin layer (called the barrier layer) made of the same material of the same type or another (n or p respectively) as the material of the porous layer makes it possible to avoid any short circuit between the semiconductor filling the pores and the transparent conductive substrate said front contact. This conductive substrate or front contact may consist of a transparent conductive commercial glass.

  Likewise, advantageously, the device according to the invention comprises a layer of thickness greater than 10 nm of a said semiconductor component which is interposed between the said porous semiconductor layer and the said layer of the conductive substrate referred to as the back contact. . This thin overcoat of this semiconductor or of another of the same type as the semiconductor filling the pores avoids any short circuit between the semiconductor of the porous layer and the conductive substrate said back contact. This conductive substrate or back contact may be made of coal or other conductive material such as metal.

  The present invention also provides a porous substrate useful in a device according to the invention consisting of a said n-type or p-type transparent semiconductor compound, preferably an n-type semiconductor metal oxide, the internal surface of which is pores is coated with a so-called absorption layer. More particularly, the porous substrate according to the invention has absorption properties with an absorption maximum of greater than 70%, for wavelengths longer than 400 nm and with an absorption threshold of at least 800 nm, and a roughness factor greater than 50, preferably greater than 100. This type of porous substrate may be useful in other charge separation devices induced by irradiation of photons on an absorber layer, but with a filling with a fluid (liquid or gaseous), which can result in a net transformation of compounds present in the fluid by radiation-induced oxidation-reduction reaction, such as in photo-catalytic devices (chemical reaction) or in photo-electrolytic devices (electrochemical reaction). An advantage of the photovoltaic device according to the invention is that it can achieve a high photovoltaic efficiency while maintaining a low manufacturing cost. This is due to the use of an interpenetrated configuration, which allows the use of materials not requiring high purity, unlike silicon-based devices, and because it can be manufactured in using less expensive techniques than those used for silicon-based devices.

  Another advantage of the photovoltaic device according to the invention compared with other prior devices with the same interpenetrated configuration but with other absorber layers is to have a significantly higher spectral response and photovoltaic efficiency. In particular, a photovoltaic device according to the invention has a spectral response threshold greater than 800 nm, which makes it possible to expect a significantly higher theoretical efficiency compared to the previous photovoltaic device made with the same interpenetrated configuration but with a layer of absorber based on bismuth sulphide or antimony.

  Another advantage of the photovoltaic device according to the invention due to the fact of using inorganic and solid components is to contribute to confer a longer life for applications outside, and more particularly for the application for roofs houses, compared to other known interpenetrating configuration devices, which use liquid and / or organic materials, and which have shown short lifetimes, caused by liquid losses by sealing and / or photo-degradation of organic compounds. Another advantage of the photovoltaic device according to the invention is that the three components do not contain toxic elements such as cadmium, selenium, etc., or element with reduced availability such as indium (In), unlike the thin-film photovoltaic devices of the CdTe / CdS and CuIn (Ga) Sei / CdS type. Finally, the porous layers according to the invention, made based on the semiconductor layer, in particular the nanocrystalline TiO 2, covered by the absorber layer, have good radiation absorption properties and keep a high porosity. . This is an advantage for applications where both of these properties are required concomitantly, such as photovoltaic applications. But other applications are possible such as optical applications, photo-catalytic or photoelectrolytic generation, such as the production of hydrogen from water. The absorber layers according to the invention have a high absorption, with an absorption maximum greater than 70%, in a wide range of wavelengths above 400 nm and an absorption threshold greater than 800 nm. , while keeping a high porosity of the final layer (no blocking or confining of pores). Other features and advantages of the present invention will become apparent in the light of the following detailed description of several embodiments with reference to FIGS. 1 to 4 in which: FIG. 1 represents a diagram of a completely solid photovoltaic device with three main components with an interpenetrating nanocrystalline configuration according to the invention; FIG. 2 represents the absorption spectra of the various layers of absorbers deposited in a porous nanocrystalline TiO 2 layer approximately 3 microns thick on a conducting glass substrate (the absorption corresponding to the conducting glass having been removed); (a) an antimony and silver sulfide absorber layer according to example 1a, (b) an antimony and copper sulfide absorber layer according to example 2a, (c) an antimony sulfide absorber layer according to Comparative Example 3; FIG. 3 represents the spectral response (or quantum yield curves (%) as a function of the wavelength) of the different photovoltaic devices manufactured with different absorber layers and with a covering of a transparent semiconductor of the type based on CuSCN: (a) an absorber layer based on antimony sulphide and silver according to example la, (b) an absorber layer based on antimony sulphide and copper according to Example 2a, (c) an antimony sulfide absorber layer according to Comparative Example 3; d) a bismuth sulfide absorber layer according to Comparative Example 4; FIG. 4 represents the current-voltage curves under an irradiation of 1000 W / m2 (according to the standard AM1.5G solar spectrum standard) corresponding to different photovoltaic devices manufactured with different layers of absorber and with a covering of one p-type transparent semiconductor based on CuSCN: (a) an antimony and silver sulfide absorber layer according to example 1a, (b) a sulfide absorber layer of antimony and copper according to Example 2a; c) a bismuth sulfide absorber layer according to Comparative Example 4. Examples 1 to 7: Absorber layers.

  The absorber layers covering the porous TiO 2 layer of Examples 1 to 7 below have been obtained by the general technique of deposition by chemical bath, known from the state of the art in particular for the manufacture of plane layers on flat substrates. The methods described for antimony sulphide flat layer deposits such as those described in Nair et al. J. Electrochem. Soc. 145, 2113 (1998) have been modified to succeed in obtaining an absorber layer inside the porous layer so that the absorber uniformly covers the inner surface of this porous layer of TiO2, making a recovery of thin and without closing or blocking the pores. More specifically, the operating method comprises the essential steps consisting in applying the general technique of deposition by chemical bath, by immersion of the substrate in one or more solutions containing precursors of antimony, silver or copper, and sulphide, maintained at a given temperature and time.

  The coverage of the absorber layer is approximately homogeneous throughout the porous TiO 2 layer described above for thicknesses of TiO 2 layers less than 4 microns. The absorber distribution was estimated by measuring the ratios of the elements present in the layer cut at different depths using the EDX (Energy-Dissipative X-ray) atomic analysis technique, coupled to an electron microscope scanning SEM (Scanning Electron Microscope). The total amount of absorber in the absorber layers (estimated by EDX measurement of the atomic ratios (Ag + Sb) / Ti or (Cu + Sb) / Ti) is of the order of 2 to 30%, more particularly of about 10 to 15%. The thicknesses of the absorber layers (estimated by transmission electron microscopy TEM measurements, in English transmission electron microscope) are of the order of 1 to 5 nm on average. The composition of the two types of absorber layers of Examples 1 and 2 is AgXSbYSZ or CuXSbYSZ respectively. This composition was estimated with the EDX technique by measuring the proportions of the different elements. However, this estimate is approximate because of the low absorber density in the porous layer and the resolution of the EDX technique. The layers may contain a certain percentage of oxide ions and / or hydroxide in partial substitution of sulphide ions. Depending on the degree of crystallization and the size of the crystalline domains, crystalline phases can be detected by the X-ray diffraction technique, XRD. The layers may contain crystalline phases corresponding to the ternary compounds AgSbS2 or CuSbS2 respectively. They may also contain crystalline phases corresponding to antimony sulphide, Sb2S3, and antimony oxide, Sb2O3. Other ternary or binary crystalline phases could also be present but in lesser amounts. As a result, the composition of these absorbers is generally given as antimony and silver sulfide or as antimony and copper sulfide, respectively, with an Ag / Sb or Cu / Sb ratio between 0.1 and 1.2. EXAMPLE 1 Preparation of an Absorbent Antimony (Sb) and Silver (Ag) Absorbent Layer, AgXSbYSZ with an Ag / Sb Ratio of 0.3, in a Nanocrystalline TiO2 Porous Layer (Medium Size) particle size of 40 nm.

  Transparent conductive commercial glasses (made with a layer of F-doped SnO2) were cut into 2.5 by 2.5 cm pieces, and cleaned with ethanol and distilled water. A compact (non-porous) layer of TiO 2 about 50 nm thick was deposited on the anterior substrate by the pyrolytic spraying method: A solution of titanium (IV) bis (acetylacetonato) di- 10% v / v isopropoxide in ethanol on the surface of the anterior glass substrate heated to about 450 ° C, as described in Kavan et al. Electrochim. Acta 40, 643 (1995). This layer is necessary during photovoltaic applications with interpenetrating configuration since it acts as a barrier against the transport of positive charges towards the electron collector side. Following the deposition of the compact layer, a porous and nanocrystalline layer of TiO2 was deposited by doctor blading (or tape casting), using a colloidal dispersion of TiO 2 prepared as described in patent EP1271580, with an average particle size of about 40 nm. The layer was heated and sintered at 450 ° C. for 30 minutes, which cleans it of organic impurities and ensures good contact of the nanocrystals with each other and with the substrate. The thickness of the porous TiO2 layer is about 3 microns. The absorber layer of antimony (Sb) sulfide and silver (Ag) was deposited using a technique using precursor solutions. First, in a first step, a layer of antimony sulphide (Sb) was deposited by chemical bath deposition. A single vessel containing 35 mL of a precursor solution was used. The bath consisted of a solution in acetone / water (20:80 by volume) containing SbCl3 at a concentration of 0.025M and sodium thiosulfate (Na25203) at a concentration of 0.25M. The sample was immersed in this bath, and the container which contains it was placed in a refrigerator at 10 C and left there for 2 hours. Then, the sample was taken out, rinsed with distilled water, and finally dried with dry nitrogen. The layer was subjected to thermal annealing treatment by heating it at 300 C under a nitrogen flow for 30 minutes. Then, in a second step, the layer with the Sb sulphide deposit was immersed in 0.05 M aqueous silver nitrate solution (AgNO 3) at room temperature for 10 minutes. Then, the sample was taken out, rinsed with distilled water, and finally dried with dry nitrogen. The layer was subjected to thermal annealing treatment by heating it at 400 ° C. under a stream of nitrogen for 30 minutes. The layer has a dark brown color. The absorber layer is, therefore, finished and ready for characterization. The Ag / Sb ratio measured by EDX is 0.3. The ratio (Ag + Sb) / S was 2.0. X-ray diffractograms showed the presence of the crystalline ternary phase of silver sulphide and antimony AgSbS2, as well as the crystalline phase of antimony sulphide Sb2S3.

  Other examples of the absorber layers of Ag-Sb-S compound (s) have been obtained with other Ag / Sb ratios. Example lb. The absorber layer was made by following the first step of Example la above, for the deposition of the antimony sulphide, but without the intermediate annealing, followed by another chemical bathing step for the sulphide. silver, like that of Comparative Example 6 below, but with an immersion time of 25 minutes. Then, the sample was taken out, rinsed with distilled water, and finally dried with dry nitrogen. The layer was subjected to thermal annealing treatment by heating it at 300 ° C. under a stream of nitrogen for 20 minutes. The absorption layer is dark brown in color and has a very broad absorption spectrum, with a threshold greater than 1200 nm, very similar to that of the silver sulfide absorber alone, Ag2S, from Comparative Example 6 Example 1c. The absorber layer was carried out as in Example 1b but with an immersion time of 15 minutes in the second step. Then, the sample was taken out, rinsed with distilled water, and finally dried with dry nitrogen. The layer was subjected to thermal annealing treatment by heating it at 300 ° C. under a stream of nitrogen for 20 minutes. The absorption layer, of yellow color, has an absorption spectrum with a threshold at 550 nm, a value similar to that published by [Ch. Laubis, PhD Thesis, 2001, Berlin] and corresponding to a flat layer of the Ag3SbS3 component.

  Example 1d. The absorber layer was made following the method of Example 1a above, with the first step but without the intermediate annealing. The second step was carried out in a manner similar to that of Example 1 but immersing the sample in a solution of AgNO3 at a concentration of 0.02 mM for 5 minutes. . Then, the sample was taken out, rinsed with distilled water, and finally dried with dry nitrogen. The layer was subjected to thermal annealing treatment by heating it at 300 ° C. under a stream of nitrogen for 20 minutes. The absorption layer has a dark brown color and has an absorption spectrum with a threshold at 880 nm, similar to that of Example 1. The Ag / Sb ratio as measured by EDX was 0.06. Example 2a: Manufacture of an Absorbent Antimony (Sb) and Copper (Cu) Absorber Layer, with a Cu / Sb Ratio of 0.2, in a Nanocrystalline TiO2 Porous Layer (Medium Particle Size) 40 nm). The substrate and the porous nanocrystalline layer of TiO 2 are the same as in Example 1.

  The manufacture of the absorber layer based on antimony sulphide and copper was carried out in two stages consisting of two deposits by chemical bath. In a first step, an antimony sulphide layer was deposited by following the same procedure as in example la until step 1. This procedure was repeated a second time, using a new solution of precursors of antimony sulphide, but without annealing at 300 C. Thereafter, it was continued with a second deposition step by chemical bath to deposit a layer of copper sulphide. The bath consisted of an aqueous solution containing copper sulfate (CuSO4) at a concentration of 0.025M, sodium thiosulfate (Na2S2O3) at a concentration of 0.05M, and thioacetamide (CH3CSNH2) at a concentration of 0.025M The sample was immersed in this bath, and the container containing it was placed in a thermostatic bath at 50 ° C and left there for 10 minutes. Then, the sample was taken out, rinsed with distilled water, and finally dried with dry air. The layer was subjected to thermal annealing treatment by heating it at 400 ° C. under a stream of nitrogen for 30 minutes. The Cu / Sb ratio measured by EDX is 0.2. The ratio (Cu + Sb) / S was -1.3. X-ray diffractograms showed the presence of the crystalline ternary phase of copper sulphide and antimony CuSbS2, as well as the crystalline phases of antimony sulphide Sb2S3 and antimony oxide Sb2O3.

  Other examples of the absorber layers of Cu-Sb-S compound (s) have been obtained with other Ag / Sb ratios. Example 2b. The absorber layer was made following the method of Example 2a above, but depositing with the antimony sulfide precursor solution only once and without the intermediate annealing. The second deposition step with the copper sulphide precursor solution and the final annealing were done as in Example 2a. The absorption layer has a dark brown color and showed an absorption spectrum with a threshold at 830 nm, similar to that of Example 2a. The Cu / Sb ratio measured by EDX was 0.5. The ratio (Cu + Sb) / S was -1.3. X-ray diffractograms showed the presence of the crystalline ternary phase of copper sulphide and antimony CuSbS2, as well as the crystalline phase of antimony oxide Sb2O3. Comparative Example 3: Manufacture of an Antimonium Sulfide Absorbent (Sb) Absorber Layer in a Nanocrystalline TiO2 Porous Layer (Average Particle Size 40 nm)

  The manufacture of the antimony sulfide absorber layer was carried out following the same procedure as in Example 1a to Step 1. The layer was dark brown in color. The absorber layer is, therefore, finished and ready for characterization.

  Comparative Example 4: Production of a bismuth sulfide absorber (Bi) absorber layer in a porous nanocrystalline TiO2 layer (average particle size 40 nm). The substrate and the porous nanocrystalline layer of TiO 2 are the same as in Example 1.

  A layer of bismuth sulphide absorber (Bi) was deposited with a chemical bath deposition technique. A container containing 50 mL of a precursor solution was used. The bath consisted of an aqueous solution containing Bi (NO3) 3 at a concentration of 0.025 M, triethanolamine ((OHCH2CH2) 3N) at a concentration of 0.12 M, and thioacetamide (CH3CSNH2) at a concentration of 0.04 M.

  The sample was immersed in this bath, and the container which contained it was placed in a thermostatic bath at 35 C and left there for 30 seconds. Then, the sample was taken out, rinsed with distilled water, and finally dried with dry air. The layer was subjected to thermal annealing treatment by heating it at 300 C under a nitrogen flow for 30 minutes. The layer has a dark brown color. The absorber layer is, therefore, finished and ready for characterization. Comparative Example 5: Manufacture of a copper sulfide I (Cu) absorber layer in a porous nanocrystalline TiO 2 layer (average particle size 40 nm). The substrate and the porous nanocrystalline layer of TiO 2 are the same as in Example 1. The copper sulphide absorber layer was made by deposition by chemical bath similar to the second stage used for the Example 2. Then, the sample was taken out, rinsed with distilled water, and finally dried with dry air. The layer has turned brown. The absorber layer is, therefore, finished and ready for characterization.

  Comparative Example 6: Production of a silver sulfide absorber (Ag) absorber layer in a porous nanocrystalline TiO 2 layer (average particle size 40 nm). The substrate and the porous nanocrystalline layer of TiO 2 are the same as in Example 1. A silver sulfide absorber layer (Ag) was deposited with a chemical bath deposition technique. A container containing 50 mL of a precursor solution was used. The bath consisted of an aqueous solution containing silver nitrate (AgNO3) at a concentration of 0.02M, and thiourea (NH2CSNH2) at a concentration of 0.02M. The sample was immersed in this bath, and the container which contains it was placed in a thermostatic bath at 60 C and left there for 15 minutes. Then, the sample was taken out, rinsed with distilled water, and finally dried with dry air. The layer has a dark brown color and has a very broad absorption spectrum, with an absorption threshold slightly greater than 1200 nm. Comparative Example 7: Fabrication of a Copper II Absorbent Absorber Layer (Cu ) in a nanocrystalline porous layer of TiO2 (average particle size of 40 nm). The substrate and the porous nanocrystalline layer of TiO 2 are the same as in Example 1.

  The copper sulfide absorber layer II was deposited according to a chemical bath deposition technique, using sequential immersions in different baths. Four containers with 50 ml solutions were used. One bath contains an aqueous solution of copper chloride II (CuCl 2) at a concentration of 0.05 M, another bath contains an aqueous solution of sodium sulphide (Na 2 S) at a concentration of 0.05 M, and the other two containers contain 'distilled water. The substrate with the TiO 2 layer was immersed sequentially and for 10 seconds in each of the previous containers and in the following order: first that with the salt of Cu, then that with water, then that with Na2S, and finally the second container with water. The procedure was repeated 10 times and the sample was finally dried with dry air. The layer is black in color and has a flat absorption spectrum with an absorption threshold well above 1200 nm.

  EXAMPLE 8 Absorption Properties of the Absorption Layers of Examples 1 to 7 The absorption spectra in the visible and near infrared, as shown in FIG. 2, were measured with a Perkin Elmer commercial spectrophotometer. Lambda20 with integration sphere accessory, used to capture all the diffuse radiation. The absorption spectra of the absorber layers of Examples 1 to 3 are shown in FIG. 2. They are similar for the 2 layers of antimony and silver sulfide absorber of Example 1a (FIG. 2a) and antimony and copper sulfide of Example 2a (Figure 2b). The absorber layers of Examples 1 and 2 have a high absorption, with an absorption maximum greater than 70%, in a wide range of wavelengths, with a threshold (or onset in English) of absorption greater than 800 nm, while keeping a high porosity of the final layer (no blocking or confinement of pores). FIG. 2 clearly shows that the absorption spectrum of the absorber layer based on antimony sulphide alone (FIG. 2c) is narrower than that of the layers made with the absorber containing silver sulphide and antimony. of example la, or with the absorber of copper sulphide and antimony of Example 2a.

  The absorption spectrum of the absorber layer of Example 4 (not shown) with bismuth sulfide is similar to that of the antimony and silver sulfide absorber of Example 1a. The absorption spectrum of the absorber layer of Example 5 (not shown) with Cu (I) sulphide exhibits an absorption threshold at a wavelength (ù650 nm) lower than that of the other absorbers of Examples 1a, 2, 3 and 4. Table 1 below gives a summary of the properties of the absorber layers on porous nanocrystalline layers of TiO2 of about 3 microns thick TABLE 1 Absorber No. Example Maximum Threshold Absorption [%] Absorption [nm] Sb and Ag sulphides 1 (a and d) 90 880 Sb and Ag sulphides lb 90 More than 1200nm Sb and Ag sulphides 1c 90 550 Sb and Cu 2 sulphides (a and b) 90 830 Sb sulphate * 3 90 750 Bi sulphide * 4 90 880 Cu sulphate (I) * 5 80 650 Ag sulphide * 6 90 More than 1200 nm Cu (II) sulphide * 7 90 More than 1200 nm ( *): Compared with other layers of absorbers manufactured according to Examples 3 to 7. In addition to photovoltaic applications, these absorber layers can be used for applications optical, photo-catalytic or photo-electrolytic generation, such as the production of hydrogen from water, where radiation absorption and porosity are required properties, among others. Example 9: Photovoltaic devices. 1. Manufacturing In Figure 1 there is shown the general structure of the photovoltaic device comprising the following components assembled according to the following steps. A transparent conductive substrate 4, such as commercial glasses of transparent conductive oxides, is covered with a compact (non-porous) and thin (a few tens of nm) barrier layer made of a semiconductor n such that the TiO2 5. A porous nanocrystalline layer based on a "transparent" semiconductor n 1, such as TiO 2, is deposited on this substrate. It will be noted here that the substrate coated with this porous TiO 2 layer appears white or translucent, although the compound is said to be transparent, the white or translucent color 24 from its light-scattering properties, regardless of its properties. restrictive absorption. A layer of the absorber material 2, with a thickness greater than or equal to 1 nm, is deposited on the surface of the nanocrystals 12 of the anterior porous layer 1. A covering or filling with a transparent semiconductor p 3, such as CuSCN or the nickel oxide is made in such a way as to cover the absorber layer present inside the nanocrystalline layer 1; this covering must be done so as to fill at least about 10% of the pore volume 11, and preferably more than 15% (ideally 100%). This overlap must also leave a thin layer thicker than about 10 nm in the upper part of the layer (the one opposite to the glass substrate). On this layer, another layer 6 is deposited, made of coal or other conductive material such as a metal, which serves as a rear contact to collect the charges on the p side. The final device is confined and sealed 7 to protect it from degradation due to moisture or air pollutants.

  In FIG. 1 the description of the symbols is thus the following: 1: porous and nanocrystalline layer of an n-type transparent semiconductor 11: pores 12: crystals or particles of the n-type transparent semiconductor 2: layer of the material absorber 3: covering and filling the pores with a transparent p-type semiconductor 4: transparent front contact conductive substrate 5: thin and compact barrier layer of an n-type semiconductor 6: conductive layer of a second substrate rear contact conductor 7: seals.

  Photovoltaic devices have been fabricated using the aforementioned porous TiO2 layers, with the absorber layers of this invention and filling the pores with a clear p-semiconductor overlay, more particularly based on CuSCN. . The method used to carry out CuSCN recovery is similar to that known in the art [Kumara et al., Solar Energy Materials Solar Cells 69, 195 (2001); O'Regan et al. Chem. Mater. 14, 5023 (2002)], which is based on the impregnation of the layer with a solution of CuSCN, followed by evaporation of the solvent.

  A solution of CuSCN at a concentration of 15 mg / mL in dipropyl sulfide (S (CH 2 CH 2 CH 3) 2) was prepared, with which the upper portion of the anterior sample placed on a hotplate was impregnated at C. The total volume of CuSCN solution poured was 100 microliters. After the deposition was finished, the sample was heated for additional 3 minutes at the same temperature to ensure that the solvent evaporated completely. Then, a gold layer about 25 nm thick was deposited in the surface of the sample using a metal evaporator (such as the Edwards-306 evaporator). In another alternative (less expensive alternative than gold), a layer of charcoal was painted on the sample using a commercial charcoal based conductive ink. The active area of these devices was 0.54 cm2. 2. Photovoltaic properties. The devices have been characterized with the two most used techniques of the state of the art. A characteristic is the spectral response or quantum efficiency curve (also known as Quantum Efficiency (QE) or Photon-to-Electron Conversion Efficiency Incident (IPCE)) as a function of wavelength, λ, expressed in nanometers, which is the percentage of incident photons (radiation) converted to electrons (electricity): QE (A) = (short-circuit electron flow) / (incident photon flow) = = (short-circuit photo-current) / (power of the incident radiation) x (1240 / A) The ideal spectral response should be the same as the absorption spectrum after correction by the absorption of the inactive components (such as the conducting glass substrate. The spectral response also allows calculate the integrated photocurrent which is the expected short-circuit photocurrent when the device is illuminated under an irradiance of 1000 W / m2 corresponding to a standard standard, in particular to the standard AM1.5G. tégré, IQE, is calculated by making the following integral: IQE = ex Integral (QE (À) x N (À) d À) of which N (À) is the solar spectrum corresponding to the norm AM1.5G (and expressed in [m-2s-1nm-1]), and e is the charge of the electron in coulombs.

  The other characteristic of the performance of the photovoltaic device is the measurement of the photovoltaic efficiency, n, calculated from the measurement of the current-voltage curves (IV) of the photovoltaic device under irradiation, and which is the percentage of the electrical power delivered by the device in the point of maximum power in relation to the power of the incident radiation: n = (electrical power at the point of maximum power) / (power of the incident radiation) This photovoltaic efficiency was measured with a test bench which uses a solar simulator and other equipment; the response of the devices is evaluated under the irradiation of 1000 W / m2 corresponding to the standard standard of the solar spectrum AM1.5G. The bench has been calibrated according to the standard procedure for the measurement of photovoltaic devices which is practiced by various recognized official institutes. Photovoltaic efficiency can also be expressed as a product of three factors, which allows a better interpretation of the performance of the device: n [%] = IscxVocxff (for a standard irradiation of 1000 W / m2) in which: -Isc is the photocurrent short-circuit (i.e., at V = 0), expressed in [mA / cm 2], - Voc is the open-circuit voltage (i.e., I = 0) , expressed in [V], and - ff is the form factor (without units), which expresses how much the curve approximates the ideal rectangular shape (ff = 1). It is calculated by comparing the electrical power at the maximum power point corresponding to the current-voltage curve and the maximum theoretical power that is equal to Isc x Voc. Photovoltaic performance was evaluated as described above by measuring spectral response and current-voltage curves. The spectral response of a photovoltaic device manufactured with the layer of Example 1a has a maximum quantum yield of 30% (curve (a) FIG. 3), a threshold at -880 nm (therefore 130 nm extended to compared to the antimony sulphide absorber), and an integrated IQE photocurrent of 4.8 mA / cm 2. The photovoltaic efficiency is 0.45% (FIG. 4) under an irradiation of 1000 W / m2 according to the AM1.5G standard. The spectral response of a photovoltaic device manufactured with the layer of Example 1d has a threshold at -750nm, therefore, below the absorption threshold and similar to the spectral response threshold of the device based on antimony sulfide alone of Example 3.

  The spectral response of a photovoltaic device manufactured with a layer according to example 2a has a maximum quantum yield of 50% (curve (b) FIG. 3), a threshold at -830 nm (hence 80 nm extended towards the infrared by ratio to the antimony sulphide absorber), and an integrated photocurrent of 6.8 mA / cm 2. The photovoltaic efficiency is 0.8% (FIG. 4) under an irradiation of 1000 W / m2 according to the AM1.5G standard. The spectral response of a photovoltaic device manufactured with a layer according to Example 2b has a maximum quantum yield of 20% (not shown), a threshold at -830 nm (similar to that of Example 2a), and an integrated photocurrent of 1.5 mA / cm 2.

  The spectral response with a layer according to Example 3 has a maximum quantum yield of '80% (curve (c) Figure 3), a threshold at -750 nm, and an integrated photocurrent of 13.5 mA / cm2. The photovoltaic efficiency is 3.4% under an irradiation of 1000 W / m2 according to the AM1.5G standard. The spectral response with a layer according to Example 4 has a maximum quantum yield of '30% (curve (d) Figure 3), a threshold at -600nm, and an integrated photocurrent of 0.7 mA / cm2. The photovoltaic efficiency is 0.3% (FIG. 4) under an irradiation of 1000 W / m2 according to the AM1.5G standard. The threshold of the spectral response, the integrated photocurrent and the photovoltaic efficiency are lower than those of the devices made with the antimony and silver sulphide absorber layers of Example 1a, or of antimony sulphide. and copper of Example 2. The spectral response with a layer according to Example 5 has a low quantum efficiency (less than 2%), a threshold at -550 nm, and a very low integrated photo-current. The photovoltaic efficiency is less than 0.1% under an irradiation of 1000 W / m2 according to the AM1.5G standard.

  The results obtained with the absorber layers of the invention demonstrate that this type of photovoltaic device has a good potential for obtaining a high photovoltaic performance, in particular a broad spectral response, with a threshold between 800 and 900 nm, which would allow a theoretical maximum photocurrent (according to the standard solar spectrum AM1.5G) between 29 and 33 mA / cm2, that is to say, from 20 to 35% higher than the theoretical photocurrent of a similar layer based on sulfide d antimony, which has a threshold at 750 nm (theoretical maximum photocurrent of 24 mA / cm 2). Improvements in the photovoltaic efficiency of these devices are still possible by optimizing the conditions of implementation of the absorber layer, the improvement of the contacts between layers, and a better filling of the semiconductor p, which can increase further the three factors that go into the performance formula. Table 2 summarizes the best photovoltaic performance of devices made with absorber layers on TiO2 nanocrystalline porous layers about 3 microns thick and a pore coating with p-type transparent semiconductor CuSCN TABLE 2 Absorber Maximum Photo Threshold Response Efficiency response spectral current spectral [%] spectral nm] integrated [%] [mA / cm2] Sb and Ag sulphides 880 4.8 0.45 Sb and Cu sulphides 35 830 6.8 0.8 Sb * 80 sulphide 750 13.5 3.4 Cu (I) sulphide * 1 550 <0.1 <0.1 Bi sulphide * 30 600 0.7 0.3 (*): Compared with other absorber layers manufactured according to examples 3 to 5.

Claims (17)

  1. An all-solid photovoltaic device of the type comprising 3 inorganic and solid materials comprising respectively: an n-type and transparent semiconductor compound (1), a p-type and transparent semiconductor compound (3), and a composition of at least one absorber compound arranged as a continuous absorption layer (2) between said n-type semiconductor compounds and which are not in contact, and said n-type or p-type semiconductor compound being in the form of a porous substrate (1), with pores (11) whose inner surface of said pores is entirely coated with a thin continuous absorption layer (2) of said absorber compound, and whose said pores are then filled in a proportion volume of at least 10%, preferably more than 15%, by a filling layer (3) consisting of the other said solid semiconductor solid compound of type p or respectively n, characterized in that the absorption layer includes at least one c antimony and silver sulphide compound or antimony and copper sulphide compound.   2. Device according to claim 1 characterized in that the composition of antimony sulphide compound (s) and silver or antimony sulphide compound (e) is of the type MXSbYSZ, in which the ratio x / y is 0.1 to 1.2 and the ratio (x + y) / z is 0.7 to 3 with M = Ag or Cu.   3. Device according to claim 1 or 2 characterized in that the composition of antimony sulphide compound (s) and silver is of the AgxSbYSzj type in which the Ag / Sb content is from 0.2 to 1.   4. Device according to claim 3 characterized in that the composition of compound (s) antimony sulfide and silver comprises the compound AgSbS2.   5. Device according to claim 1 or 2 characterized in that the composition of antimony sulphide compound (s) and silver is of the CuxSbYSzj type in which the Cu / Sb content is 0.2 to 1.   6. Device according to claim 5 characterized in that the composition of compound (s) antimony sulfide and silver comprises the compound CuSbS2   7. Device according to one of claims 4 or 6 characterized in that the composition of the absorption layer further comprises Sb2S3 compound.   8. Device according to one of claims 1 to 7 characterized in that said n-type transparent semiconductor material is a metal oxide such as TiO2, ZnO and SnO2i preferably TiO2.   9. Device according to one of claims 1 to 8 characterized in that said p-type transparent semiconductor material is selected from materials based on Cu (I), preferably CuSCN, CuAl or CuAlO 2, and metal oxides. , preferably nickel oxide. 10   10. Device according to one of claims 1 to 9 characterized in that the roughness factor of said porous substrate, before deposition of said absorber layer, is greater than 50, preferably greater than 100.   11. Device according to one of claims 1 to 10 characterized in that the porous substrate consists of particles or crystals (12). 15   12. Device according to one of claims 1 to 11 characterized in that the average size of the particles or crystals is 30 to 50 nm.   13. Device according to one of claims 1 to 12 characterized in that said n-type semiconductor compound is in the form of a porous substrate (1) of metal oxide, and said filler layer (3) is consisting of a p-type semiconductor solid compound.   14. Device according to one of claims 1 to 13 characterized in that said filler layer also constitutes an overlayer layer of at least 10 nm thick over one of the surfaces of the porous substrate.   15. Device according to one of claims 1 to 14 characterized in that said filler layer is made of CuSCN or nickel oxide.   16. Device according to one of claims 1 to 15 characterized in that: - said porous substrate (1) is a layer thicker than 1 pm, preferably from 2 to 10 pm, and - said absorption layer (2) has a thickness of 1 to 25 nm, preferably 1 to 10 nm, and said filler layer (3) fills at least 10% of the pore volume, preferably more than 15%, and preferably, said filler layer also constitutes an overlap layer at least 10 nm thick between said porous substrate, and the conductive layer of said second rear contact conductive substrate (6), and - said porous substrate and its layers absorption and filling devices (1,2,3) being confined (7) between two conductive contact substrates (4) and rear contact (6).   17. Porous substrate useful in a device according to one of claims 1 to 11 consisting of a said n-type or p-type semiconductor compound, preferably an n-type metal oxide semiconductor. , whose inner surface of the pores is coated with an absorption layer (2) of said absorber compound.