SILICON AND PEROVSKITE TANDEM SOLAR CELL

Description

[0001] The present invention relates to the photovoltaic field, in particular to the field of tandem cells based on silicon and perovskite. The present invention relates more particularly to a multilayer structure intended to form a tandem photovoltaic cell with two terminals. According to a second aspect, the invention relates to the method of manufacturing such a multilayer structure.

[0002] Solar cells incorporating a perovskite absorber layer have to date exceeded efficiencies of more than 25%. Due to the optically tunable band gap, potentially low-cost manufacturing processes, and high efficiencies, perovskite absorber layers are the ideal partner for tandem integration as the upper sub-cell (i.e., the partial solar cell facing the light in the tandem solar cell) with lower sub-cells made of various materials, including silicon.

[0003] Typically, the interface layers between the two tandem sub-cells consist of deposits of transition metal oxides or conjugated organic polymers. These deposits can be made using either a liquid or dry method. However, these solutions have the following disadvantages: significant optical absorption due to a thickness typically between 30 and 50 nm of the interface layers. This results in parasitic absorption losses (i.e. less current produced by the solar cell). a lack of conformity of the deposits, these deposits being typically made on a lower silicon sub-cell resulting from chemical polishing, or textured on the front face. The deposited layers can be discontinuous, of inhomogeneous quality, which results in inhomogeneous performance or the appearance of short circuits.

[0004] DE 10 2019 116851 B3 describes a multifunctional solar cell comprising a lower silicon sub-cell and an upper perovskite sub-cell between which a stack of at least two layers A and B are deposited. These layers may constitute a tunnel junction and the A layer may be a SAM layer conformally deposited on the B layer.

[0005] One of the aims of the present invention is to overcome at least one of the aforementioned drawbacks. To this end, the present invention provides a multilayer structure intended to form a tandem photovoltaic cell, the multilayer structure comprising: a silicon-based lower sub-cell, a perovskite-based upper sub-cell, a tunnel junction comprising doped silicon disposed at the interface of the lower sub-cell and the upper sub-cell, the upper sub-cell comprising a lower charge transport layer composed of a monolayer of self-assembled molecules.

[0006] The term 'self-assembled monolayer' is well known to those skilled in the art under the acronym 'SAM' for Self-Assembled Monolayer. Conventionally, the SAM is a monolayer of molecules grafted to the surface by a strong bond, such as a covalent bond or electrostatic interactions.

[0007] In the context of the present invention, the monolayer of self-assembled molecules is grafted with the silicon of the tunnel junction by Si-C covalent bonds.

[0008] As will be seen in more detail below, this Si-C covalent bond is formed by a hydrosilylation reaction of a precursor compound of the grafted molecule, with the Si-H bonds available on the silicon surface. The precursor compound comprises an anchoring group formed by a C-C multiple bond, of the alkene or alkyne type, which reacts with the Si-H bonds on the silicon surface to be functionalized, forming Si-C bonds. The reaction is self-limited by the number of accessible reactive Si-H sites. The monolayer of self-assembled molecules thus formed according to the present invention is thus conformal. The monolayer uniformly covers the texture of the Si surface it covers, so that the thickness of said monolayer is the same at all points. It follows the geometry of the surface at the microscopic scale, regardless of the surface roughness, avoiding any interruption that could lead to an electrical short circuit.

[0009] In addition, said monolayer has a reduced thickness, in particular less than or equal to 10 nm. This thickness is at most equivalent to the largest dimension of the grafted molecule. This makes it possible to limit the optical adsorption of the interface and ensures better photovoltaic efficiency.

[0010] The Si-C bonds formed are covalent in nature and are known to be extremely robust from a chemical and thermal point of view.

[0011] Furthermore, said grafted molecule also comprises a functional group linked to the anchoring group by a carbon chain creating a dipole moment. The nature and charge of the functional group make it possible to influence the strength of the dipole moment, as well as the work function of the functionalized silicon surface. The functional group also influences the surface energy of the layer, so as to promote the growth of the perovskite layer of the upper sub-cell.

[0012] In addition, the carbon chain between the functional group and the anchoring group allows the grafted molecules to be oriented, which creates an orientation at the dipole moment and allows for a homogeneous surface from a chemical and surface energy point of view. For example, the existence of intermolecular Van der Waals forces between the carbon chains allows the grafted molecules to be aligned with each other, this "orientation" leading to better charge transfer and better homogeneity of the surfaces obtained.

[0013] According to one possibility, the monolayer of self-assembled molecules is constituted by first fragments of formula I (with R = H, Me, OMe and n = 1, 2 or 3) or of formula II (with R = H, Me, OMe and n = 1, 2 or 3) linked by Si-C covalent bonds to silicon, so as to form a hole transport layer. These first fragments can indeed play the role of hole transport due to the localization / delocalization of electrons which can be extended over the entire conjugated system of the first fragments.

[0014] Alternatively, the monolayer of self-assembled molecules consists of the second fragments of formulae III to VI linked by Si-C covalent bonds to silicon, so as to form an electron transport layer, as described in the review article by K Choi et al. Sol. RRL 2020, 4, 1900251. These second fragments are indeed favorable to the conduction of electrons.

[0015] According to one arrangement, the silicon of the tunnel junction is selected from hydrogenated and doped amorphous silicon, hydrogenated and doped polycrystalline silicon, hydrogenated and doped microcrystalline silicon or hydrogenated and doped nanocrystalline silicon, hydrogenated and doped oxygenated nanocrystalline silicon (nc-SiOx:H) or hydrogenated and doped carbonaceous nanocrystalline silicon (nc-SiCx:H). In the case of hydrogenated and doped oxygenated nanocrystalline silicon "nc-SiOx:H", it is understood in this document that it is a material composed primarily of silicon crystals, and a portion of silica intended to lower its refractive index. This material is therefore very different from the native oxide SiOx.

[0016] Of course, the silicon involved in the hydrosilylation reaction is devoid of native oxide. The Si-H hydrogen bonds present on the surface are made accessible by a deoxidation treatment to allow the reaction with the precursor compound of said monolayer.

[0017] According to one arrangement, the tunnel junction comprises a lower layer and an upper layer, at least one of the upper layer and the lower layer is an n-type layer or a p-type layer. Depending on the NIP or PIN configuration of the photovoltaic cell, the monolayer of self-assembled molecules is selected to form an electron transport layer or to form a hole transport layer, regardless of whether the underlying layer of the tunnel junction is p-type or n-type.

[0018] According to a particular embodiment, the lower layer is an n-type layer or a p-type layer and the monolayer of self-assembled molecules forms at least in part the upper layer of the tunnel junction. In other words, said monolayer forms a charge transport layer at the same time as it acts as the upper layer of the tunnel junction. Thus, the upper layer of the tunnel junction is a hole transport layer consisting of a monolayer of self-assembled molecules containing the first fragments of formula I or formula II when formed on a lower n-type layer. The upper layer is an electron transport layer consisting of a monolayer of self-assembled molecules containing second fragments chosen from formulas III to VI when formed on a lower p-type layer.

[0019] Alternatively, the lower layer of the tunnel junction forms a charge selection layer for the lower silicon sub-cell. A simplified architecture can then be achieved in which the lower layer of the tunnel junction is implemented by a charge selection layer for the lower sub-cell and the upper layer of the tunnel junction is implemented by a charge transport layer for the perovskite sub-cell.

[0020] According to another variant, the upper layer and the lower layer are respectively an n-type layer or a p-type layer or vice versa.

[0021] According to one possibility the multi-layer structure includes: a silicon-based lower sub-cell, a perovskite-based upper sub-cell, a tunnel junction comprising doped silicon disposed at the interface of the lower sub-cell and the upper sub-cell, the upper sub-cell comprising a lower charge transport layer composed of a monolayer of self-assembled molecules, the monolayer of self-assembled molecules is grafted with the doped silicon of the tunnel junction by Si-C covalent bonds.

[0022] According to a second aspect, the present invention provides a method of manufacturing the multilayer structure as previously described, the method comprising the steps of: a) providing the lower silicon-based sub-cell, b) depositing at least the doped and hydrogenated silicon layer to form at least the lower layer or the upper layer of the tunnel junction, c) forming the monolayer of self-assembled molecules so as to form the charge transport layer, d) depositing the perovskite layer on the monolayer of self-assembled molecules.

[0023] A conformal and low-absorption charge transport layer (of light flux) is thus easily formed between the two sub-cells.

[0024] According to one possibility, the lower sub-cell provided in step a) comprises at least a silicon substrate, passivation layers and a charge selection layer of the lower sub-cell.

[0025] According to one arrangement, the method also comprises depositing a second charge transport layer on the perovskite layer so as to finalize the upper sub-cell.

[0026] This step is followed by the deposition of a layer of TCO (transparent conductive oxide) before forming metal contacts on the front and back faces of the tandem photovoltaic cell.

[0027] Concretely, before the formation of the monolayer of self-assembled molecules of step c), the method comprises a step i) of applying a treatment to the silicon surface of the tunnel junction with an aqueous solution of a mixture of 2% HF and 1% HCl so as to remove the native oxide.

[0028] The treatment of step i) advantageously makes it possible to make all the Si-H bonds on the silicon surface accessible in order to carry out hydrosilylation.

[0029] The aqueous solution allows the removal of the native oxide and also contaminants.

[0030] Step i) is carried out for a duration of a few tens of seconds, for example for a duration of approximately 60 seconds.

[0031] Step c) of the method comprises the hydrosilylation of a precursor compound selected from the first precursor compounds of formula P1 or P.II (with R = H, Me, OMe and n = 1, 2 or 3) and the second precursor compounds of formula P.III to P.VI on the hydrogenated silicon surface so as to obtain the monolayer of self-assembled molecules.

[0032] According to one possibility, hydrosilylation is obtained by a radical reaction activated by heat treatment.

[0033] Concretely, at least the hydrogenated silicon surface is placed in the presence of a precursor compound under an inert atmosphere and the whole is heated to a temperature between 130 and 180°C for a period of 12 to 18 hours so as to lead to hydrosilylation.

[0034] According to one arrangement, step b) of the method comprises deposition of a single lower layer of the doped silicon tunnel junction and step c) comprises formation of the monolayer of self-assembled molecules forming a charge transport layer and the upper layer of the tunnel junction.

[0035] According to other features, the multilayer structure of the invention comprises one or more of the following optional features considered alone or in combination: The multilayer structure is intended to form a two-terminal tandem photovoltaic cell. The monolayer of self-assembled molecules has a thickness of 10 nm or less. The monolayer of self-assembled molecules is at the interface of the perovskite absorber layer of the upper sub-cell and the tunnel junction. The monolayer of self-assembled molecules is in direct contact with the perovskite absorber layer of the upper sub-cell. The monolayer of self-assembled molecules is in direct contact with the bottom layer or the top layer of the tunnel junction. The monolayer of self-assembled molecules originates from a hydrosilylation reaction with the silicon-hydrogen bonds on the surface of the underlying hydrogen-doped silicon layer, optionally amorphous, microcrystalline, nanocrystalline, hydrogenated oxygenated nanocrystalline or nc-SiCx:H.The underlying layer constitutes one of the lower layer and the upper layer of the tunnel junction. The self-assembled molecules form silicon-carbon covalent bonds with the underlying silicon atoms. The covalent bonds are formed by hydrosilylation. The doped silicon of the tunnel junction with which the self-assembled molecule monolayer is grafted has a deoxidized surface. The doped silicon of the tunnel junction with which the self-assembled molecule monolayer is grafted is cleaned of its native oxide. The surface of the doped silicon layer at the interface between the lower sub-cell and the upper sub-cell is devoid of native oxide. The first precursor compound and the second precursor compound of the self-assembled molecule mono-layer are defined by the formulas P1 or P.II and P.III to P, respectively.VI corresponding respectively to the molecules of the first fragments of formula I or II and the second fragments of formulas III to VI. The n-type layer of the tunnel junction is a silicon layer comprising between 10 e< 19 at / cm3 and 10 e< 21 at / cm3 of phosphorus. The p-type layer of the tunnel junction is a silicon layer comprising between 10 e< 19 at / cm3 and 10 e< 20 at / cm3 of boron. The monolayer of self-assembled molecules constitutes the entire upper layer of the tunnel junction. The charge transport layer conforms to the morphology of the silicon surface of the tunnel junction. The silicon surface is generally planar and said monolayer is also generally planar.The silicon surface is obtained by CMP polishing (roughness < 10 nm), or by chemical polishing (surface topology of the order of microns high by a few tens of microns wide), or textured (with morphologies of the order of a few µm in height and width). The lower layer of the tunnel junction is formed by PECVD deposition, for example for hydrogenated microcrystalline silicon, on the lower sub-cell of the multilayer structure. The lower layer and / or the upper silicon layer of the tunnel junction each have a thickness between 10 and 100 nm. The perovskite layer is deposited to a thickness between 200 and 1000 nm by wet chemical means (spin coating or printing processes). The perovskite layer is deposited by dry means by vacuum deposition or by mixed means (evaporation of Pbl2 followed by liquid impregnation of the perovskite precursors).The lower silicon-based sub-cell includes passivation layers of undoped hydrogenated amorphous silicon or silicon oxide.

[0036] The term "perovskite" herein refers to the "perovskite structure" and not specifically to the perovskite material CaTiO3. For the purposes of this disclosure, "perovskite" preferably includes and refers to any material that has the same type of crystal structure as calcium titanium oxide and materials in which the divalent cation is replaced by two separate monovalent cations. The perovskite structure generally has an AMX3 type stoichiometry, where "A" and "M" are cations and "X" is an anion.

[0037] The "A" and "M" cations can have different charges, and in the original perovskite mineral (CaTiO3), the A cation is divalent and the M cation is tetravalent. Perovskite formulas encompass structures having one, two, three, or four cations, which may be the same or different, and / or one or two of the anions and / or metal atoms that carry two or three positive charges.

[0038] Furthermore, the organic-inorganic perovskite material has a perovskite structure of one of the following formulas: A1A2A3A4MX3, A1A2A3MX3, A1A2MX4; A1MX3; A1A2N2 / 3X4; A1N2 / 3X3; BN2 / 3X4; A12NMX6; BMX4, where A1, A2, A3, A4 are either monovalent organic cations or mixtures thereof which are independently selected from primary, secondary, tertiary or quaternary organic ammonium compounds, including heterocycles and nitrogenous ring systems. A and A' independently of each other have 1 to 60 carbon atoms and 1 to 20 heteroatoms (such as methylammonium or formamidinium) or inorganic cations (such as Na, K, Rb, Cs).B is a divalent organic cation selected from primary, secondary, tertiary or quaternary organic ammonium compounds having from 1 to 60 carbon atoms and from 2 to 20 heteroatoms and with two positively charged nitrogen atoms; M is a divalent metal cation selected from the group consisting of Cu, Ni, Co2+, Fe2+, Mn2+, Cr2+, Pd2+, Cd2+, Ge2+, Sn2+, Pb2+, Eu2+ or Yb2+; N is a divalent metal cation selected from the group of Bi and Sb and, X is an anion independently selected from Cl -< Br -< , I -< , NCS -< , CN -< and NCO- and a mixture thereof.

[0039] Other characteristics and advantages will appear on reading the detailed description below, of 7 non-limiting examples of implementation, made with reference to the appended figures in which: There figure 1 represents a schematic view of a photovoltaic cell having a PIN configuration, comprising a multilayer structure according to a first embodiment of the invention. The figure 2 represents a schematic view of a photovoltaic cell comprising a multilayer structure according to a second embodiment of the invention. The figure 3 represents a schematic view of a photovoltaic cell comprising a simplified multilayer structure of the multilayer structure of the first embodiment of the invention (third embodiment). The figure 4 represents a schematic view of a photovoltaic cell comprising a simplified multilayer structure of the multilayer structure of the second embodiment of the invention (fourth embodiment). The figure 5 represents a schematic view of a photovoltaic cell comprising a simplified multilayer structure according to a fifth embodiment of the invention. The figure 6 represents a schematic view of a photovoltaic cell comprising a simplified multilayer structure according to a sixth embodiment of the invention. The figure 7 represents a schematic view of a photovoltaic cell according to the first embodiment of the invention and having a NIP configuration.

[0040] The structures illustrated in figures 1 à 7 represent the multilayer structure according to the invention implemented in photovoltaic cells.

[0041] The multilayer structure 100 of the invention comprises a lower sub-cell 10 comprising a silicon substrate 1 of c-Si (n), an upper sub-cell 20 comprising a perovskite absorbing layer 2, a tunnel junction 30, arranged at the interface of the lower sub-cell 10 and the upper sub-cell 20, based on silicon, a charge transport layer 4 on the rear face of the upper sub-cell 20 being composed of a mono-layer of self-assembled molecules grafted onto the silicon of the tunnel junction 30.

[0042] The photovoltaic cells formed from the multilayer structure 100 described above further comprise a full-plate metal contact 5 made of silver on a TCO layer 6 (transparent conductive oxide) for making electrical contact on the rear face of the cell. The multilayer structure 100 also comprises a conventional charge transport layer 7 on the perovskite absorber layer 2 (front side relative to the light flux), a TCO layer 6' and a gate metal contact 5' for making electrical contact on the front face (in the direction of light penetration).

[0043] As illustrated on the figures 1 à 3 , the lower silicon sub-cell 10 comprises on either side of the silicon substrate a passivation layer 8, 8' conventionally made of (i) a-Si:H ( figures 1 , 3 , 5 And 7 ) or in SiOx ( figures 2 And 4). A p, 9p charge selection layer, in (p) a-Si:H ( figure 1 , 3 And 5 et 6 ) or in Poly-Si (p+) ( figure 2 ) covers the backside passivation layer and a charge selection layer n, 9n, respectively in (n) a-Si:H ( figures 1 , 3 And 5 et 6 ) or Poly-Si (n+) ( figures 2 And 4 ) covers the 8' passivation layer on the front face in a PIN type configuration ( figure 1 ). The 9p, 9n charge selection layers are inverted in a PIN-like configuration ( figure 7 ).

[0044] At the interface of the lower silicon sub-cell 10 and the upper perovskite sub-cell 20, the structure 100 according to the present invention comprises a tunnel junction 30 made of silicon and a charge transport layer made by a mono-layer of self-assembled molecules 4 covalently bonded to the underlying silicon.

[0045] As shown in the figure 1 illustrating a PIN type architecture, the tunnel junction 30 consists of a lower n-type, 3n, layer of silicon (n+) µc-Si:H and an upper p-type, 3p, layer of silicon (p+) µc-Si:H. A mono-layer of self-assembled molecules 4 forming a hole transport layer of the upper sub-cell 20 is formed by hydrosilylation of a first precursor compound of formula PI or P.II on the Si-H bonds of the deoxidized surface of the upper p-type, 3p layer.

[0046] According to an embodiment variant illustrated in the figure 2 , the charge selection layer n, 9n, made of Poly-Si:H (n+) of the lower sub-cell 10 also acts as the lower layer 3n of the tunnel junction 30. An upper layer 3p of the tunnel junction 30 is formed of Poly-Si:H (p+). A monolayer of self-assembled molecules 4 of first fragments of formula I or II is formed by hydrosilylation on the Si:H bonds of the upper p-type layer to form a hole transport layer.

[0047] According to yet another embodiment variant illustrated in the figure 3 , the invention this time proposes a simplified architecture of the multilayer structure 100 as previously described in relation to the figure 1 . In this simplified configuration, the hole transport layer 4 is obtained by hydrosilylation of a first precursor compound of formula PI or P.II on the lower n-type 3n layer of the tunnel junction 30. The hole transport layer plays the role of the p-type, 3p, layer of the tunnel junction 30.

[0048] This simplification ( figure 4 ) is applied to the configuration of the figure 2 wherein the lower n-type layer, 3n, of the tunnel junction 30 is obtained by the electron selection layer 9n made of Poly-Si:H (n+) of the lower sub-cell 10. The electron selection layer 9n is covered by the monolayer of self-assembled molecules 4 (hole transport layer). Thus, the two aforementioned layers form respectively the lower layer 3n and the upper layer 3p of the tunnel junction 30.

[0049] This same simplification ( figure 5 ) is also brought to the variant of the figure 3 in which the electron selection layer 9n, in (n) a-Si:H, forms the lower 3n layer of the tunnel junction 30 and the hole transport monolayer 4 plays the role of the upper 3p layer of the tunnel junction 30.

[0050] The simplified 100 multilayer structure illustrated in figure 6 differs from that of the figure 5 , in that the electron selection layer 9n forming the lower n-type layer, 3n, of the tunnel junction is made of (n+) nc-SiOx: H. (According to a variant not illustrated, the lower layer is made of nc-SiCx: H (n+)). This material has a refractive index intermediate between that of the absorbing layer of peroskite 2 and that of the silicon substrate 1, which makes it possible to reduce parasitic reflection.

[0051] The multilayer structures 100 may be constructed to produce a cell of a PIN type configuration ( figure 1 à 6 ) as a PIN type configuration ( figure 7 ) without significant modification of the implementation of the invention except the choice of the type of first fragment of formula I and II or of second fragment of formula III to VI of the grafted molecules and of the precursor compounds used for the hydrosilylation reaction.

[0052] According to the manufacturing method of the present invention, the layer 3,9 of doped hydrogenated silicon (optionally nc-SiOx or nc-SiCx according to embodiment variants) intended to receive the charge transport layer 4 is first cleaned of its native oxide by a treatment according to a step b) which provides for the immersion of said layer in an aqueous solution of 2% HF and 1% HCl at room temperature for approximately 60 seconds. The deoxidized surface having accessible Si-H bonds is rinsed with deionized water, then dried with nitrogen. The plate having the deoxidized layer is immediately placed in a reactor previously purged with nitrogen and containing the first precursor compound of formula PI with R = H before being placed at 180°C for 16 hours so as to obtain the monolayer 4.

[0053] The plate is then rinsed with hexane, dichloromethane and dried under nitrogen.

[0054] The deposition of an absorbent layer of perovskite 2 is then carried out by spin coating or any other conventional method.

[0055] Thus configured, the multilayer structure 100 of the present invention comprises a monolayer of self-assembled molecules 4 of low optical absorption on a silicon layer 3, 9 of the tunnel junction 30. It is also a conformal layer regardless of the roughness of the underlying silicon surface and has a strong coupling achieved by the creation of a robust Si-C covalent bond. In addition, said monolayer 4 allows the achievement of selectivity via the formation and orientation of a dipole moment at the surface of the system and this layer has an adjustable surface energy to promote the wettability and growth of the overlying perovskite layer 2.

Claims

1. A multilayer structure (100) intended to form a tandem photovoltaic cell, the multilayer structure (100) comprising: - a lower silicon-based sub-cell (10), - an upper perovskite-based sub-cell (20), - a tunnel junction (30) including doped silicon disposed at the interface of the lower sub-cell (10) and of the upper sub-cell (20), the upper sub-cell (20) comprising a lower charge transport layer (4) composed of a monolayer of self-assembled molecules, characterized in that the monolayer of self-assembled molecules (4) is grafted with the doped silicon of the tunnel junction (30) by Si-C covalent bonds.

2. The multilayer structure (100) according to claim 1, wherein the silicon of the tunnel junction (30) is selected from hydrogenated and doped amorphous silicon, hydrogenated and doped polycrystalline silicon, hydrogenated and doped microcrystalline silicon, hydrogenated and doped nanocrystalline silicon, hydrogenated and doped oxygenated nanocrystalline silicon, or hydrogenated and doped carbonaceous nanocrystalline silicon.

3. The multilayer structure (100) according to any of claims 1 and 2, wherein the monolayer of self-assembled molecules (4) consists of first fragments of formula I or formula II linked by covalent bonds to the silicon, so as to form a hole transport layer.

4. The multilayer structure (100) according to any of claims 1 and 2, wherein the monolayer of self-assembled molecules (4) consists of second fragments of formula III to VI linked by covalent bonds to the silicon, so as to form an electron transport layer.

5. The multilayer structure (100) according to any of claims 1 to 4, wherein the tunnel junction (30) comprises a lower layer and an upper layer, at least one of the upper layer and of the lower layer is an n-type layer (3n) or a p-type layer (3p).

6. The multilayer structure (100) according to claim 5, wherein the lower layer of the tunnel junction (30) is an n-type layer (3n) or a p-type layer (3p) and the monolayer of self-assembled molecules (4) forms at least part of the upper layer.

7. The multilayer structure (100) according to any of claims 5 and 6, wherein the lower layer of the tunnel junction (30) forms a charge selection layer (9n, 9p) of the lower silicon sub-cell (10).

8. The multilayer structure (100) according to any of claims 1 to 7, wherein the surface of the doped silicon layer at the interface between the lower sub-cell (10) and the upper sub-cell (20) is devoid of native oxide.

9. A method for manufacturing the multilayer structure (100) according to claims 1 to 8, the method comprising the steps of: - a) providing the lower silicon-based sub-cell (10), - b) depositing at least the doped and hydrogenated silicon layer to form at least the lower layer or the upper layer (3n, 3p) of the tunnel junction (30), - c) forming the monolayer of self-assembled molecules (4) so as to form the charge transport layer, - d) depositing the perovskite layer (2) on the monolayer of self-assembled molecules (4).

10. The method for manufacturing the multilayer structure (100) according to claim 9 comprising, before the formation of the monolayer of self-assembled molecules (4) of step c), a step i) of applying a treatment of the silicon surface with an aqueous solution of a mixture of HF at 2% and HCl at 1%, so as to remove the native oxide.