WO2019116031A1

WO2019116031A1 - Multi-junction photovoltaic device

Info

Publication number: WO2019116031A1
Application number: PCT/GB2018/053606
Authority: WO
Inventors: Simon KIRNER; Daniel Kirk
Original assignee: Oxford Photovoltaics Limited
Priority date: 2017-12-15
Filing date: 2018-12-12
Publication date: 2019-06-20
Also published as: GB201721066D0

Abstract

A multi-junction photovoltaic device comprises a first sub-cell (100, 101, 102, 103, 104, 105) stacked on and in electrical contact with a second sub-cell (109, 110, 111, 112, 113), the first sub-cell comprising a layer of a perovskite material (104), the first sub-cell comprising the front of the multijunction photovoltaic device adapted to be exposed to incident light in use. The second sub-cell comprises a rear emitter p-type silicon structure. A tunnel junction (108, 107, 106) is provided between the first and second sub-cells.

Description

MULTI-JUNCTION PHOTOVOLTAIC DEVICE

FIELD OF THE INVENTION

The present invention relates to a perovskite on silicon multi-junction photovoltaic device that comprises a perovskite material that has both a band gap that makes it suitable for use in multi-junction photovoltaic devices and improved stability, and methods of fabricating such a photovoltaic device.

BACKGROUND OF THE INVENTION

Over the past forty years or so there has been an increasing realisation of the need to replace fossil fuels with more sustainable energy sources. The new energy sources should also have low environmental impact, be highly efficient and be easy to use and cost effective to produce. To this end, solar energy is one of the most promising technologies. However, the high cost of manufacturing devices that capture solar energy, including high material costs, has historically hindered its widespread use.

Every solid has its own characteristic energy-band structure which determines a wide range of electrical characteristics. Electrons can transition from one energy band to another, but each transition requires a specific minimum energy and the amount of energy required will be different for different materials. The electrons acquire the energy needed for the transition by absorbing either a phonon (heat) or a photon (light). The term “band gap” for crystalline materials refers to the energy difference range in a solid where no electron states can exist, and generally means the energy difference (in electron volts) between the top of the valence band and the bottom of the conduction band. The efficiency of a material used in a photovoltaic device, such as a solar cell, under normal daylight conditions is a function of the band gap for that material. If the band gap is too high, most daylight photons cannot be absorbed and are lost to transmission or reflection; if it is too low, then most photons have much more energy than necessary to excite electrons across the band gap, and the rest will be lost to thermalisation. The Shockley-Queisser limit refers to the theoretical maximum efficiency that can be obtained with a single junction solar cell. The highest possible efficiency is about 33% and can be obtained with a 1.34eV band gap. The focus of much of the recent work on photovoltaic devices has been the quest for materials which have a band gap as close as possible to this maximum.

One class of photovoltaic materials that has attracted significant interest has been the hybrid organic-inorganic halide perovskites. Materials of this type form an ABX₃ crystal structure which has been found to show a favourable band gap, a high absorption coefficient and long diffusion lengths, making such compounds ideal as an absorber in photovoltaic devices. Early examples of hybrid organic-inorganic metal halide perovskite materials are reported by Kojima, A et al. (2009) Organometal Halide Perovskites as Visible-Light Sensitizers for Photovoltaic Cells. J. Am. Chem. Soc., 131 (17), pp.6050-6051 in which such perovskites were used as the sensitizer in liquid electrolyte based photoelectrochemical cells. Kojima et al. report that a highest obtained solar energy conversion efficiency (or power energy conversion efficiency, PCE) of 3.8%, although in this system the perovskite absorbers decayed rapidly and the cells dropped in performance after only 10 minutes.

Subsequently, Lee, M et al. “Efficient Hybrid Solar Cells Based on Meso-Superstructured Organometal Halide Perovskites”, Science, 338(6107), pp.643-647 (2012) reported a“meso- superstructured solar cell” in which the liquid electrolyte was replaced with a solid-state hole transporting material, (HTM), spiro-MeOTAD. Lee et al. reported a significant increase in the conversion efficiency achieved, whilst also achieving greatly improved cell stability through avoiding the use of a liquid solvent. In the examples described therein, CHsNHsPbL perovskite nanoparticles assumed the role of the sensitizer within the photovoltaic cell, injecting electrons into a mesoscopic T1O2 scaffold and holes into the solid-state HTM. Both the T1O2 and the HTM act as selective contacts through which the charge carriers produced by photoexcitation of the perovskite nanoparticles are extracted.

Further work described in WO2013/171517 disclosed how the use of mixed-anion perovskites as a sensitizer/absorber in photovoltaic devices, instead of single-anion perovskites, results in more stable and highly efficient photovoltaic devices. In particular, this document disclosed that the superior stability of the mixed-anion perovskites is highlighted by the finding that the devices exhibit negligible colour bleaching during the device fabrication process, whilst also exhibiting full sun power conversion efficiency of over 10%. In comparison, equivalent single anion perovskites are relatively unstable, with bleaching occurring quickly when fabricating films from the single halide perovskites in ambient conditions.

More recently, WO2014/045021 described planar heterojunction (PHJ) photovoltaic devices comprising a thin film of a photoactive perovskite absorber disposed between n-type (electron transporting) and p-type (hole transporting) layers. Unexpectedly it was found that good device efficiencies could be obtained by using a compact (i.e. without effective/open porosity) thin film of the photoactive perovskite, as opposed to the requirement of a mesoporous composite, demonstrating that perovskite absorbers can function at high efficiencies in simplified device architectures. Recently the application of perovskites in photovoltaic devices has focussed on the potential of these materials to boost the performance of conventional silicon-based solar cells by combining them with a perovskite-based cell in a tandem/multi-junction arrangement. In this regard, a multi-junction photovoltaic device comprises multiple separate sub-cells (i.e. each with their own photoactive region) that are“stacked” on top of one another such that the photoactive regions overlie one another, and convert more of the solar spectrum into electricity - increasing the overall efficiency of the device. To do this, the photoactive region of each sub cell is selected so that the band gap of the sub-cell ensures that it will efficiently absorbs photons from a specific segment of the solar spectrum. Multi-junction solar cells are a way to circumvent the above-mentioned Shockley-Queisser limit if an appropriate band gap combination is used. The combination of multiple photoactive regions/sub-cells with different band gaps provides that on the one hand, a wide range of incident photons can be absorbed in the lower bandgap sub cells, while on the other hand, each photoactive region/sub-cell will be more effective at extracting energy from the photons within the relevant part of the spectrum in the higher bandgap sub cells. In theory, the lowest band gap of a multi-junction photovoltaic device will be lower than that of a typical single junction device, such that a multi-junction device will be able to absorb photons that possess less energy than those that can be absorbed by a single junction device. Furthermore, for those photons that would be absorbed by both a multi-junction device and a single junction device, the multi-junction device will absorb those photons more efficiently, as having band gaps closer to the photon energy reduces thermalization losses.

In a typical multi-junction device the top photoactive region/sub-cell in the stack has the highest band gap, with the band gap of the lower photoactive regions/sub-cells reducing towards the bottom of the device. This arrangement maximizes photon energy extraction as the top photoactive region/sub-cell absorbs the highest energy photons first whilst allowing the transmission of photons with less energy. Each subsequent photoactive region/sub-cell then extracts energy from photons closest to its band gap thereby minimizing thermalization losses. The bottom photoactive region/sub-cell then absorbs all remaining photons with energy above its band gap. When designing multi-junction cells it is therefore important to choose photoactive regions/sub-cells with the right bandgaps in order to optimise harvesting of the solar spectrum. In this regard, for a tandem photovoltaic device that comprises two photoactive regions/sub-cells, a top photoactive region/sub-cell and a bottom photoactive region/sub-cell, it has been shown that the bottom photoactive region/sub-cell should have a band gap of around 1.1 eV whilst the top photoactive region/sub-cell should have a band gap of around 1.7eV (Coutts, T et al, (2002)“Modelled performance of polycrystalline thin-film tandem solar cells”, Progress in Photovoltaics: Research and Applications, 10(3), pp.195- 203).

Today’s photovoltaic (PV) market is dominated by single-junction solar cells made of silicon (bandgap = 1.12 eV). A low cost material with a complementing bandgap of about 1.7 eV has been sought for a long time within the scientific community. Consequently, there has been interest in developing hybrid organic-inorganic perovskite solar cells for use in tandem photovoltaic devices given that the band gap of these perovskite materials can be tuned from around 1 5eV to over 2eV by varying the halide composition of organometal halide perovskites (Noh, J. et al, (2013)“Chemical Management for Colorful, Efficient, and Stable Inorganic- Organic Hybrid Nanostructured Solar Cells”, Nano Letters, p.1303211 12645008). In particular, by varying the halide composition it is possible to tune the band gap of an organometal halide perovskite to around 1 7eV, such that it is then ideal for use as the top sub-cell in a tandem structure when combined with a crystalline silicon bottom sub-cell.

In this regard, Schneider, B.W. et al (Schneider, B.W. et al., [2014]“Pyramidal surface textures for light trapping and antireflection in perovskite-on-silicon tandem solar cells”, Optics Express, 22(S6), p.A1422) reported on the modelling of a perovskite-on-silicon tandem cell in which the modelled cell has a 4-terminal, mechanically stacked structure. Loper, P. et al (Loper, P. et al., 2015. Organic-inorganic halide perovskite/crystalline silicon four-terminal tandem solar cells. Physical chemistry chemical physics: PCCP, 17, p.1619) reported on the implementation of a four-terminal tandem solar cell consisting of a methyl ammonium lead triiodide (OHbNHbR b) (i.e. organometal halide perovskite) top sub-cell mechanically stacked on a crystalline silicon heterojunction bottom sub-cell. Similarly, Bailie, C. et al. (Bailie, C. et al., 2015. Semi-transparent perovskite solar cells for tandems with silicon and CIGS. Energy Environ. Sci. , pp.1-28) reported on a mechanically-stacked tandem solar cell consisting of a methyl ammonium lead triiodide (CHsNHsPbL) top sub-cell on a copper indium gallium diselenide (CIGS) or low-quality multi-crystalline silicon bottom sub-cell. Filipic, M. et al. (Filipic, M. et al., 2015. CHsNHsPbb perovskite / silicon tandem solar cells: characterization based optical simulations. Optics Express, 23(7), pp.480-484) reported on the simulation of both mechanically stacked (four terminal) and monolithically integrated (two terminal) tandem devices consisting of a methyl ammonium lead triiodide (CHsNHsPbL) top sub-cell and a crystalline silicon bottom sub-cell. Mailoa, J.P. et al. (Mailoa, J.P. et al., [2015],“A 2-terminal perovskite/silicon multi-junction solar cell enabled by a silicon tunnel junction”, Applied Physics Letters, 106(12), p.121 105) then reported on the fabrication of a monolithic tandem solar cell consisting of a methyl ammonium lead triiodide (CHsNHsPbL) top sub-cell and a crystalline silicon bottom sub-cell. In a mechanically stacked multi-junction photovoltaic device the individual sub-cells are stacked on top of each other and are each provided with their own separate electrical contacts, such that the individual sub-cells are connected in parallel and do not require current matching. This contrasts with a monolithically integrated multi-junction photovoltaic device in which the individual sub-cells are electrically connected in series between a single pair of terminals, which results in the need for a recombination layer or a tunnel junction and current matching between adjacent sub-cells. Whilst a mechanically stacked multi-junction photovoltaic device does not require current matching between the sub-cells, the additional size and cost of the additional contacts and substrates, and the high losses due to lateral transport, make mechanically stacked structures less favourable than monolithically integrated structures.

SUMMARY OF THE PRESENT INVENTION

According to a first aspect of the present invention there is provided a multijunction photovoltaic device as specified in claims 1 to 13.

According to a second aspect of the present invention there is provided a method of fabricating a multi-junction photovoltaic device comprising a perovskite sub-cell in electrical contact with a silicon sub-cell, as specified in claims 14 to 20.

The present invention outlines a fabrication method and final device stack that is very compatible with the existing manufacturing equipment for high efficiency monocrystalline p- type silicon solar cells as are installed to allow multi GW annual production. The present invention would allow the owner of such equipment to upgrade its existing line to allow tandem cell manufacturing with only small investment. It is therefore commercially highly relevant.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will now be more particularly described, by way of example only, with reference to the accompanying drawings, in which:

Figure 1 illustrates schematically the method of manufacture of a p-type silicon substrate suitable for use in the manufacture of p-type silicon solar cells;

Figure 2 illustrates schematically the manufacture of an aluminium-back surface field or Al- BSF silicon solar cell using the substrate of figure 1 ; Figure 3 illustrates schematically the manufacture of a passivated emitter and rear contact (PERC) silicon solar cell using the substrate of figure 1 ;

Figure 4 shows a first embodiment of a multi-junction solar cell according to the present invention;

Figure 5 shows a fifth embodiment of a multi-junction solar cell according to the present invention; and

Figure 6 shows a method of making the embodiment shown in Figure 4, wherein the perovskite sub-cell on the top light-facing side of a multi-junction solar cell is deposited upon the flat“rear” surface of a conventional PERC silicon sub-cell which has been inverted so that the flat surface faces the incoming sunlight in use.

DETAILED DESCRIPTION

Definitions

The term“photoactive”, as used herein, refers to a region, layer or material that is capable of responding to light photoelectrically. A photoactive region, layer or material is therefore capable of absorbing the energy carried by photons in light that then results in the generation of electricity (e.g. by generating either electron-hole pairs or excitons).

The term“perovskite”, as used herein, refers to a material with a three-dimensional crystal structure related to that of CaTiCh or a material comprising a layer of material, which layer has a structure related to that of CaTiC>3. The structure of CaTiCh can be represented by the formula ABX₃, wherein A and B are cations of different sizes and X is an anion. In the unit cell, the A cations are at (0,0,0), the B cations are at (1/2, 1/2, 1/2) and the X anions are at (1/2, 1/2, 0). The A cation is usually larger than the B cation. The skilled person will appreciate that when A, B and X are varied, the different ion sizes may cause the structure of the perovskite material to distort away from the structure adopted by CaTiCh to a lower-symmetry distorted structure. The symmetry will also be lower if the material is in the form of a layer that has a structure related to that of bulk CaTi03. Materials comprising a layer of perovskite material are well known. For instance, the structure of materials adopting the K2N1F4 type structure comprises a layer of perovskite material. The skilled person will appreciate that a perovskite material can be represented by the formula [A][B][X]₃, wherein [A] is at least one cation, [B] is at least one cation and [X] is at least one anion. When the perovskite comprises more than one A cation, the different A cations may distributed over the A sites in an ordered or disordered way. When the perovskite comprises more than one B cation, the different B cations may distributed over the B sites in an ordered or disordered way. When the perovskite comprise more than one X anion, the different X anions may distributed over the X sites in an ordered or disordered way. The symmetry of a perovskite comprising more than one A cation, more than one B cation or more than one X cation, will often be lower than that of CaTiC>3.

As mentioned in the preceding paragraph, the term“perovskite”, as used herein, refers to (a) a material with a three-dimensional crystal structure related to that of CaTiC>3 or (b) a material comprising a layer of material, wherein the layer has a structure related to that of CaTiCh. Although both of these categories of perovskite may be used in the devices according to the invention, it is preferable in some circumstances to use a perovskite of the first category, (a), i.e. a perovskite having a three-dimensional (3D) crystal structure. Such perovskites typically comprise a 3D network of perovskite unit cells without any separation between layers. Perovskites of the second category, (b), on the other hand, include perovskites having a two- dimensional (2D) layered structure. Perovskites having a 2D layered structure may comprise layers of perovskite unit cells that are separated by (intercalated) molecules; an example of such a 2D layered perovskite is [2-(1-cyclohexenyl)ethylammonium]₂PbBr₄. 2D layered perovskites tend to have high exciton binding energies, which favours the generation of bound electron-hole pairs (excitons), rather than free charge carriers, under photoexcitation. The bound electron-hole pairs may not be sufficiently mobile to reach the p-type or n-type contact where they can then transfer (ionise) and generate free charge. Consequently, in order to generate free charge, the exciton binding energy has to be overcome, which represents an energetic cost to the charge generation process and results in a lower voltage in a photovoltaic cell and a lower efficiency. In contrast, perovskites having a 3D crystal structure tend to have much lower exciton binding energies (on the order of thermal energy) and can therefore generate free carriers directly following photoexcitation. Accordingly, the perovskite semiconductor employed in the devices and processes of the invention is preferably a perovskite of the first category, (a), i.e. a perovskite which has a three-dimensional crystal structure. This is particularly preferable when the optoelectronic device is a photovoltaic device.

The perovskite material employed in the present invention is one which is capable of absorbing light and thereby generating free charge carriers. Thus, the perovskite employed is a light absorbing perovskite material. However, the skilled person will appreciate that the perovskite material could also be a perovskite material that is capable of emitting light, by accepting charge, both electrons and holes, which subsequently recombine and emit light. Thus, the perovskite employed may be a light-emitting perovskite.

As the skilled person will appreciate, the perovskite material employed in the present invention may be a perovskite which acts as an n-type, electron-transporting semiconductor when photo-doped. Alternatively, it may be a perovskite which acts as a p-type hole-transporting semiconductor when photo-doped. Thus, the perovskite may be n-type or p-type, or it may be an intrinsic semiconductor. In preferred embodiments, the perovskite employed is one which acts as an n-type, electron-transporting semiconductor when photo-doped. The perovskite material may exhibit ambipolar charge transport, and therefore act as both n-type and p-type semiconductor. In particular, the perovskite may act as both n-type and p-type semiconductor depending upon the type of junction formed between the perovskite and an adjacent material.

Typically, the perovskite semiconductor used in the present invention is a photosensitizing material, i.e. a material which is capable of performing both photogeneration and charge transportation.

The term“mixed-anion”, as used herein, refers to a compound comprising at least two different anions. The term“halide” refers to an anion of an element selected from Group 17 of the Periodic Table of the Elements, i.e., of a halogen. Typically, halide anion refers to a fluoride anion, a chloride anion, a bromide anion, an iodide anion or an astatide anion.

The term“metal halide perovskite”, as used herein, refers to a perovskite, the formula of which contains at least one metal cation and at least one halide anion. The term“organometal halide perovskite”, as used herein, refers to a metal halide perovskite, the formula of which contains at least one organic cation.

The term“organic material” takes its normal meaning in the art. Typically, an organic material refers to a material comprising one or more compounds that comprise a carbon atom. As the skilled person would understand it, an organic compound may comprise a carbon atom covalently bonded to another carbon atom, or to a hydrogen atom, or to a halogen atom, or to a chalcogen atom (for instance an oxygen atom, a sulphur atom, a selenium atom, or a tellurium atom). The skilled person will understand that the term“organic compound” does not typically include compounds that are predominantly ionic such as carbides, for instance.

The term "organic cation" refers to a cation comprising carbon. The cation may comprise further elements, for example, the cation may comprise hydrogen, nitrogen or oxygen. The term "inorganic cation" refers to a cation that is not an organic cation. By default, the term “inorganic cation” refers to a cation that does not contain carbon. The term“semiconductor”, as used herein, refers to a material with electrical conductivity intermediate in magnitude between that of a conductor and a dielectric. A semiconductor may be an n-type semiconductor, a p-type semiconductor or an intrinsic semiconductor.

The term "dielectric", as used herein, refers to material which is an electrical insulator or a very poor conductor of electric current. The term dielectric, as used herein, typically refers to materials having a band gap of equal to or greater than 3.0 eV, preferably greater than 4 eV.

The term“n-type”, as used herein, refers to a region, layer or material that comprises an extrinsic semiconductor with a larger concentration of electrons than holes. In n-type semiconductors, electrons are therefore majority carriers and holes are the minority carriers, and they are therefore electron transporting materials. The term“n-type region”, as used herein, therefore refers to a region of one or more electron transporting (i.e. n-type) materials. Similarly, the term“n-type layer” refers to a layer of an electron-transporting (i.e. an n-type) material. An electron-transporting (i.e. an n-type) material could be a single electron transporting compound or elemental material, or a mixture of two or more electron-transporting compounds or elemental materials. An electron-transporting compound or elemental material may be undoped or doped with one or more dopant elements.

The term“p-type”, as used herein, refers to a region, layer or material that comprises an extrinsic semiconductor with a larger concentration of holes than electrons. In p-type semiconductors, holes are the majority carriers and electrons are the minority carriers, and they are therefore hole transporting materials. The term“p-type region”, as used herein, therefore refers to a region of one or more hole transporting (i.e. p-type) materials. Similarly, the term“p-type layer” refers to a layer of a hole-transporting (i.e. a p-type) material. A hole transporting (i.e. a p-type) material could be a single hole-transporting compound or elemental material, or a mixture of two or more hole-transporting compounds or elemental materials. A hole-transporting compound or elemental material may be undoped or doped with one or more dopant elements.

The term “band gap”, (sometimes spelled bandgap) as used herein, refers to the energy difference between the top of the valence band and the bottom of the conduction band in a material. The skilled person may readily measure the band gap of a material without undue experimentation.

The term“layer”, as used herein, refers to any structure which is substantially laminar in form (for instance extending substantially in two perpendicular directions, but limited in its extension in the third perpendicular direction). A layer may have a thickness which varies over the extent of the layer. Typically, a layer has approximately constant thickness. The“thickness” of a layer, as used herein, refers to the average thickness of a layer. The thickness of layers may easily be measured, for instance by using microscopy, such as electron microscopy of a cross section of a film, or by surface profilometry for instance using a stylus profilometer.

The term“porous”, as used herein, refers to a material within which pores are arranged. Thus, for instance, in a porous material the pores are volumes within the body of the material where there is no material. The individual pores may be the same size or different sizes. The size of the pores is defined as the“pore size”. The limiting size of a pore, for most phenomena in which porous solids are involved, is that of its smallest dimension which, in the absence of any further precision, is referred to as the width of the pore (i.e. the width of a slit-shaped pore, the diameter of a cylindrical or spherical pore, etc.). To avoid a misleading change in scale when comparing cylindrical and slit-shaped pores, one should use the diameter of a cylindrical pore (rather than its length) as its "pore-width " (Rouquerol, J. et al, (1994) Recommendations for the characterization of porous solids (Technical Report). Pure and Applied Chemistry, 66(8)). The following distinctions and definitions were adopted in previous IUPAC documents (J. Haber. (1991) Manual on catalyst characterization (Recommendations 1991). Pure and Applied Chemistry.): micropores have widths (i.e. pore sizes) smaller than 2 nm; Mesopores have widths (i.e. pore sizes) of from 2 nm to 50 nm; and Macropores have widths (i.e. pore sizes) of greater than 50 nm. In addition, nanopores may be considered to have widths (i.e. pore sizes) of less than 1 nm.

Pores in a material may include“closed” pores as well as open pores. A closed pore is a pore in a material which is a non-connected cavity, i.e. a pore which is isolated within the material and not connected to any other pore and which cannot therefore be accessed by a fluid to which the material is exposed. An“open pore” on the other hand, would be accessible by such a fluid. The concepts of open and closed porosity are discussed in detail in J. Rouquerol et al..

Open porosity, therefore, refers to the fraction of the total volume of the porous material in which fluid flow could effectively take place. It therefore excludes closed pores. The term “open porosity” is interchangeable with the terms“connected porosity” and“effective porosity”, and in the art is commonly reduced simply to“porosity”. The term“without open porosity”, as used herein, therefore refers to a material with no effective porosity. Thus, a material without open porosity typically has no macropores and no mesopores. A material without open porosity may comprise micropores and nanopores, however. Such micropores and nanopores are typically too small to have a negative effect on a material for which low porosity is desired.

In addition, polycrystalline materials are solids that are composed of a number of separate crystallites or grains, with grain boundaries at the interface between any two crystallites or grains in the material. A polycrystalline material can therefore have both interparticle/interstitial porosity and intraparticle/internal porosity. The terms “interparticle porosity” and“interstitial porosity”, as used herein, refer to pores between the crystallites or grains of the polycrystalline material (i.e. the grain boundaries), whilst the terms“intra particle porosity” and“internal porosity”, as used herein, refer to pores within the individual crystallites or grains of the polycrystalline material. In contrast, a single crystal or monocrystalline material is a solid in which the crystal lattice is continuous and unbroken throughout the volume of the material, such that there are no grain boundaries and no interparticle/interstitial porosity.

The term “compact layer”, as used herein, refers to a layer without mesoporosity or macroporosity. A compact layer may sometimes have microporosity or nanoporosity.

The term "scaffold material", as used herein, therefore refers to a material that is capable of acting as a support for a further material. The term "porous scaffold material", as used herein, therefore refers to a material which is itself porous, and which is capable of acting as a support for a further material.

The term“transparent”, as used herein, refers to material or object allows visible light to pass through almost undisturbed so that objects behind can be distinctly seen. The term“semi transparent”, as used herein, therefore refers to material or object which has a transmission (alternatively and equivalently referred to as a transmittance) to visible light intermediate between a transparent material or object and an opaque material or object. Typically, a transparent material will have an average transmission for visible light (generally light with a wavelength of from 370 to 740 nm) of around 100%, or from 90 to 100%. Typically, an opaque material will have an average transmission for visible light of around 0%, or from 0 to 5%. A semi-transparent material or object will typically have an average transmission for visible light of from 10 to 90%, typically 40 to 60%. Unlike many translucent objects, semi-transparent objects do not typically distort or blur images. Transmission for light may be measured using routine methods, for instance by comparing the intensity of the incident light with the intensity of the transmitted light. The term“electrode”, as used herein, refers to a conductive material or object through which electric current enters or leaves an object, substance, or region. The term “negative electrode”, as used herein, refers to an electrode through which electrons leave a material or object (i.e. an electron collecting electrode). A negative electrode is typically referred to as an “anode”. The term“positive electrode”, as used herein, refers to an electrode through which holes leave a material or object (i.e. a hole collecting electrode). A positive electrode is typically referred to as a“cathode”. Within a photovoltaic device, electrons flow from the positive electrode/cathode to the negative electrode/anode, whilst holes flow from the negative electrode/anode to the positive electrode/cathode.

The term“front electrode”, as used herein, refers to the electrode provided on that side or surface of a photovoltaic device that it is intended will be exposed to sun light. The front electrode is therefore typically required to be transparent, semi-transparent, or at least light transmissive so as to allow light to pass through the electrode to the photoactive layers provided beneath the front electrode. The term“back electrode”, as used herein, therefore refers to the electrode provided on that side or surface of a photovoltaic device that is opposite to the side or surface that it is intended will be exposed to sun light.

The term“charge transporter” refers to a region, layer or material through which a charge carrier (i.e. a particle carrying an electric charge), is free to move. In semiconductors, electrons act as mobile negative charge carriers and holes act as mobile positive charges. The term“electron transporter” therefore refers to a region, layer or material through which electrons can easily flow and that will typically reflect holes (a hole being the absence of an electron that is regarded as a mobile carrier of positive charge in a semiconductor). Conversely, the term“hole transporter” refers to a region, layer or material through which holes can easily flow and that will typically reflect electrons.

The term“consisting essentially of” refers to a composition comprising the components of which it consists essentially as well as other components, provided that the other components do not materially affect the essential characteristics of the composition. Typically, a composition consisting essentially of certain components will comprise greater than or equal to 95 wt% of those components or greater than or equal to 99 wt% of those components.

The term“bifacial”, as used herein, refers to a photovoltaic device/solar cell/sub-cell that can collect light and generate electricity through both of its faces; the front, sun-exposed face and the rear face. Bifacial devices/cells achieve a power gain by making use of diffuse and reflected light as well as direct sunlight. In contrast, the term “monofacial” refers to a photovoltaic device/solar cell/sub-cell that can only collect light and generate electricity through its front, sun-exposed face.

The term“conform”, as used herein, refers to an object that is substantially the same in form or shape as an another object. A“conformal layer”, as used herein, therefore refers to a layer of material that conforms to the contours of the surface on which the layer is formed. In other words, the morphology of the layer is such that the thickness of the layer is approximately constant across the majority of the interface between the layer and the surface on which the layer is formed.

The term“passivation”, as used herein, refers to the reduction of electron hole recombination usually present at the surface and/or within the bulk of a defect rich material. There are different kinds of electronic passivation. One example is the saturation of open silicon bonds (dangling bonds) with hydrogen. Another is the formation of a silicon oxide layer on the surface.

The terms“tunnel recombination junction”,“tunnel junction” or“tunnel diode”, as used herein, refers to the n/p or p/n junctions between the sub cells of a multi junction solar cell. Usually, the adjacent layers are highly doped such that the potential barrier is so narrow that charge carriers can tunnel through it and recombine. Thus, the junction loses its rectifying characteristic. Ideally, the recombination mechanism is iso-energetic meaning that the one charge carrier type coming from the top cell is at the same potential as the opposite one coming from the bottom cell. In this ideal case, the mechanism is loss free.

A“rear emitter p-type cell” as described herein means a cell made from a p-type silicon wafer having a p-n junction adjacent the surface opposite to the surface carrying the perovskite top cell. This p-n junction will emit electrons into the bulk p-type region of the silicon sub-cell in use. The rear emitter junction can be a homojunction comprising an n-doped layer (for example fabricated by diffusion of an n-type dopant such as phosphorous). Alternatively it can comprise a heterojunction.

Device Structure - General

The present invention relates to a tandem or multijunction photovoltaic device such as a solar cell including a first sub-cell comprising a perovskite layer carried on a further sub-cell comprising a crystalline silicon layer. Currently over 80 % of commercial silicon solar cells are based on a p-type wafer. Figure 1 shows an exemplary process for making a starting substrate for a solar cell on a p-type silicon wafer. An ingot of p-type silicon is sawn into wafers consisting of p-type silicon (109). The sawing produces rough major surfaces on either side of the wafer. The saw damage (typically of the order of 10 microns in thickness or less depending on the wafering method) is then removed using a chemical etch, and the surfaces exposed to a further etch to provide a textured surface to promote light scattering in the finished device. Depending on the etch solution used (i.e. either acidic or basic) and the resulting etch-mechanism (i.e. isotropic or preferential along a direction given by crystal orientation), the surface structure can be either randomly distributed, faceted or pyramidal. The pyramidal surface structure is preferred to enhance light in-coupling and the optical path length in the final device. The wafer is then cleaned and doped n-type (usually by furnace heating in a phosphorous containing ambient). This produces thin n-doped layers (1 10) on both sides of the wafer with the p-type bulk material (109) in the middle. The phosphorous doping produces a surface layer of phosphorous doped glass which can be removed by wet etching if required. One side of the wafer (109) is etched to remove the n-doped layer on one side (1 10). This etching step is usually performed using an acidic etch solution containing e.g. HN03 and HF. The etch mechanisms is isotropic and the resulting surface is smoother than before the etching step. Typically 0.5 to 20 micrometers are removed. The resulting surface roughness (120) can be controlled e.g. by the etch solution composition and etch duration. A front passivation layer (11 1) with a thickness between 40 - 200 nm is then deposited on one side of the wafer. This layer will also act as an antireflective coating. This layer is usually silicon nitride, but could be silicon oxide, alumina or titanium oxide or a combination of these (see for example Chapter 3.5 in“Photovoltaic Solar Energy: From Fundamentals to Applications” edited by Angele Reinders and Pierre Verlinden, Wiley- Blackwell (2017) ISBN-13: 978-11 18927465). In one other manufacturing process sequence, which doesn’t have the one side etch step before the deposition of the passivation layer (1 11), the opposite side of the wafer to the side having the passivation layer is etched to remove the n-type diffused layer and to form a substantially flat p-type surface (120). Using this manufacturing process, the passivation layer (11 1) acts as a mask to protect one side of the wafer. As mentioned above, the roughness can be controlled. The optimum roughness of the surface (120) is dependent on the following cell process.

This substrate is the building block used to make a commercial solar cell. Using this substrate there are two main ways to fabricate silicon solar cells.

The first method is shown schematically in Figure 2, and produces what are known as aluminium-back surface field or AI-BSF cells. In this method a back electrode 130 typically comprising aluminium is deposited onto the flat p-type surface 120, and front contacts 140 comprising a reactive metal such as silver and other materials in a paste are screen printed over the front surface passivation layer The wafer is then fired to produce a p+ doped layer of aluminium silicate 131 adjacent the back electrode and to allow the front contacts 140 to penetrate the passivation layer 111 to provide low resistance contacts to the front and back of the cell. Such AI-BSF cells have a typical efficiency of 16 - 19%.

The second method is shown schematically in Figure 3, and produces what are known as passivated emitter and rear contact (PERC) cells. In this method a thin dielectric film or typically a double layer of two dielectric films (150 and 151) is placed between the back electrode 130 and the p-type silicon substrate. Typically, a stack of AIOx and SiNx is used but also other dielectrics such as Si02 or Ti02 can be used. These layers act to reduce both optical absorption losses and recombination. An array of holes (152) in the dielectric film are then made (for example by laser ablation), covering about 1% of the surface allowing direct electrical contact to the rear surface 120. Back and front electrodes are then deposited, and fired as before to produce the finished cell. The slightly increased cost of manufacture of a PERC cell is counterbalanced by an increased efficiency - typically up to 22%. The rear surface roughness (120) in the PERC cell is of higher importance than for the AL-BSF cell because it is usually found that there is a trade-off between enhancing the light path on the one hand and the rear side passivation quality on the other hand. Thus, smooth rear sides are not uncommon in the industry due to the dissemination of the PERC cell.

A description of both types of cell can be found in Chapter 3.2 of“Photovoltaic Solar Energy: From Fundamentals to Applications” edited by Angele Reinders and Pierre Verlinden, Wiley- Blackwell (2017) ISBN-13: 978-1118927465). A typical PERC sub-cell structure is described, for example, by D. Zielke et al.“21.7 % Efficient PERC Solar Cells with AIOx T unneling Layer”, published in 26th European Photovoltaic Solar Energy Conference and Exhibition [2011] pp 1115-1119, ISBN 3-936338-27-2.

The present invention starts from the same building block shown in Figure 1 , and therefore is very compatible with conventional silicon solar cell manufacturing requiring minimal re-tooling. However, instead of using the flat surface silicon 120 as the back surface facing away from the incident sunlight, the p-type silicon sub-cell is flipped over so that a perovskite sub cell is formed on the flat surface 120 to provide the front sub-cell facing the incident sunlight, and the textured surface having the passivation layer 110 becomes the rear surface of the rear cell.

Therefore, instead of an opaque electrode being formed on the flat surface 120 the structure must be modified to include light transmissive, preferably almost transparent conductive layers carried by the surface 120. The top sub-cell comprising an n-i-p perovskite sub-cell is then deposited on top of the transparent conductive layer. Depending on the exact nature of the junction between the two sub-cells, several different specific structures can be made falling within the scope of the present invention. In essence, the transparent conductive layers are selected to passivate the surface (120) and to form a tunnel junction (also called tunnel diode) between the two sub cells.

In a first embodiment, the multijunction photovoltaic device comprises a perovskite sub-cell in electrical contact with a crystalline silicon sub-cell, as shown in figure 4. The fabrication sequence is illustrated in Figure 6a to 6e. In this sequence, Figures 6a and 6b show the fabrication of a passivated emitter and rear local diffusion (PERL) type cell, but without the deposition of the opaque back contact. The structure is then inverted, and the perovskite sub cell is fabricated on top of the“rear” flat surface of the silicon cell, which is now located on the “front” side of the silicon cell - i.e. the side facing the incoming sunlight in use. The fabrication of the top perovskite sub-cell of the multijunction device is then shown in sequence in Figures 6c, 6d and 6e.

The silicon sub-cell is thus located towards the rear of the multijunction device, such that incident sunlight passes through the top perovskite sub-cell first before entering the bottom silicon sub-cell. The silicon sub cell comprises a rear emitter p-type cell having a substantially opaque rear electrode (113), which is also preferably reflective (e.g. aluminium or silver can be used with a thickness of e.g. 0.1 - 20 micrometre, which reflects e.g. >80% and preferably >90% of the incoming light in the wavelength range of 300 - 1200 nm), this rear electrode making contact with an n-type diffused layer(1 10), which doping profile and depth can be controlled during the firing process (e.g. having an concentration of donors of 0.1 - 5 x 10¹⁹ cm^-3 at the surface and decreasing exponentially to the base doping within 0.1 - 10 micrometres) through via holes filled with metal (1 12) in a passivation layer (11 1). In another embodiment (not shown) the reflecting layer (1 13) can be left out. The resulting device would then be bifacial, which could be desirable for some applications. The passivation layer (11 1) and the metal filled via holes (1 12) are - in accordance with the idea to maximize compatibility to existing processes - made of e.g. silicon nitride with the standard thickness (40 - 200 nm) and the via holes are made by reactively etching a conventional screen printed silver paste through the silicon nitride using a conventional furnace firing process. In this embodiment, the screen printing and firing may be performed before or after the passivation step. The front surface (120) of the p-type silicon substrate (109) is provided with a thin tunnel passivation layer 108, in the present example comprising silicon dioxide with a thickness of e.g. 0.1 - 5 nm and preferably between 0.5 and 1.5 nm. The passivation layer can be formed e.g. wet chemically, using UV/ozone growth or together with the following silicon layer using an oxygen precursor such as C02, 02 or N20 in a CVD process. A p-type silicon layer 107 is provided over the passivation layer 108. The p-type layer (107) can be either amorphous or poly crystalline and has a thickness between 1 and 200 nm and preferably a thickness between 5 and 30 nm and an acceptor density of 1 - 100 x 10¹⁹ cm^-3 and can be deposited e.g. by CVD. After deposition of the silicon layer (107), the Si0₂/Si stack can be annealed to increase the crystalline volume fraction and reduce the surface recombination velocity at the Si/Si0₂ interfaces. An additional hydrogen plasma treatment can be applied to saturate dangling bonds. Such Si/Si0₂/Si passivated contacts are well known under names such as SIPOS (compare e.g. Yablonovitch et al., Applied Physics Letters, vol. 47, p. 121 1 (1985)), TOPCON (compare e.g. Feldmann et al., proceedings of the 28^th European PV Solar Energy Conference and Exhibition, 2013, Paris, France) or POLO (compare e.g. Peibst et al., proceedings of the 32nd European Photovoltaic Solar Energy Conference and Exhibition, 2016, Munich, Germany). Further details regarding the fabrication methods are described in these references.

The n-type silicon layer 106 is provided on top of the 108 and can be either amorphous or poly-crystalline and has a thickness between 1 and 200 nm and preferably a thickness between 5 and 30 nm and a donor density of 1 - 100 x 10¹⁹ cm^-3. The next layer to be deposited is a solid-state electron transporting layer (105), typically a fullerene derivative such as C60 or C60PCBM or C60IPB or C60 IPH with a layer thickness of 1 - 20 nm, and preferably between 2 - 15 nm. Such materials can be deposited from solution or via PVD and are commercially available from Solenne BV, Zernikepark 6, 9747AN Groningen, The Netherlands. Next, a nominally intrinsic perovskite layer 104 is deposited. On top of the perovskite layer a solid state hole transporting p-type layer 103 is provided.

Layer 103 may comprise an inorganic or an organic p-type material. Typically, the p-type region comprises a layer of an organic p-type material.

Suitable p-type materials may be selected from polymeric or molecular hole transporters. The p-type layer employed in the photovoltaic device of the invention may for instance comprise spiro-OMeTAD (2,2’,7,7’-tetrakis-(N,N-di-p-methoxyphenylamine)9,9’-spirobifluorene)), P3HT (poly(3-hexylthiophene)), PCPDTBT (Poly[2, 1 ,3-benzothiadiazole-4,7-diyl[4,4-bis(2- ethylhexyl)-4H-cyclopenta[2, 1-b:3,4-b']dithiophene-2,6-diyl]]), PVK (poly(N-vinylcarbazole)), HTM-TFSI (1-hexyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide), Li-TFSI (lithium bis(trifluoromethanesulfonyl)imide) or tBP (tert-butylpyridine). The p-type region may comprise carbon nanotubes. Usually, the p-type material is selected from spiro-OMeTAD, P3HT, PCPDTBT and PVK. Preferably, the p-type region consists of a p-type layer that comprises spiro-MeOTAD. The p-type layer (103) is between 1 - 500 nm and prefereably between 5 - 400 nm in thickness.

The next layer in the stack (102) is a thin barrier layer comprising, for example, a transition metal oxide, preferable molybdenum trioxide, although tungsten, aluminium or vanadium oxides can be used. This layer is optional, and can be deposited by physical vapour deposition, atomic layer deposition or from solution. The thickness of this barrier layer (102) is between 1 - 100 nm, and preferably between 5 - 30 nm. On top of this is deposited a transparent conductive oxide layer 101. This may comprise, for example, indium tin oxide (ITO) typically deposited by physical vapour deposition. Finally, a grid of front electrodes 100 is printed on the top of the stack to complete the multijunction solar cell. Such front electrodes are typically screen printed using a paste containing a conductive material, usually a metal such as silver. The grid electrodes only cover a small percentage of the surface area of the device to allow the maximum amount of light into the device.

The layers 108, 107 and 106 in the above embodiment are one example of a suitable device stack. A number of other embodiments may be fabricated as described next. In all these alternative embodiments the other layers in the device remain substantially as described for the first embodiment shown in figure 1.

In a second embodiment, the tunnel passivation layer 108 is omitted, and layer 107 comprises a p+ layer of mono- or polycrystalline or amorphous silicon. As in the first embodiment, 107 has a thickness between 1 and 200 nm and preferably a thickness between 5 and 30 nm and an acceptor density between 1 - 100 x 10¹⁹ cm^-3 and can be deposited e.g. by chemical vapor deposition. Layer 106 in this embodiment comprises n-type silicon or a transparent conductive oxide layer such as ITO. The thickness of 106 is between 1 and 200 nm and preferably between 5 and 30 nm. In a third embodiment, the layer of ITO is omitted, and layer 107 comprising p+ silicon is in direct contact with electron transporting layer 105.

In a fourth embodiment, layer 108 is present and comprises a layer of intrinsic amorphous silicon with a thickness of 1 - 20 nm and preferably 3 - 8 nm. In this example, the layer 107 comprises p+ silicon, silicon carbide or silicon oxide, which can be amorphous or partially crystalline. This layer can either be in contact with a layer 106 of a transparent conductive oxide as before, or directly in contact with electron transporting layer 105. In this embodiment, the stack comprising layers 108 and 107 are similar to the Panasonic HIT passivation/emitter stack. Since the intrinsic amorphous silicon passivation quality depends on a high amount of hydrogen being present at the wafer surface to passivate dangling bonds, the process sequence has to be done such that the firing of the metal fingers through the silicon nitride passivation layer (111) to form metal filled via holes (112) is being performed before the deposition of the layers 108 and 107 to avoid hydrogen effusion and to avoid crystallisation of the amorphous layer passivation layer.

A fifth embodiment is shown in figure 5, and is related very closely to a conventional PERC cell which has been inverted.

A thin dielectric film or a combination of dielectric films (150 and 151) is disposed on the flat un-textured surface 120 of a p-type silicon substrate (109). This film acts to reduce recombination (passivation). It typically comprises a layer of alumina 150 adjacent the p type substrate and a layer 151 of silicon nitride. An array of holes in the dielectric film are then made (for example by laser ablation, wet or dry etching), covering about 1% of the surface allowing direct electrical contact to the surface 120 for a subsequently deposited p+ layer 107, which consists, like, for example, in the first embodiment, of silicon. The layer has a thickness of 1-200 nm and preferably 5-30 nm and preferably a crystalline volume fraction of >50%. An n-type layer 106 is then deposited as in the first embodiment, followed by the n-i-p layers comprising the top perovskite sub-cell as before

The photovoltaic device of the present invention has a multi-junction structure comprising a first sub-cell disposed over a second sub-cell, the first sub-cell comprising a photoactive region comprising the perovskite material. The photovoltaic device may have a monolithically integrated structure. In a monolithically integrated multi-junction photovoltaic device the two or more photovoltaic sub-cells are deposited directly onto one another and are therefore electrically connected in series.

The photoactive region 104 of the perovskite sub-cell may comprise a layer of the perovskite material without open porosity. Alternatively, the layer of the perovskite material may be in contact with a porous scaffold material that is disposed between an n-type region and a p-type region. The porous scaffold material may comprise or consist essentially of any of a dielectric material and a semiconducting/charge transporting material. The layer of the perovskite material may then be disposed within pores of/be conformal with a surface of the porous scaffold material. Alternatively, the layer of the perovskite material may fill the pores of the porous scaffold material and form a capping layer on the porous scaffold material, which capping layer consists of a layer of the photoactive material without open porosity. The layer thickness of the photoactive region 104 may be between 100 and 5000 nm and preferably between 300 and 1000 nm. The multi-junction photovoltaic device further comprises a first electrode and a second electrode, with the first sub-cell and the second sub-cell disposed between the first and second electrodes.

The first electrode (101 , 100) is in contact with the p-type region of the first (perovskite) sub- cell, and comprises a transparent, semi-transparent or light transmissive electrically conductive material. The first electrode may then be an electron collecting electrode, whilst the second electrode is a hole collecting electrode. In a tandem device, the second electrode will then be in contact with the second (crystalline silicon) sub-cell, and will comprise a electrically conductive light reflective material such as a metal (for example Al or Ag 112, 113). The perovskite sub cell preferably comprises a n-i-p structure, in which the n-type layer is deposited first (to be in contact with the p+ layer of the underlying silicon sub-cell), followed by the intrinsic layer and then a p-type layer on top, followed by the light transmissive first electrode.

The silicon sub cell may comprise a single crystal (mono-crystalline), poly- or micro-crystalline semiconducting material.

In the above described multi-junction photovoltaic devices, the hole transporting layer (103) of the first (perovskite) sub-cell typically comprises one or more p-type layers. Often, the p-type region is a p-type layer, i.e. a single p-type layer. In other examples, however, the p-type region may comprise a p-type layer and a p-type exciton blocking layer or electron blocking layer. If the valence band (or highest occupied molecular orbital energy levels) of the exciton blocking layer is closely aligned with the valence band of the photoactive material, then holes can pass from the photoactive material into and through the exciton blocking layer, or through the exciton blocking layer and into the photoactive material, and we term this a p-type exciton blocking layer. An example of such is tris[4-(5-phenylthiophen-2-yl)phenyl]amine, as described in Masaya Hirade, and Chihaya Adachi,“Small molecular organic photovoltaic cells with exciton blocking layer at anode interface for improved device performance” Appl. Phys. Lett. 99, 153302 (2011). A p-type layer is a layer of a hole-transporting (i.e. a p-type) material. The p-type material may be a single p-type compound or elemental material, or a mixture of two or more p-type compounds or elemental materials, which may be undoped or doped with one or more dopant elements. A p-type layer may comprise an inorganic or an organic p-type material. Typically, the p-type region comprises a layer of an organic p-type material.

Suitable p-type materials may be selected from polymeric or molecular hole transporters. The p-type layer employed in the photovoltaic device of the invention may for instance comprise spiro-OMeTAD (2,2’,7,7’-tetrakis-(N,N-di-p-methoxyphenylamine)9,9’-spirobifluorene)), P3HT (poly(3-hexylthiophene)), PCPDTBT (Poly[2, 1 ,3-benzothiadiazole-4,7-diyl[4,4-bis(2- ethylhexyl)-4H-cyclopenta[2, 1-b:3,4-b']dithiophene-2,6-diyl]]), PVK (poly(N-vinylcarbazole)), HTM-TFSI (1-hexyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide), Li-TFSI (lithium bis(trifluoromethanesulfonyl)imide) or tBP (tert-butylpyridine). The p-type region may comprise carbon nanotubes. Usually, the p-type material is selected from spiro-OMeTAD, P3HT, PCPDTBT and PVK. Preferably, the p-type region consists of a p-type layer that comprises spiro-MeOTAD.

A p-type layer may for example comprise spiro-OMeTAD (2,2’,7,7’-tetrakis-(N,N-di-p- methoxyphenylamine)9,9’-spirobifluorene)), P3HT (poly(3-hexylthiophene)), PCPDTBT (Poly[2, 1 ,3-benzothiadiazole-4,7-diyl[4,4-bis(2-ethylhexyl)-4H-cyclopenta[2, 1-b:3,4- b']dithiophene-2,6-diyl]]), or PVK (poly(N-vinylcarbazole)).

Suitable p-type materials also include molecular hole transporters, polymeric hole transporters and copolymer hole transporters. The p-type material may for instance be a molecular hole transporting material, a polymer or copolymer comprising one or more of the following moieties: thiophenyl, phenelenyl, dithiazolyl, benzothiazolyl, diketopyrrolopyrrolyl, ethoxydithiophenyl, amino, triphenyl amino, carbozolyl, ethylene dioxythiophenyl, dioxythiophenyl, or fluorenyl. Thus, a p-type layer employed in the photovoltaic device of the invention may for instance comprise any of the aforementioned molecular hole transporting materials, polymers or copolymers.

Suitable p-type materials also include m-MTDATA (4, 4', 4"- tris(methylphenylphenylamino)triphenylamine), MeOTPD (N,N,N',N'-tetrakis(4- methoxyphenyl)-benzidine), BP2T (5,5'-di(biphenyl-4-yl)-2,2'-bithiophene), Di-NPB (N,N'-Di- [(1-naphthyl)-N,N'-diphenyl]-1 , 1 '-biphenyl)-4,4'-diamine), a-NPB (N,N’-di(naphthalen-1-yl)- N,N’-diphenyl-benzidine), TNATA (4,4',4"-tris-(N-(naphthylen-2-yl)-N- phenylamine)triphenylamine), BPAPF (9,9-bis[4-(N,N-bis-biphenyl-4-yl-amino)phenyl]-9H- fluorene), spiro-NPB (N2,N7-Di-1-naphthalenyl-N2,N7-diphenyl-9,9'-spirobi[9H-fluorene]-2,7- diamine), 4P-TPD (4,4-bis-(N,N-diphenylamino)-tetraphenyl), PEDOT:PSS and spiro- OMeTAD. A p-type layer may be doped, for instance with tertbutyl pyridine and LiTFSI. A p-type layer may be doped to increase the hole-density. A p-type layer may for instance be doped with NOBF4 (Nitrosonium tetrafluoroborate), to increase the hole-density.

In other examples, a p-type layer may comprise an inorganic hole transporter. For instance, a p-type layer may comprise an inorganic hole transporter comprising an oxide of nickel, vanadium, copper or molybdenum; Cul, CuBr, CuSCN, Cu20, CuO or CIS; a perovskite; amorphous Si; a p-type group IV semiconductor, a p-type group lll-V semiconductor, a p-type group ll-VI semiconductor, a p-type group l-VII semiconductor, a p-type group IV-VI semiconductor, a p-type group V-VI semiconductor, and a p-type group ll-V semiconductor, which inorganic material may be doped or undoped. A p-type layer may be a compact layer of said inorganic hole transporter.

A p-type layer may for instance comprise an inorganic hole transporter comprising an oxide of nickel, vanadium, copper or molybdenum; Cul, CuBr, CuSCN, Cu20, CuO or CIS; amorphous Si; a p-type group IV semiconductor, a p-type group lll-V semiconductor, a p-type group ll-VI semiconductor, a p-type group l-VII semiconductor, a p-type group IV-VI semiconductor, a p- type group V-VI semiconductor, and a p-type group ll-V semiconductor, which inorganic material may be doped or undoped. A p-type layer may for instance comprise an inorganic hole transporter selected from Cul, CuBr, CuSCN, Cu20, CuO and CIS. A p-type layer may be a compact layer of said inorganic hole transporter.

The p-type region may have a thickness of from 5 nm to 1000 nm. For instance, the p-type region may have a thickness of from 50 nm to 500 nm, or from 100 nm to 500 nm. In the above described multi-junction photovoltaic devices, the p-type region 112 of the first sub-cell preferably has a thickness from 10 nm to 50 nm, and more preferably of approximately 20 nm. The p-type region could also comprise a bi-layer or multilayer structure consisting of 2 or more layers having different materials.

Electron transporting layers 105 suitable for use in perovskite photovoltaic sub cells in the present embodiments have recently been described in the review paper“Current status of electron transport layers in perovskite solar cells: materials and properties”, Mahmood, Sarwar and Mehran, RSC Adv. 2017.7.17044.

The electron transporting layers typically comprise n-type regions. In the above described multi-junction photovoltaic device, the n-type region of the first sub-cell comprises one or more n-type layers. Often, the n-type region is an n-type layer, i.e. a single n-type layer. In other examples, however, the n-type region may comprise an n-type layer and a separate n-type exciton blocking layer or hole blocking layer.

An exciton blocking layer is a material which is of wider band gap than the photoactive material, but has either its conduction band or valance band closely matched with those of the photoactive material. If the conduction band (or lowest unoccupied molecular orbital energy levels) of the exciton blocking layer are closely aligned with the conduction band of the photoactive material, then electrons can pass from the photoactive material into and through the exciton blocking layer, or through the exciton blocking layer and into the photoactive material, and we term this an n-type exciton blocking layer. An example of such is bathocuproine (BCP), as described in P. Peumans, A. Yakimov, and S. R. Forrest,“Small molecular weight organic thin-film photodetectors and solar cells” J. Appl. Phys. 93, 3693 (2001) and Masaya Hirade, and Chihaya Adachi,“Small molecular organic photovoltaic cells with exciton blocking layer at anode interface for improved device performance” Appl. Phys. Lett. 99, 153302 (201 1)}.

The n-type layer (105) is a layer of an electron-transporting (i.e. an n-type) material. The n- type material may be a single n-type compound or elemental material, or a mixture of two or more n-type compounds or elemental materials, which may be undoped or doped with one or more dopant elements.

The electron transporting material employed may comprise an inorganic or an organic n-type material.

A suitable inorganic n-type material may be selected from a metal oxide, a metal sulphide, a metal selenide, a metal telluride, a perovskite, amorphous or nanocrystalline Si, an n-type group IV semiconductor, an n-type group lll-V semiconductor, an n-type group ll-VI semiconductor, an n-type group l-VII semiconductor, an n-type group IV-VI semiconductor, an n-type group V-VI semiconductor, and an n-type group ll-V semiconductor, any of which may be doped or undoped.

More typically, the n-type material is selected from a metal oxide, a metal sulphide, a metal selenide, and a metal telluride.

Thus, an n-type layer may comprise an inorganic material selected from oxide of titanium, tin, zinc, niobium, tantalum, tungsten, indium, gallium, neodymium, palladium, or cadmium, or an oxide of a mixture of two or more of said metals. For instance, an n-type layer may comprise T1O2, SnC>2, ZnO, Nb₂05, Ta₂Os, WO3, W2O5, ln₂C>3, Ga₂C>3, Nd₂03, PbO, or CdO.

Other suitable n-type materials that may be employed include sulphides of cadmium, tin, copper, or zinc, including sulphides of a mixture of two or more of said metals. For instance, the sulphide may be FeS2, CdS, ZnS, SnS, BiS, SbS, or Cu2ZnSnS₄.

An n-type layer may for instance comprise a selenide of cadmium, zinc, indium, or gallium or a selenide of a mixture of two or more of said metals; or a telluride of cadmium, zinc, cadmium or tin, or a telluride of a mixture of two or more of said metals. For instance, the selenide may be Cu(ln,Ga)Se2. Typically, the telluride is a telluride of cadmium, zinc, cadmium or tin. For instance, the telluride may be CdTe.

An n-type layer may for instance comprise an inorganic material selected from oxide of titanium, tin, zinc, niobium, tantalum, tungsten, indium, gallium, neodymium, palladium, cadmium, or an oxide of a mixture of two or more of said metals; a sulphide of cadmium, tin, copper, zinc or a sulphide of a mixture of two or more of said metals; a selenide of cadmium, zinc, indium, gallium or a selenide of a mixture of two or more of said metals; or a telluride of cadmium, zinc, cadmium or tin, or a telluride of a mixture of two or more of said metals.

Examples of other semiconductors that may be suitable n-type materials, for instance if they are n-doped, include group IV elemental or compound semiconductors; amorphous Si; group lll-V semiconductors (e.g. gallium arsenide); group ll-VI semiconductors (e.g. cadmium selenide); group l-VII semiconductors (e.g. cuprous chloride); group IV-VI semiconductors (e.g. lead selenide); group V-VI semiconductors (e.g. bismuth telluride); and group ll-V semiconductors (e.g. cadmium arsenide).

When an n-type layer is an inorganic material, for instance T1O2 or any of the other materials listed above, it may advantageously be a compact layer of said inorganic material. Preferably the n-type layer is a compact layer of T1O2.

Other n-type materials may also be employed, including organic and polymeric electron transporting materials, and electrolytes.

Particularly preferred materials for the electron transporting layer 105 are fullerene or fullerene derivative, For example, the n-type region may comprise a n-type layer comprising one or more of C60, C70, C84, C60-PCBM, C70-PCBM, C84-PCBM and carbon nanotubes. It may comprise C60-IPB, C60-IPH, C70-IPB, C70IPH or mixtures thereof. Such materials are commercially available from Solenne BV, Zernikepark 6, 9747AN Groningen, The Netherlands.

Alternative organic electron transporting materials include comprising perylene or a derivative thereof, or poly{[N,N0-bis(2-octyldodecyl)-naphthalene-1 ,4,5,8-bis(dicarboximide)-2,6-diyl]- alt-5,50-(2, 20-bithiophene)} (P(NDI20D-T2)).

The n-type region may have a thickness of from 3 nm to 1000 nm. Where the n-type region comprises a compact layer of an n-type semiconductor, the compact layer has a thickness of from 5 nm to 200 nm. In the above described multi-junction photovoltaic device, the n-type region 1 11 of the first sub-cell 210 preferably has a thickness from 5 to 20 nm.

Device Structure - Transparent Electrode

In the above described multi-junction photovoltaic device, the thin light transmissive conductive layer (101) is the electrode provided on that side or surface of the photovoltaic device that it is intended will be directly exposed to sun light. The first electrode (101) is therefore required to be light transmissive, preferably transparent, so as to maximise the transmission of the light through the electrode to the photoactive layers of the first and second sub-cells provided beneath, whilst also having sufficient electrical conductivity. In particular, for multi-junction devices, the first electrode should transmit a large proportion of light over the complete optical window (i.e. from 300 nm to 1200 nm in wavelength) as transmission of the longer wavelengths is highly important for achieving useful power conversion efficiencies.

The first electrode 101 therefore preferably consists of material that has a sheet resistance (Rs) from 10 ohms per square (W/sq) to 100 W/sq and an average transmission for visible and infrared light of at least 85% (i.e. transmits at least 85% of light from 300 nm to 1200 nm in wavelength). More preferably, the first electrode 15 consists of material that has a sheet resistance (Rs) of equal to or less than 50 W/sq and an average transmission for visible and infrared light of greater than 90%, and preferably has an average transmission for visible and infrared light of at least 95%.

Particularly suitable materials for use as the transparent front electrode include transparent conductive oxides (TCO). Transparent conductive oxides (TCO) are doped metal oxides that are electrically conductive and have a comparably low absorption of light. TCOs can have greater than 80% transmittance of incident light as well as conductivities higher than 10⁴ S/cm (i.e. resistivity of ~10^-4 W-cm) for efficient carrier transport. Examples of suitable TCO materials include indium tin oxide (ITO), hydrogen doped indium oxide (IOH), aluminium doped zinc oxide (AZO), fluorine doped tin oxide (FTO), indium-doped zinc oxide (IZO), niobium-doped titanium dioxide (Nb:TiC>2), etc..

The first electrode 101 therefore preferably comprises of a layer of transparent conductive oxides (TCO). By way of example, the first electrode 15 can comprise of a layer of any of indium tin oxide (ITO), aluminium doped zinc oxide (AZO), fluorine doped tin oxide (FTO), indium-doped zinc oxide (IZO), and niobium-doped titanium dioxide (Nb:Ti0₂). More preferably, the first electrode 101 therefore preferably consists of a layer of indium tin oxide (ITO). When the first electrode 101 consists of a layer of indium tin oxide (ITO) it is preferable that the layer has a thickness of from 100 nm to 200 nm, and more preferably of 150 nm.

Conventional techniques for fabrication of layers of TCO materials typically involve a magnetron sputtering process. To minimise ion impact damage to underlying layers, an additional protective inorganic buffer layer may be employed. Alternatively atomic layer deposition, or a remotely generated plasma may be used to deposit a layer of TCO, as described in WO2016/198898. Such methods do not require a high temperature annealing step.

The TCO preferably has a sheet resistance (R_s) equal to or less than 50 ohms per square (W/sq) and an average transmission for visible and infrared light of greater than 90% (i.e. transmits at least 90% of light above 300nm in wavelength), and preferably has an average transmission for visible and infrared light of at least 95 %.

Device Structure - Perovskite Material

In the above described multi-junction photovoltaic devices, the perovskite material in the photoactive region of the first sub-cell 104 is configured to function as a light absorber/a photosensitizer within the photoactive region. The perovskite material then preferably has a band gap from 1.50eV to 1.75eV, and more preferably from 1.65eV to 1.70eV. The second PERC sub-cell comprising silicon then preferably has a band gap of around 1.1 eV.

The perovskite material may have general formula (I):

[A][B][X]s (I)

wherein [A] is one or more monovalent cations, [B] is one or more divalent inorganic cations, and [X] is one or more halide anions. [X] preferably comprises one or more halide anions selected from fluoride, chloride, bromide, and iodide, and preferably selected from chloride, bromide and iodide. More preferably [X] comprises one or more halide anions selected from bromide and iodide. In some examples, [X] preferably comprises two different halide anions selected from fluoride, chloride, bromide, and iodide, and preferably selected from chloride, bromide and iodide, and more preferably comprises bromide and iodide.

[A] preferably comprises one or more organic cations selected from methylammonium (CH₃NH₃ ⁺), formamidinium (HC(NH)2)2⁺), and ethyl ammonium (CH₃CH2NH₃ ⁺), and preferably comprises one organic cation selected from methylammonium (CH₃NH₃ ⁺) and formamidinium (HC(NH₂)₂ ⁺). [A] may comprise one or more inorganic cations selected from Cs⁺, Rb⁺, Cu⁺, Pd⁺, Pt⁺, Ag⁺, Au⁺, Rh⁺, and Ru⁺.

[B] preferably comprises at least one divalent inorganic cation selected from Pb²⁺ and Sn²⁺, and preferably comprises Pb²⁺.

In preferred examples, the perovskite material has the general formula:

AxA’i-xB(X_yXV_y)₃ (IA)

wherein A is formamidinium (FA), A’ is a caesium cation (Cs⁺), B is Pb²⁺, X is iodide and X’ is bromide, and wherein 0 < x £ 1 and 0 < y £ 1. In these preferred embodiments, the perovskite material can therefore comprise a mixture of two monovalent cations. In addition, in the preferred embodiments, the perovskite material can therefore comprise either a single iodide anion or a mixture of iodide and bromide anions. The present inventors have found such perovskite materials can have band gaps in from 1.50eV to 1.75eV and that layers of such perovskite materials can be readily formed with suitable crystalline morphologies and phases. More preferably, the perovskite material is FAi-_xCs_xPbl3-_yBr_y.

In order to provide highly efficient photovoltaic devices, the absorption of the absorber should ideally be maximised so as to generate an optimal amount of current. Consequently, when using a perovskite as the absorber in a photovoltaic device or sub-cell, the thickness of the perovskite layer should ideally be in the order of from 300 to 600nm, in order to absorb most of the sun light across the visible spectrum. Typically, therefore, the thickness of the layer of the perovskite material is greater than 100nm. The thickness of the layer of the perovskite material in the photovoltaic device may for instance be from 100 nm to 1000 nm. The thickness of the layer of the perovskite material in the photovoltaic device may for instance be from 200 nm to 700 nm, and is preferably from 300nm to 600nm. In the above described multi-junction photovoltaic devices, the planar layer of perovskite material 1 1 in the photoactive region of the first/top sub-cell 210 preferably has a thickness from 350 nm to 450 nm, and more preferably of approximately 400nm.

The perovskite layer may be prepared as described in WO2013/171517, WO2014/045021 , WO2016/198889, WO2016/005758, WO2017/089819, and in the reference books “Photovoltaic Solar Energy: From Fundamentals to Applications” edited by Angele Reinders and Pierre Verlinden, Wiley- Blackwell (2017) ISBN-13: 978-1 118927465 and “Organic- Inorganic Halide Perovskite Photovoltaics: From Fundamentals to Device Architectures” edited by Nam-Gyu Park et al., Springer (2016) ISBN-13: 978-3319351124.

In the fifth embodiment shown in figure 5, the tunnel recombination junction (TRJ) to the top cell is made by selectively opening the passivating film 150. This can be achieved a number of techniques.

The first technique is known as laser ablation. Short pulse laser ablation is described, for example in“Extended study on short pulse laser ablation of dielectric layers” by Theobald, Mayerhofer, Grosser, Harney and Schneider, 28^th European Photovoltaic Solar Energy Conference and Exhibition, Paris 2014 pp 1231 - 35, or“Influence of ultra-short pulse laser ablation of silicon nitride passivation layers” by Trusheim, Schultz-Ruhtenberg, Smeets, Das and Wieduwilt, 26^th European Photovoltaic Solar Energy Conference and Exhibition, pp 1623 - 27.

The opening of the passivating contact should be less than 5% of the total front area, preferably less than 1 %. A large number, meaning a high density, of small openings (<100 urn diameter, preferably <30 urn) is preferable over a smaller number of larger openings. The density of openings should be >10 cm^-2, preferably >200 cm^-2. After opening the contact, the perovskite sub-cell can be deposited. Another cleaning step prior to the perovskite deposition is advantageous.

In the above described examples, the multi-junction photovoltaic device could be considered to be monofacial, such that it is configured to only collect light and generate electricity through its front, sun-exposed face. However, the majority of the features described above are equally applicable to a bifacial multi-junction photovoltaic device that can collect light and generate electricity through both of its faces; the front, sun-exposed face and the rear face. In particular, the present inventors have recognised that the multi-junction photovoltaic device could be configured into a bifacial architecture. It will be appreciated that individual items described above may be used on their own or in combination with other items shown in the drawings or described in the description and that items mentioned in the same passage as each other or the same drawing as each other need not be used in combination with each other. Furthermore, although the invention has been described in terms of preferred embodiments as set forth above, these embodiments are illustrative only. Those skilled in the art will be able to make modifications and alternatives in view of the disclosure which are contemplated as falling within the scope of the appended claims.

Claims

1. A multi-junction photovoltaic device comprising a first sub-cell (100, 101 , 102, 103, 104, 105) stacked on and in electrical contact with a second sub-cell (109, 110, 111 , 112, 113), the first sub-cell comprising a layer of a perovskite material (104), the first sub-cell comprising the front of the multijunction photovoltaic device adapted to be exposed to incident light in use, characterised in that the second sub-cell comprises a rear emitter p-type silicon structure.

2. A multijunction photovoltaic device as claimed in claim 1 in which a tunnel junction (108, 107, 106) is provided between the first and second sub-cells.

3. A multi-junction photovoltaic device as claimed in claim 2, wherein the multi-junction photovoltaic device has a monolithically integrated structure.

4. A multi-junction photovoltaic device as claimed in any preceding claim, wherein the multi-junction photovoltaic device has a tandem structure consisting of the first sub cell and the second sub-cell.

5. A multi-junction photovoltaic device as claimed in any preceding claim in which the first sub cell is provided with a light transmissive front electrode (100, 101) and the second sub-cell is provided with a textured surface to promote scattering of light after it has passed through the first then the second sub-cell.

6. A multi-junction photovoltaic device as claimed in any preceding claim in which the second sub-cell is provided with a planar, non-textured surface (120) adjacent the first sub-cell.

7. A multi-junction photovoltaic device as claimed in claim 2 in which the tunnel recombination junction comprises a layer comprising a p-type semiconductor (107).

8. A multi-junction photovoltaic device as claimed in claim 7 in which the tunnel recombination junction further comprises a transparent conducting oxide.

9. A multi-junction photovoltaic device as claimed in claim 7 in which the tunnel recombination junction further comprises an insulating passivation layer (108) a layer comprising a p-type semiconductor (107), and a layer comprising an n-type semiconductor (106).

10. A multi-junction photovoltaic device as claimed in claim 9 in which the insulating passivation layer (108) comprises silicon dioxide or undoped silicon.

11. A multi-junction photovoltaic device as claimed in claim 7 in which the tunnel recombination junction comprises a dielectric stack having via holes between the p doped crystalline substrate (104) and the p-type semiconductor layer.

12. A multi-junction photovoltaic device as claimed in any preceding claim in which the perovskite sub-cell comprises photoactive region comprising a compact layer of perovskite material without open porosity that is disposed between the n-type region and the p-type region and that forms a planar heterojunction with one or both of the n- type region and the p-type region; wherein the perovskite material is of general formula

[A][B][X]s (I)

wherein [A] is one or more organic cations selected from methylammonium (CH₃NH₃ ⁺), formamidinium (HC(NH2)2⁺), and ethyl ammonium (CH₃CH2NH₃ ⁺), and/or one or more more inorganic cations selected from Cs⁺, Rb⁺, Cu⁺, Pd⁺, Pt⁺, Ag⁺, Au⁺, Rh⁺, and Ru⁺, [B] is one or more divalent inorganic cations, and [X] is one or more halide anions.

13. A multi-junction photovoltaic device as claimed in any preceding claim in which the perovskite sub-cell comprises photoactive region comprising a compact layer of perovskite material without open porosity that is disposed between the n-type region and the p-type region and that forms a planar heterojunction with one or both of the n- type region and the p-type region; wherein the perovskite material is of general formula (IA):

AxAVxB(X_yXV_y)₃ (IA)

wherein A is a formamidinium cation (FA), A’ is a caesium cation (Cs⁺), B is Pb²⁺, X is iodide and X’ is bromide, and wherein 0 < _x £ 1 and 0 < _y £ 1.

14. A method of fabricating a multi-junction photovoltaic device comprising a perovskite sub-cell in electrical contact with a silicon sub-cell, the method comprising:

a. forming an electrically insulating passivation layer over a silicon sub-cell comprising a p-type silicon substrate, said passivation layer having a plurality of openings, b. forming a perovskite sub-cell over the passivation layer, wherein there is electrical contact between the silicon sub-cell and the perovskite sub-cell through the plurality of openings.

15. A method as claimed in claim 14 in which step a. comprises selective deposition of a patterned passivation layer.

16. A method as claimed in claim 14 in which step a. comprises blanket deposition of a layer of passivation material followed by selective removal of the passivation material in specific areas to form openings.

17. A method as claimed in claim 16 in which the selective removal comprises laser ablation, or depositing a photo-patternable resist followed by patterning of the resist then wet or dry etching.

18. A method as claimed in any one of claims 14 to 17 in which step b. comprises depositing the perovskite layers by vacuum deposition.

19. A method as claimed in any one of claims 14 to 17 in which step b. comprises depositing the perovskite layers by vacuum deposition of a precursor followed by exposure to a vapour.

20. A method as claimed in any one of claims 14 to 17 in which step b. comprises depositing the perovskite layers by wet chemical methods from a solution or suspension.