WO2020260891A1 - Semi-transparent perovskite films - Google Patents

Abstract

This invention relates to semi-transparent perovskite photoactive films, and more particularly, to semi-transparent perovskite films containing microgels. The present invention also relates to processes for the preparation of these films and to their use in perovskite solar cells.

Description

Semi-Transparent Perovskite Films

[0001] This invention relates to semi-transparent perovskite photoactive films, and more particularly, to semi-transparent perovskite films containing microgels. The present invention also relates to processes for the preparation of these films and to their use in perovskite solar cells.

BACKGROUND

[0002] Perovskite solar cells (PSCs) continue to generate enormous research interest in the field of renewable energy generation, because of their meteoric rise in p ower conversion efficiencies (PCEs), thereby offering the potential for low-carbon energy generation. PSCs contain a light harvesting photoactive layer, the layer being formed from a compound with an ABX₃ perovskite crystal structure.

[0003] Hybrid organic-inorganic halide perovskites are an important class of perovskite compounds. This is due to the near optimum balance of perovskite material properties that suit their use as light harvesting layers in solar cells. These properties include panchromatic absorption (Kazim et al., Angewandt. Chem. Int. Ed., 2014, 53, 2812-2824), low exciton binding energies (Miyata et al., Nat. Phys., 2015, 11 , 582), high exciton diffusion lengths (Stranks et al., Science, 2013, 342, 341-344) and defect tolerant performance (De Marco et al., Nano Lett., 2016, 16, 1009-1016). Pb-based perovskites provide the highest power conversion efficiencies and are based on earth-abundant materials (Frost et al., Acc. Chem. Res. 2016, 49, 528-535).

[0004] PSCs are also attractive because they are defect tolerant (Akkerman et al., Nature Mater. 2018, 17, 394-405) which has enabled a wide range of additives to be used whilst still achieving viable PCEs (Zhang et al., Adv. Mater. 2019, 31 , 1805702). Additives that interact with Pb have been added to perovskite precursor solutions to increase grain size (Lee et al., Acc. Chem. Res. 2016, 49, 311-319) by decreasing the nucleation rate (Han et al., Solar RRL 2018, 2, 1800054). These species have also included solvents which form Pb adducts and are removed by heating (Ahn et ai, J. Amer. Chem. Soc. 2015, 137, 8696- 8699). The crystallization rate may also be slowed by increasing solution viscosity and using higher boiling point co-solvents. These approaches have mostly involved small molecules or polymers where the additive affects the whole solution uniformly. PSCs containing high proportions of micrometer-sized swellable microgel particles have been reported recently (Dokkhan et al., Phys. Chem. Chem. Phys. 2018, 20, 27959-27969), wherein microgels are crosslinked polymer colloid particles that swell in a thermodynamically good solvent (Saunders et al., Adv. Coll. Inter†. Sci., 1999, 80, 25). Inclusion of these relatively large colloidal additives was found to give viable devices.

[0005] An area of future energy generation is building-integrated solar cells, which involve semi-transparent solar cells for windows or interior walls. Hence, there is considerable interest in establishing semi-transparent (ST) PSCs, which offer great potential to reduce the carbon footprint of buildings. Neutral colour ST PSCs have been prepared with microstructured perovskite films (Eperon et al., ACS Nano, 2014, 8, 591-598). Because ST PSCs have a target application of power windows, aesthetic requirements are important. The minimum average visible transmittance (AVT) should be 25% for such applications (Chen et al. , Energy Environ. Sci. 2012, 5, 9551-9557). Unfortunately, the dual goals of high AVT and good PCE require compromise because they show a negative correlation. Whilst decreasing the thickness of PSC has been widely used to increase the AVT, retaining a uniform pinhole-free morphology is challenging. Uncontrolled pinhole formation is a common problem for such films that obstructs development, especially for large area devices. This problem has led to the addition of linear polymer additives, such as PVDF- HFP (Zhang et al., Sol. Energy Mater. Sol. Cells 2017, 170, 178-186). A problem with linear polymer additives is that it is difficult to identify and control their location within the resulting perovskite/polymer composites.

[0006] It is therefore an object of the present invention to obviate or mitigate one or more of the problems of the prior art, whether identified herein or elsewhere. In particular, the present invention seeks to provide semi-transparent PSCs having both good AVT and PCE.

SUMMARY OF THE DISCLOSURE

[0007] The present invention provides novel semi-transparent perovskite films comprising sub-micrometer microgel particles within the photoactive perovskite layer.

[0008] The use of sub-micrometer microgel particles to prepare thin perovskite films (less than 100 nm thickness) has resulted in films with AVT values greater than 25%. Surprisingly, it has been discovered that the sub-micrometer microgel particles have major effects on the perovskite morphology and passivation and were found to delay perovskite crystallization, resulting in a more uniform perovskite film being deposited with fewer pinholes than equivalent thin films prepared in the absence of microgels. When thin semi transparent perovskite films according to the present invention were incorporated into PSCs, the devices were found to have improved PCE values when compared to PSCs constructed using ST microgel-free perovskite films.

[0009] Therefore, in a first aspect, the present invention provides a photoactive layer for a perovskite solar cell, the layer comprising: a hybrid inorganic-organic perovskite of formula ABX₃, wherein:

A is Ci-6alkyl-NH3⁺ and optionally also includes one or more of Cs⁺, Rb⁺, guanidinium and formamidinium;

B is selected from Pb²⁺ Ba²⁺, and Sn²⁺; and

X is selected from one or more of Br, Cl and I ;

provided that A and B balance the X charge, so that overall A is singly- charged and B is doubly-charged; and

a plurality of microgel particles formed from a hydrophilic crosslinked polymeric material capable of swelling in polar aprotic solvents,

wherein the z-average diameter of the swollen microgel particles in a polar aprotic solvent is less than 1000 nm; and

the photoactive layer has a thickness of less than 100 nm.

[0010] In another aspect, the present invention provides a method of forming a photoactive layer as described herein, comprising the steps of:

a) swelling particles of the microgel in a polar aprotic solvent so that the z-average diameter of the swollen microgel particles in the polar aprotic solvent is less than 1000 nm;

b) adding hybrid inorganic-organic perovskite precursors to the dispersion of swollen microgel particles from step a);

c) coating the dispersion from step b) onto a substrate; and

d) evaporating the solvent;

wherein the dispersion resulting from step b) comprises the microgel particles at a concentration of 0.1 to 5 % w/w and the hybrid inorganic-organic perovskite precursors at a concentration of 2 to 20 % w/w.

[0011] In a further aspect, the present invention provides a perovskite solar cell comprising a photoactive layer as described herein.

[0012] The above and further aspects of the invention are described in further detail herein.

BRIEF DESCRIPTION OF THE DRAWINGS

[The code used to denote the films referred to below is (x, y) where x and y are the concentrations of MAPbh-_zCl_z and MG (in wt%) respectively used to prepare the films. Note that MA is CH₃NH₃ ⁺].

[0013] Particular embodiments of the invention are further described hereinafter with reference to the accompanying drawings, in which: Figure 1 shows (A) size distribution measured using dynamic light scattering (DLS) for the MGs prepared according to Method 2 dispersed in ethanol and DMSO; (B) and (C) SEM images of the MGs deposited from ethanol and DMSO respectively (scale bars are 5 mhi).

Figure 2 shows X-ray scattering patterns for thin (x, y) perovskite films containing MGs ((20, 2.4) and (10, 1.2)) and the corresponding films without MGs ((20, 0) and (10, 0)).

Figure 3 shows (A) - (E) SEM images showing the morphology of the thin (x, y) films with decreasing perovskite precursor (x) and MG (y) concentrations. The arrows highlight MGs, which appear dark or almost transparent; (F) Thickness of various perovskite/MG films; (G) Depiction of proposed morphological changes as the films became thinner; (H) and (I) show SEM images of MG-free controls. The scale bar values shown in the first row apply to all images.

Figure 4 shows (A) Peak-force tapping mode AFM image for the thin (5.0, 0.6) perovskite film; (B) Line profile across two sites previously occupied by swollen MG particles for the same film. The numbers show the regions of the AFM image measured.

Figure 5 shows digital photographs of the formation of (A) (20, 2.4) and (B) (20, 0) films on glass when heated at 100 °C on a hotplate for the times shown. The film dimensions were 15 mm x 20 mm.

Figure 6 shows XPS core-level (A) N 1s spectra measured for MGs and (39, 0) and (10, 1.2) films; (B) Pb 4f spectra measured for (39, 0) and (10, 1.2) films. The arrows in (A) show the binding energy shifts. The arrows in (B) show that binding energy shifted to lower values when MG was present.

Figure 7 shows (A) PL spectra and (B) variation of maximum PL intensity with x for various (x, y) thin films. The small peaks at around 820 nm in (A) are due to light scattering.

Figure 8 shows (A) Transmittance spectra for the thin (x, y) films; (B) AVTs for the films determined from the spectra in (A). Digital photographs for the films (15 mm x 20 mm) are shown as well as a photo of a building through the (10, 1.2) film.

Figure 9 shows (A) PL spectra for the thin films; (B) Variation of the wavelength at maximum PL intensity with x for the films.

Figure 10 shows (A) Representative SEM image and (B) transmittance spectra for (10, 1 2)^mes° The arrow in (A) highlights a MG. The AVT from the data in (B) was 40.7%. Figure 11 shows device architecture for (A) a planar solar cell of the present invention without a mesoporousTiC>2 layer; and (B) solar cell of the present invention comprising a thin layer of mesoporous PO2 layer.

Figure 12 shows (A) J-V data for (5.0, 0) and (10, 1.2) devices (device data for the (10, 1.2) device are also shown); (B) J-V data for (5.0, 0), (10, 1.2) and (10, 1.2)^meso devices (device data for the (10, 1.2)^meso device are also shown); (C) Time-dependent current density for (5.0, 0), (10, 1.2) and (10, 1.2)^meso solar cells; (D) Box charts and PCE distributions for the (5.0, 0), (10, 1.2) and (10, 1.2)^meso solar cells.

DETAILED DESCRIPTION

Thin perovskite layers

[0014] The present invention provides a photoactive layer for a perovskite solar cell, the layer comprising:

a hybrid inorganic-organic perovskite of formula ABX₃, wherein:

A is Ci-6alkyl-NH3⁺ and optionally also includes one or more of Cs⁺, Rb⁺, guanidinium and formamidinium;

B is selected from Pb²⁺, Ba²⁺ and Sn²⁺; and

X is selected from one or more of Br, Cl and h;

provided that A and B balance the X charge, so that overall A is singly-charged and B is doubly-charged; and

a plurality of microgel particles formed from a hydrophilic crosslinked polymeric material capable of swelling in polar aprotic solvents;

wherein the z-average diameter of the swollen microgel particles in a polar aprotic solvent is less than 1000 nm; and

the photoactive layer has a thickness of less than 100 nm.

[0015] Photoactive layers are used in solar cells to absorb light. In perovskite solar cells (PSCs), a photoactive layer includes a perovskite-structured material with a crystal structure of general formula ABX₃. Hybrid inorganic-organic perovskite compounds are a major class of compounds used as photoactive layers in PSCs.

[0016] The photoactive layer of the present invention comprises a hybrid inorganic-organic perovskite of formula ABX₃, as described herein. A and B must balance the X charge, so that overall A is singly-charged and B is doubly-charged. In other words, there are three X anions, so to balance the charge, overall, even though A and B may be combinations of different cations, A must be A⁺ and B must be B²⁺. For example, A is Ci-6alkyl-NH3⁺ and B is (Sn²⁺)o.₃( Pb²⁺)₀₇; or A is (Cs⁺Ks C^alkyl-NHs⁺Vs and B is Pb²⁺; or A is (CH₃NH₃ ⁺)o.i5 (HC(NH₂)=NH₂ ⁺)O.85 and B is Pb²⁺

[0017] In an embodiment, A is Ci-6alkyl-NH3⁺. Ci-₆alkyl refers to a branched or unbranched alkyl chain containing between 1 and 6 carbon atoms. In an embodiment, Ci-₆alkyl is methyl. In an embodiment, A is CH3NH3⁺. Formamidinium refers to the protonated form of formamidine. Guanidinium refers to the protonated form of guanidine.

[0018] In an embodiment, B is Pb²⁺.

[0019] In an embodiment, X is selected from one or more of Br and I . In an embodiment, X is selected from one or more of Br and Cl . In an embodiment, X is selected from one or more of Cl and I . In an embodiment, X is a combination of Cl and I .

[0020] In an embodiment, A is Ci-6alkyl-NH3⁺ and B is Pb²⁺. In an embodiment, A is CH₃NH₃ ⁺ and B is Pb²⁺. In an embodiment, A is Ci-6alkyl-NH3⁺, B is Pb²⁺ and X is selected from one or more of Cl and I , preferably a combination of Cl and I . In an embodiment, A is CH₃NH₃ ⁺, B is Pb²⁺ and X is selected from one or more of Cl and I , preferably I . In an embodiment, the photoactive layer comprises a hybrid inorganic-organic perovskite of formula (CH₃NH₃ ⁺)(Pb²⁺)(^|-)3.

[0021] In an embodiment, A is a mixture of Ci-6alkyl-NH3⁺ (such as CH3NH3⁺) and formamidinium and B is Pb²⁺. In an embodiment, A is a mixture of Ci-6alkyl-NH3⁺ (such as CH₃NH₃ ⁺) and formamidinium, B is Pb²⁺ and X is a combination of Br and I . In an embodiment, the photoactive layer is a hybrid inorganic-organic perovskite of formula [(CH₃NH3⁺)(Pb²⁺)(Br-)₃]p [(HC(NH₂)=NH₂ ⁺)(Pb²⁺)(l )₃]_q, wherein p + q = 1. In an embodiment, p is 0.05 to 0.25 and q is 0.75 to 0.95. In an embodiment, p is 0.15 and q is 0.85.

[0022] The hybrid inorganic-organic perovskites are typically prepared by mixing perovskite precursors in a suitable solvent. A suitable solvent, or solvent system, is one in which the precursors dissolve. Suitable solvents include polar aprotic solvents, such as dimethyl formamide (DMF) or dimethyl sulfoxide (DMSO). Suitable solvents may be selected from g- butyrolactone, dimethyl formamide, dimethyl sulfoxide, or a combination of these solvents. In an embodiment, the hybrid inorganic-organic perovskite is formed from perovskite precursors. The hybrid inorganic-organic perovskite precursors are compounds which, when combined, are capable of forming a hybrid inorganic-organic perovskite of formula ABX3, as defined herein. In a preferred embodiment, the precursors are of the formula AX and BX₂. In a preferred embodiment, the precursors are Ci-6alkyl-NH3X and PbX₂, such as CH3NH3X and PbX₂ (for example, CH3NH3I and PbCI₂). In an embodiment, the hybrid inorganic-organic perovskite is of formula (CH3NH3⁺)(Pb²⁺)(l )3-_z(CI )_z where z is 0 to 3; and is formed from the precursors CH3NH3I and PbCI₂. In an embodiment, the hybrid inorganic- organic perovskite is of formula (CH3NH3⁺)(Pb²⁺)(l )3; and is formed from the precursors CH3NH3I, CH3NH3CI and Pb . As further encompassed by the present invention, the hybrid inorganic-organic perovskite may be a mixed perovskite and the precursors of formulae AX and BX2 may comprise several components. For example, when the photoactive layer is a hybrid inorganic-organic mixed perovskite of formula [(CH3NH3⁺)(Pb²⁺)(Br)3] [(HC(NH2)=NH2⁺)(Pb²⁺)(l )3]q, it may be formed from the precursors CHsNHsBr, formamidinium iodide, PbBr2 and Pb .

[0023] In the present invention, microgels are used as an additive for hybrid inorganic- organic perovskite photoactive layers. The microgels act as pore-forming agents around which the hybrid inorganic-organic perovskite crystallises during layer deposition. In contrast to some previously reported methods to prepare porous perovskite layers (Meng et al., Nano Lett., 2016, 16, 4166-4173; Zhou et al., Adv. Mater., 2017, 29, 170368) where the additive was removed, the photoactive perovskite layers of the present invention are prepared via a single-step deposition method without the need to subsequently remove the microgel particles.

[0024] In order to prepare the porous photoactive layers according to the present invention, microgels were used which dispersed in solvents suitable for hybrid inorganic-organic perovskite preparation (polar aprotic solvents such as g-butyrolactone, dimethyl formamide, and dimethyl sulfoxide), without dissolving. It is important that the microgels are also capable of swelling in polar aprotic solvents. ‘Good’ solvents for swelling microgels as described herein are therefore polar aprotic solvents such as g-butyrolactone, dimethyl formamide, and dimethyl sulfoxide.‘Poor’ solvents do not swell the microgels and are polar protic solvents such as ethanol or methanol, or non-polar solvents such as toluene, hexane or diethyl ether.

[0025] Accordingly, the photoactive layer of the present invention further comprises a plurality of microgel particles formed from a hydrophilic crosslinked polymeric material capable of swelling in polar aprotic solvents.

[0026] The crosslinked polymeric material should be hydrophilic to enable it to be dispersed and swelled in polar aprotic solvents (such as g-butyrolactone, dimethyl formamide, and dimethyl sulfoxide). Hydrophilic polymers may result from the co-polymerisation of monomers bearing hydrophilic moieties (such as hydrogen bond donor/acceptor moieties). In an embodiment, the microgel particles are prepared by non-aqueous dispersion polymerisation. Preferably, the microgel particles are prepared by non-aqueous dispersion polymerisation of monomers bearing hydrophilic moieties. Preferably the monomers are vinyl monomers bearing hydrophilic moieties. Preferably, the polymerisation is a free-radical co-polymerisation initiated via a suitable free-radical source such as AIBN. In an embodiment, the polymerisation is carried out in the presence of additional vinyl co-polymer stabilisers, such as polyvinylpyrrolidone/polyvinyl acetate co-polymer (PVP-co-PVA).

[0027] The swelling capability of the polymerized microgel particles can be assessed by comparing the z-average diameters ( d_z ) of the microgel particles (MGs) dispersed in a poor solvent (e.g. ethanol) and a good solvent (e.g. DMSO) via a suitable technique such as dynamic light scattering. This comparison can be seen in Fig. 1A, where the d_z values for the microgel particles in ethanol and DMSO were measured at 354 nm and 495 nm respectively. According to this invention, a microgel is capable of swelling in polar aprotic solvents, if the particles have z-average diameters in polar aprotic solvents 1.2 to 100 times greater than the z-average diameters in poor solvents such as ethanol, which do not swell the MGs (this is termed the‘linear swelling ratio’). In an embodiment, the photoactive layer of the present invention comprises a microgel comprising a plurality of microgel particles formed from a hydrophilic crosslinked polymeric material capable of swelling in polar aprotic solvents, with a linear swelling ratio of between 1.2 and 100 compared to the unswollen particles. For the Fig. 1A example the linear swelling ratio was 1.40. In a preferred embodiment, the linear swelling ratio is between 1.2 and 50, such as between 1.2 and 25, between 1.2 and 10, or between 1.2 and 5 compared to the unswollen particles. The MG particle volume swelling ratios can also be estimated from the dynamic light scattering data. For the Fig. 1A example the volume swelling ratio was 2.7. In a preferred embodiment, the volume swelling ratio is between 1.5 and 10, such as between 1.5 and 5, between 1.5 and 3, or between 2 and 3 compared to the unswollen particles. Swollen MGs have good dispersion stability (i.e. they remain separated in dispersion), because the particles have a negligible effective Hamaker constant (Saunders et ai, Adv. Coll. Inter†. Sci., 1999, 80, 25).

[0028] The present invention uses sub-micrometer microgel particles as additives in the preparation of thin perovskite layers. By the term ‘sub-micrometer’ is meant that the z- average diameter of the swollen microgel particles in a polar aprotic solvent is less than 1000 nm (1 mhi). In an embodiment, the z-average diameter of the swollen microgel particles in g-butyrolactone, dimethyl formamide, dimethyl sulfoxide, or a combination of these solvents is less than 1000 nm. In a preferred embodiment, the z-average diameter of the swollen microgel particles in dimethyl sulfoxide is less than 1000 nm. In an embodiment, the z-average diameter of the swollen microgel particles in a polar aprotic solvent is less than 750 nm. In an embodiment, the z-average diameter of the swollen microgel particles in g-butyrolactone, dimethyl formamide, dimethyl sulfoxide, or a combination of these solvents is less than 750 nm. In a preferred embodiment, the z-average diameter of the swollen microgel particles in dimethyl sulfoxide is less than 750 nm. In an embodiment, the z- average diameter of the swollen microgel particles in a polar aprotic solvent is less than 500 nm. In an embodiment, the z-average diameter of the swollen microgel particles in g- butyrolactone, dimethyl formamide, dimethyl sulfoxide, or a combination of these solvents is less than 500 nm. In a preferred embodiment, the z-average diameter of the swollen microgel particles in dimethyl sulfoxide is less than 500 nm.

[0029] In an embodiment, the microgel particles comprise a co-polymer of monomers (I) and (II):

wherein:

Y is selected from:

Z is selected from one of the following linkers:

R¹ , R², R³, R⁴, R⁵, R⁶, R⁷, R⁸ and R⁹ are independently selected from hydrogen and Ci-3alkyl;

R¹⁰ and R¹¹ are independently selected from hydrogen and Ci-3alkyl; or

R¹⁰ and R¹¹ are taken together with the moieties to which they are attached to form a

4- to 9-membered lactam; R¹² is selected from OH, NR¹³R¹⁴ and -(OCH₂CH₂)_P-OR¹⁵;

R¹³, R¹⁴ and R¹⁵ are independently selected from hydrogen and Ci-3alkyl;

U and L² are independently selected from divalent alkyl, divalent alkylether, divalent alkylamine, divalent alkylamide and divalent alkylester linker groups;

Ar is a divalent optionally substituted aryl or heteroaryl group

n is 1 to 20; and

p is 2 to 20.

[0030] In an embodiment, there is provided a microgel comprising a co-polymer of monomers (I) and (II), as described herein, wherein Y is:

ally R¹⁰ and R¹¹ are both hydrogen.

nt monomer (I) is:

[0032] In an embodiment monomer (I) is:

are hydrogen.

[0033] In an embodiment, there is provided a microgel comprising a co-polymer of monomers (I) and (II), as described herein, wherein Y is:

and R¹⁰ and R¹¹ are taken together with the moieties to which they are attached to form a 4- to 9-membered lactam.

[0034] In an embodiment monomer (I) is: [0035] In an embodiment monomer (I) is:

are hydrogen.

[0036] In an embodiment, there is provided a microgel comprising a co-polymer of monomers (I) and (II), as described herein, wherein Y is:

R¹²

, and optionally R¹² is NR¹³R¹⁴. In a preferred embodiment R¹³ and R¹⁴ are both hydrogen. In a preferred embodiment R¹³ is hydrogen and R¹⁴ is isopropyl.

[0037] In an embodiment monomer (I) is:

are hydrogen.

[0038] In an embodiment monomer (I) is:

are hydrogen.

[0039] In an embodiment monomer (I) is:

[0040] In an embodiment monomer (I) is:

methyl and R² and R³ are hydrogen.

[0041] In an embodiment monomer (I) is:

hydrogen or methyl and R² and R³ are hydrogen.

[0042] In an embodiment, there is provided a microgel comprising a co-polymer of monomers (I) and (II), as described herein, wherein Z is selected from:

[0043] In an embodiment, Z is:

U is selected from divalent alkyl, divalent alkylether, divalent alkylamine, divalent alkylamide and divalent alkylester linker groups.

[0044] As used herein‘divalent alkyl’ group refers to a branched or unbranched, optionally substituted 1 to 20 carbon alkylene linker. Examples of suitable divalent alkyl groups include methylene (CH₂) ethylene (CH₂CH₂), propylene or butylene.

[0045] As used herein ‘divalent alkylether’ group refers to a branched or unbranched, optionally substituted 1 to 20 carbon alkylene linker wherein one or more carbon atoms have been replaced by oxygen atoms. Examples of suitable divalent alkylether groups include -CH₂OCH₂-, -CH₂CH₂OCH₂CH₂-, -CH₂OCH₂CH₂- and -CH₂OCH₂CH₂OCH₂-. [0046] As used herein ‘divalent alkylamine’ group refers to a branched or unbranched, optionally substituted 1 to 20 carbon alkylene linker wherein one or more carbon atoms have been replaced by nitrogen atoms. Examples of suitable divalent alkylamine groups include -CH₂NHCH₂-, -CH₂CH₂NHCH₂CH₂-, -CH₂NHCH₂CH₂- and -CH₂NHCH₂CH₂NHCH₂-. The N atoms may be optionally substituted with Ci-3alkyl groups.

[0047] As used herein ‘divalent alkylamide’ group refers to a branched or unbranched, optionally substituted 1 to 20 carbon alkylene linker wherein one or more carbon atoms have been replaced by amide moieties (-NHC(O)- or -C(O)NH-). Examples of suitable divalent alkylamide groups include -CH₂-C(0)NH-CH₂-, -CH₂CH₂-C(0)NH-CH₂CH₂- and - CH₂-NHC(0)CH₂CH₂-. The N atoms may be optionally substituted with Ci-3alkyl groups.

[0048] As used herein ‘divalent alkylester’ group refers to a branched or unbranched, optionally substituted 1 to 20 carbon alkylene linker wherein one or more carbon atoms have been replaced by ester moieties (-OC(O)- or -C(O)O-). Examples of suitable divalent alkylester groups include -CH₂-C(0)0-CH₂-, -CH₂CH₂-C(0)0-CH₂CH₂- and -CH₂-

0C(0)CH₂CH₂-.

[0049] In an embodiment, Z is:

divalent alkyl or divalent alkylether linker group, preferably U is a divalent alkylether linker group.

[0050] In an embodiment, Z is

[0051] In an embodiment monomer (II) is:

R⁴ R⁵ R⁶, R⁷, R⁸ and R⁹ are hydrogen.

[0052] In an embodiment, there is provided a microgel comprising a co-polymer of monomers (I) and (II), as described herein, wherein Z is: selected from divalent alkyl, divalent alkylether, divalent alkylamine, divalent alkylamide and divalent alkylester linker groups.

[0053] In an embodiment, Z is:

divalent alkyl or divalent alkylether linker group, preferably L¹ is a divalent alkyl linker group.

[0054] In an embodiment, there is provided a microgel comprising a co-polymer of monomers (I) and (II), as described herein, wherein Z is:

[0055] In an embodiment monomer (II) is:

hydrogen.

[0056] In an embodiment, Z is:

20, such as 1-5, or preferably n is 1.

[0057] In an embodiment monomer (II) is:

are hydrogen and R⁶ and R⁷ are methyl. [0058] In an embodiment, there is provided a microgel comprising a co-polymer of monomers (I) and (II), as described herein, wherein Z is:

and R¹⁶ is halo, nitro, cyano, hydroxy or CO2H (preferably nitro).

[0059] In an embodiment, the microgel particles comprise a co-polymer of one or more monomers of formula (I) and one or more monomers of formula (II), as described herein.

[0060] In an embodiment, the microgel particles comprise a co-polymer of:

This co-polymer is poly(/\/-vinylformamide-co-2-(/\/-vinylformamido)ethyl ether) (PNVF- NVEE).

[0061] In an embodiment, the microgel particles comprise a co-polymer of:

This co-polymer is poly(/\/-vinylcaprolactam-co-/\/,/\/-methylenebis(acrylamide)) (PNVC-BA).

[0062] In an embodiment, the microgel particles comprise a co-polymer of:

This co-polymer is poly(acrylamide-co-/\/,/\/-methylenebis(acrylamide)) (PA-BA).

[0063] In an embodiment, the microgel particles comprise a co-polymer of: This co-polymer is poly(/\/-isopropylacrylamide-co-/\/,/\/ -methylenebis(acrylamide)) (PNIPAM-BA).

[0064] In an embodiment, the microgel particles comprise a co-polymer of:

wherein n and p are as defined herein. This co-polymer is poly(polyethylene glycol methacrylate-co-polyethylene glycol dimethacrylate) (PPEGMA-PEDGMA).

[0065] Wherein the sub-micrometer microgel particles are formed from the polymerisation of vinyl monomers in the presence of an additional vinyl co-polymer stabilizer (such as PVP- co-PVA) as described herein, the ratio of the total mmols of monomers relative to the mass of the vinyl co-polymer stabilizer is less than 50:1. In an embodiment, the ratio of the total mmols of monomers relative to the mass of the vinyl co-polymer stabilizer is less than 40:1 , such as less than 30:1.

[0066] The present invention relates to the preparation of thin, semi-transparent perovskite photoactive layers. The thickness of the photoactive perovskite layer as described herein should be less than 100 nm, in order to be suitable for ST PSCs. In a preferred embodiment, the photoactive layer has a thickness of less than 75 nm, such as less than 50 nm. In an even more preferred embodiment, the photoactive layer has a thickness of less than 20 nm.

[0067] In an embodiment, the photoactive perovskite layer as described herein has an average visible transmittance (AVT) of at least 25%. In a preferred embodiment, the photoactive perovskite layer as described herein has an AVT of at least 30%, such as at least 35%, or at least 40%.

Formation of perovskite layers

[0068] In an aspect of the present invention, there is provided a method of forming a photoactive layer as described herein, comprising the steps of: a) swelling particles of the microgel in a polar aprotic solvent so that the z- average diameter of the swollen microgel particles in the polar aprotic solvent is less than 1000 nm;

b) adding hybrid inorganic-organic perovskite precursors to the dispersion of swollen microgel particles from step a);

c) coating the dispersion from step b) onto a substrate; and

d) evaporating the solvent;

wherein the dispersion resulting from step b) comprises the microgel particles at a concentration of 0.1 to 5 % w/w and the hybrid inorganic-organic perovskite precursors at a concentration of 2 to 20 % w/w.

[0069] In step a) microgel particles as described herein, are swollen in a polar aprotic solvent (a thermodynamically ‘good’ solvent for the microgels). In an embodiment, the swollen microgel particles are 1.2-100 times the size of the unswollen particles. In an embodiment, the linear swelling ratio of the microgel particles is in the range 1.2 to 100. In an embodiment, in step a) the particles are swollen to 1.2 to 50 times, such as 1.2 to 25, 1.2 to 10, or 1.2 to 5 times the size of the unswollen particles. In an embodiment, the linear swelling ratio of the microgel particles is in the range 1.2 to 50, such as 1.2 to 25, 1.2 to 10, or 1.2 to 5. Alternatively, step a) comprises swelling particles of the microgel in a polar aprotic solvent to 1.5 to 10 times the volume (this is the volume swelling ratio) of the unswollen particles. Preferably, the volume swelling ratio of the microgel particles is between 1.5 and 5, between 1.5 and 3, or between 2 and 3, compared to the unswollen particles. In an embodiment, the solvent is selected from g-butyrolactone, dimethyl formamide, dimethyl sulfoxide, or a combination of these solvents. In a preferred embodiment, the solvent is dimethyl sulfoxide.

[0070] In step b) hybrid inorganic-organic perovskite precursors are added to the dispersion of swollen microgel particles from step a), so that after deposition the hybrid inorganic- organic perovskite crystallises on the substrate. The hybrid inorganic-organic perovskite precursors are compounds which, when combined, are capable of forming a hybrid inorganic-organic perovskite of formula ABX₃, as defined herein. In a preferred embodiment, the precursors are of the formula AX and BX₂, wherein A, B and X are as defined herein. For example, the perovskite precursors may be Ci-₆alkyl-NH₃X and PbX₂, such as CH₃NH₃X and PbX₂ (for example, CH₃NH₃I and Pbh).

[0071] In a preferred embodiment, steps a) and b) are carried out concurrently. In other words, the microgel particles and the perovskite precursors are dispersed in a polar aprotic solvent in a single step. [0072] In order to prepare the thin, semi-transparent photoactive layers according to the present invention low concentrations of both perovskite precursors and microgels are employed. More specifically, the dispersion resulting from step b) comprises the microgel particles at a concentration of 0.1 to 5 % w/w (y) and the hybrid inorganic-organic perovskite precursors at a concentration of 2 to 20 % w/w (x). The balance of the dispersion is typically made up of one or more polar aprotic solvents, such as g-butyrolactone, dimethyl formamide, or dimethyl sulfoxide, preferably dimethyl sulfoxide. In an embodiment, the dispersion resulting from step b) comprises the microgel particles at a concentration of 0.1 to 5 % w/w and the hybrid inorganic-organic perovskite precursors at a concentration of 2 to 15 % w/w. In an embodiment, the dispersion resulting from step b) comprises the microgel particles at a concentration of 0.2 to 5 % w/w and the hybrid inorganic-organic perovskite precursors at a concentration of 3 to 15 % w/w. In an embodiment, the dispersion resulting from step b) comprises the microgel particles at a concentration of 0.5 to 2.5 % w/w and the hybrid inorganic-organic perovskite precursors at a concentration of 5 to 15 % w/w. In an embodiment, the dispersion resulting from step b) comprises the microgel particles at a concentration of 0.3 to 3 % w/w and the hybrid inorganic-organic perovskite precursors at a concentration of 4 to 14 % w/w. In an embodiment, the dispersion resulting from step b) comprises the hybrid inorganic-organic perovskite precursors and the microgel particles at a combined concentration (x + y) of 5 to 20 % w/w, such as 7 to 15 % w/w.

[0073] The nominal weight fraction of microgel particles ( M/MG) in the thin layers can be calculated, wherein M MG = 100y/(x + y). In an embodiment, the M/MG of the photoactive layer is 2-40 %, such as 5-15 %, 8-12 %, or 10-11 %.

[0074] Step c) involves deposition of the perovskite precursor/microgel dispersion onto a substrate. A suitable substrate for forming the porous photoactive layer on, is any layer which may be conventionally used in the production of solar cells and which is stable to the solvent used in step a). Preferably, the substrate is a T1O2 hole-blocking layer (bl-TiC>2) or a mesoporous T1O2 layer (mp-TiC>2).

[0075] A skilled person may envisage various deposition methods suitable for achieving the coating in step c). Such methods may include casting, doctor blading, screen printing, inkjet printing, pad printing, knife coating, meniscus coating, slot die coating, gravure coating, reverse gravure coating, kiss coating, micro-roll coating, roll-to-roll coating, curtain coating, slide coating, spin coating, spray coating, flexographic printing, offset printing, rotatory screen printing, evaporative coating or dip coating. In a preferred embodiment, the coating in step c) is carried out by spin coating.

[0076] In step d) the solvent is evaporated, either passively (under ambient conditions) or actively (e.g. via the application of heat and/or vacuum to the coated dispersion). This process is also referred to as annealing. In an embodiment, step d) is carried out at 80 to 200 °C. In a preferred embodiment, step d) is carried out at 90 to 110 °C. In an embodiment, step d) is carried out for between 5 minutes and 4 hours. In a preferred embodiment, step d) is carried out for between 5 and 15 minutes (such as 10 minutes).

[0077] Solvent evaporation drives crystallization of the hybrid inorganic-organic perovskite. To enhance perovskite crystallization, an anti-solvent may be added during the coating or evaporating steps. In an embodiment, step c) further comprises the addition of an anti solvent during coating. Preferably the anti-solvent is added at or towards the end of the coating step. An anti-solvent is defined as a solvent in which the hybrid inorganic-organic perovskite, as described herein, has poor solubility. In an embodiment, the anti-solvent is selected from chlorobenzene, benzene, xylene, toluene, methanol, ethanol, ethylene glycol, 2-propanol, chloroform, THF, acetonitrile, and benzonitrile, or a combination thereof. In a preferred embodiment, the anti-solvent is toluene.

[0078] A proposed mechanism by which the thin perovskite layers of the present invention are formed, involves the MGs acting as colloidal sponges and delaying the release of perovskite precursors, thereby delaying perovskite crystallization. This results in thin, uniform ST perovskite layers with few pinholes compared to equivalent thin films prepared in the absence of MGs. The uniformity of the films of the present invention results in surprisingly high PCEs when incorporated into PSCs, while the inclusion of the sub micrometer MGs within the thin films was unexpectedly found to decrease light scattering and thereby improve the AVT of the films.

[0079] In a further aspect, there is provided a photoactive layer directly obtained by, obtained by, or obtainable by a method as described herein.

Perovskite solar cells

[0080] PSCs can be constructed comprising photoactive layers according to the present invention. Therefore, in an aspect of the invention, there is provided a perovskite solar cell comprising a photoactive layer as described herein.

[0081] Perovskite solar cells may be constructed using processes and techniques familiar to those in the field. A PSC is typically formed from a number of layers selected from one or more of glass, indium tin oxide (ITO), T1O2 hole-blocking layer (bl-Ti0₂), mesoporous T1O2 layer (mp-Ti0₂), perovskite photoactive layer (capping layer), hole transport layer and a top electrode (e.g. gold). Preferably, the PSC layers are deposited on top of one another in the order glass, ITO, T1O2 hole-blocking layer (bl-Ti0₂), mesoporous T1O2 layer (mp-Ti0₂), perovskite photoactive layer, hole transport layer and top electrode. In PSCs comprising photoactive layers according to the present invention, the photoactive layer is the perovskite photoactive layer. In one embodiment, the mesoporous T1O2 layer is omitted from the PSC device architecture, as shown in Fig. 11 A. Accordingly, in an embodiment, there is provided a perovskite solar cell containing the following layers: glass, ITO, T1O2 hole-blocking layer (bl-TiC>2), perovskite photoactive layer, hole transport layer and top electrode; wherein the perovskite photoactive layer is as described herein. In an alternative embodiment, the perovskite solar cell comprises a mesoporous PO2 layer, as shown in Fig. 11 B. Accordingly, in an embodiment, there is provided a perovskite solar cell containing the following layers: glass, ITO, T1O2 hole-blocking layer (bl-Ti0₂), mesoporous T1O2 layer (mp-Ti0₂), perovskite photoactive layer, hole transport layer and top electrode; wherein the perovskite photoactive layer is as described herein. Preferably, the mesoporous T1O2 layer (mp-Ti0₂), when present, has a thickness no greater than 100 nm. In an embodiment, the mesoporous T1O2 layer (mp-Ti0₂), when present, has a thickness less than 100 nm, less than 90 nm, less than 80 nm, less than 70 nm, less than 60 nm, or less than 50 nm.

[0082] In an embodiment the top electrode is gold. In another embodiment, the top electrode is transparent, such as an Mo0₃/Ag/MoC>₃ electrode (described in Cho et ai, ACS Appl. Mater. Interfaces 2019, 20864-20872).

[0083] A process for forming a PSC may comprise coating the perovskite precursor-MG dispersion onto a glass/ITO/bl-TiC>2 substrate, and then applying a hole transport layer, followed by a top electrode (e.g. gold) coating (see Fig. 11 A for a schematic of the solar cell architecture).

[0084] Alternatively, a process for forming a PSC may comprise coating the perovskite precursor-MG dispersion onto a glass/IT0/bl-Ti0₂/mp-Ti0₂ substrate, and then applying a hole transport layer, followed by a top electrode (e.g. gold) coating (see Fig. 11 B for a schematic of this solar cell architecture).

[0085] In an embodiment, the perovskite solar cell is semi-transparent. In an embodiment, the semi-transparent perovskite solar cell as described herein has an average visible transmittance (AVT) of at least 25%. In a preferred embodiment, the semi-transparent perovskite solar cell as described herein has an AVT of at least 30%, such as at least 35% or at least 40%.

[0086] PSCs formed with photoactive layers according to the present invention may have significantly higher power conversion efficiency (PCE) values compared to analogous PSCs prepared with analogous MG-free photoactive layers. In an embodiment, the semi transparent perovskite solar cell as described herein has a PCE greater than 5%, such as greater than 7.5%. In an embodiment, the semi-transparent perovskite solar cell as described herein has an average visible transmittance (AVT) of at least 25% and a PCE greater than 5%. In an embodiment, the semi-transparent perovskite solar cell as described herein has an average visible transmittance (AVT) of at least 40% and a PCE greater than 7.5%.

[0087] Throughout the description and claims of this specification, the words“comprise” and“contain” and variations of them mean“including but not limited to”, and they are not intended to (and do not) exclude other moieties, additives, components, integers or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

EXAMPLES

[0088] The invention will now be described in more detail in relation to the following illustrative examples.

Physical Measurements

[0089] Unless stated otherwise, the following methodology was used to obtain physical measurements.

[0090] Dynamic light scattering (DLS) measurements was obtained using a Malvern Zetasizer Nano ZS instrument (via cumulants analysis). The z-average diameter ( d_z ) is an average value from five runs.

[0091] The top view SEM was obtained using a Philips XL30 FEGSEM.

[0092] AFM images were obtained using either a Bruker Multimode 8 or a Bruker Catalyst; images were captured in ScanAsyst™ (Peak Force Tapping) mode.

[0093] UV-visible spectra were recorded using a Perkin Elmer Lamda 25 UV-Vis spectrometer. The average visible transmittance was measured between 370 and 740 nm.

[0094] Film thickness measurements were conducted using a Dektak 8 Stylus Profilometer (Bruker).

[0095] XRD patterns were conducted using a Bruker D8 Advance diffractometer (Cu-Ka). Films were scanned with a step size of 0.02°. The films were prepared under nitrogen atmosphere and measured using an airtight holder.

[0096] Photoluminescence (PL) spectra were obtained using an Edinburgh Instruments FLS980 spectrometer. The beam was incident on the film surface side and an excitation wavelength of 480 nm was used.

[0097] X-ray photoelectron spectroscopy (XPS) measurements were performed with a SPECS XPS spectrometer, equipped with a monochromated Al K_c X-ray source that can produce X-rays with photon energy of 1486.7 eV. Photoelectrons produced were collected by a 150 mm hemispherical energy analyzer (Phoibos 150 SPECS). Binding energies (BEs) were calibrated to C 1s from CHsNHsPbU (MAPI) at 285.3 eV, or to C 1s from adventitious carbon at 284.8 eV. All XPS spectra were analysed using CasaXPS software. A Shirley background and pseudo-Voigt (GL(30)) function (30% Lorentzian and 70% Gaussian) were applied to fit the spectra acquired. BE values are quoted to an accuracy of ± 0.1 eV due to instrumental precision. To obtain the surface composition of the films, the built-in CasaXPS Kratos sensitivity factors were used to calculate the stoichiometries.

Device Measurements

[0098] Unless stated otherwise, the following methodology was used to obtain device measurements.

[0099] The current density-voltage (J-V) characteristics were measured using a Keithley 2420 Sourcemeter and lOOmWcm² illumination (AM 1.5G) and a calibrated NREL certified Oriel Si-reference cell. An Oriel solar simulator (SOL3A) was used for these measurements. The active area of the devices (0.080 cm²) was determined using a square aperture within a mask. The data shown are from the reverse scan unless otherwise stated (V_oC to _sc) and the sweep rate was 100 mV s ¹.

Materials

[00100] /V-vinylformamide (NVF, 98%), azoisobutyronitrile (AIBN, 98%), potassium-tert- butoxide (95%), bis(2-bromoethyl)ether (BBE, 95%), dicyclohexyl- 18-crown-6 (98%), anhydrous tetrahydrofuran (THF, 99.9%), and ethanol (99.9%) , poly(1-vinylpyrrolidone-co- vinyl acetate) (PVP-co-PVA, M_n -50,000 g/mol) , anhydrous sodium sulphate (100%) , chloroform (99.9%), sodium chloride (NaCI, 100%), toluene (99.8%), chlorobenzene (CBZ, 99.8%), isopropanol (IPA, anhydrous, 99.5%), 4-tert-butylpyridine (TBP, 96%) and lithium bistrifluoromethanesulfonimidate (LiTFSI, 99.95%) were all purchased from Aldrich and used as received. Methyl amine solution (33 wt.% in absolute EtOH) and hydroiodic acid (57 wt.%), titanium diisopropoxide bis(acetylacetonate) (TDB, 75 wt % in IPA), lead (II) chloride (PbCh, 98%) and dimethyl sulfoxide (DMSO, 99.7%) were also purchased from Aldrich and used as received. Methylammonium iodide (MAI) and methylammonium chloride (MAC) were synthesised and purified using the method previously reported (Etgar et a!., J. Amer. Chem. Soc., 2012, 134, 17396-17399). Titania paste (T1O2, 18 NRT) was purchased from Dyesol and used as received. Spiro-MeOTAD (Spiro, L/²,L/²,L/² ,L/² ,L/⁷,L/⁷,L/⁷ , \/⁷-octakis(4-methoxyphenyl)-9,9’-spirobi[9H-fluorene]-2,2’,7,7’- tetramine, Fenglin Chemicals, 99.5%) was also used as received. Water was of ultra-high purity and de-ionised.

Method 1 - Preparation of 2-(/V-vinylformamido)ethyl ether (NVEE)

[00101] NVEE was synthesized in a 250 ml_ reactor equipped with an overhead stirrer. Firstly, a mixture of NVF (7.1 g, 100 mmol), potassium-tert-butoxide (12 g, 105 mmol) and dicyclohexyl-18-crown-6 (1 g, 2.65 mmol) were dissolved in anhydrous THF (100 ml_). Then this mixture was stirred vigorously at room temperature for 45 min and was cooled to 0 °C in an ice bath for 20 min. Bis(2-bromoethyl)ether (9.3 g, 40 mmol) was then added dropwise to the mixture during cooling and the mixture was stirred at room temp for 72 h. After that, KBr was removed from the mixture by filtration and the reaction mixture was concentrated under rotary evaporator and diluted with water (100 ml_). The product was repeatedly extracted with chloroform (5 x 40 ml_) and washed twice with brine and then dried over anhydrous sodium sulphate (40 g) for 24 h. Finally, the product was recovered as a liquid after concentration using rotary evaporation.

Method 2 - Preparation of poly(NVF-co-NVEE) [PNVF-9NVEE] sub-micrometer microgel particles

[00102] NVF-9NVEE sub-micrometer microgel particles were prepared by non-aqueous dispersion polymerization. The MGs nominally contained 9.0 wt% NVEE based on monomer. A mixture of NVF (3.0 g, 42.75 mmol), PVP-co-PVA (1.8 g) and NVEE (0.9 g, 4.16 mmol) were added to EtOH in a four-necked round bottomed flask equipped with overhead stirrer, nitrogen supply and a reflux condenser. The solution was heated to 70 °C and stirred vigorously. Then, AI BN (0.12 g, 0.73 mmol) in EtOH (2.0 ml_) was added to the reaction flask and the polymerization allowed to continue for 1 h. The dispersion was filtered with a 50 pm mesh filter after cooling to 0 °C and was then purified by three centrifugation and re-dispersion steps in EtOH. To transfer the MGs from EtOH to DMSO, the MGs in EtOH were centrifuged and then re-dispersed in DMSO. This process was repeated and finally diluted to 6.0 wt% in DMSO.

[00103] In Method 2 the total mass of monomers (NVF and NVEE) relative to the mass of the vinyl co-polymer stabilizer (PVP-co-PVA) is 2.17:1. The total mmols of monomers (NVF and NVEE) relative to the mass of the vinyl co-polymer stabilizer (PVP-co-PVA) is 26.1:1 The sub-micrometer PNVF-NVEE MGs prepared according to this method were more than a factor of two smaller in diameter than the PNVF-NVEE MGs prepared according to similar methodology in Dokkhan et aL, Phys. Chem. Chem. Phys. 2018, 20, 27959-27969, in which the total mass of monomers relative to the mass of the vinyl co-polymer stabilizer was 4.33:1 and the total mmols of monomers relative to the mass of the vinyl co-polymer stabilizer was 52.1 :1.

[00104] The MGs prepared according to Method 2, when deposited from ethanol and DMSO, had z-average diameters ( d_z ) of 354 nm (PDI = 0.012) and 495 nm (PDI = 0.041) respectively, as measured by dynamic light scattering (Fig. 1A, Table A). The volume swelling ratio for the MG particles is estimated as 2.7 from these d_z values. They exhibited low polydispersity (Table A).

[00105] The MGs deposited from ethanol (Fig. 1 B) were spherical and had a number- average diameter of 336 nm when measured using SEM. The MGs deposited from DMSO (Fig. 1C) had a much larger diameter of 731 nm when measured using SEM. The MGs are highly deformable and flatten when deposited from the swollen state.

Table A: Characterisation data for MGs prepared according to Method 2 dispersed in ethanol and DMSO

a z-average diameter determined from dynamic light scattering. ^b Polydispersity index. ^c Number-average diameter determined from SEM. ^d Standard deviation for CISEM-

Method 3 - Preparation of (CH₃NH₃ ⁺)( Pb²⁺)(l )₃ [MAPblJ / MG films

[00106] Indium tin oxide (ITO)-coated glass substrates (20 W/sq) were cleaned by ultrasonication in a 1.0 wt% Hellmanex solution, rinsed with water, followed by I PA, NaOH (2.5 M) and then rinsed with water and dried. A T1O2 hole blocking layer (bl-TiC>2, 45 nm) was spin-coated at 2000 rpm for 60 s onto the ITO using TDB solution in 1 -butanol (0.30 M) and subsequent heating at 125 °C for 5 min. After that, the bl-TiC>2 film was annealed at 500 °C for 30 min. Upon cooling to room temperature, a MAPbU precursor solution* containing MGs (total volume 100 mI_) was spin-coated onto the ITO/bl-TiC>2 substrate at 4000 rpm for 25 s (using a Laurell WS-650 Mz-23NPP spin processor). During the spin coating process, toluene (200 mI_) was dripped onto the surface at a uniform rate onto the film during the last 10 sec. The films were annealed at 100 °C for 10 min and consisted of glass/ITO/bl-TiC>2/(x, y). The Cl content of the perovskite part of the (x, y) films is assumed to be negligible. All films were stored in a desiccator over P2O5 in the dark until further investigation.

*The precursor solution contained MAI, Pb _, MAC and DMSO (1 :1 :0.6:1 molar ratio) in 1.0 ml_ of DMF, as well as MGs in DMSO at various concentrations. The nominal compositions of the (x, y) films prepared were (30, 1.0), (25, 3), (20, 2.4), (15, 1.8), (10, 1.2) and (5, 0.6). An example preparation is given for the (10, 1.2) film: the precursor perovskite solution contained MAI (0.072g), Pbl₂ (0.208 g), MAC (0.0203 g) and DMSO (0.0990 g) in DMF (1.0 ml_) and the total concentration was 20 wt.%; the precursor / MG mixed solution was prepared by mixing the perovskite precursor solution (0.50 ml_, 20 wt%) with MGs in DMSO (0.50 ml_, 2.4 wt%); the mixed solution was sonicated in an ultrasonic bath for 10 min before deposition. MG-free control films were prepared as above, with the compositions (39, 0), (30, 0), (20, 0), (10, 0) and (5, 0). A perovskite-free MG control film was also prepared according to the same methodology coating only with MGs.

[00107] The perovskite/MG films are denoted in terms of the concentrations of MAPbU - also referred to herein as MAPI (x) and MG (y) used to spin coat the film, i.e. , (x, y). The MGs were first characterized and then the morphologies of the (x, y) films were investigated.

[00108] X-ray scattering patterns were measured for (20, 2.4), (10, 1.2) and the MG-free (20, 0) and (10, 0) films (see Fig. 2). Pbh was absent from all films which shows excellent conversion. Furthermore, there is a noticeable increase in the (110) peak broadness for the MG-containing (10, 1.2) and (20, 2.4) films compared to the respective MG-free (10, 0) and (20, 0) films. This indicates that the grain size decreases in the presence of the MGs. The (10, 1.2) and (20, 2.4) films showed multiple perovskite peaks and evidence of a more polycrystalline structure. In contrast (10, 0) and (20, 0) films showed fewer peaks and evidence of more preferred growth of larger grains.

[00109] SEM images for the (x, y) films (Fig. 3A - 3E) show dramatic changes as the films became thinner, i.e., as x and y decreased. Measured perovskite layer thicknesses are shown in Fig. 3F and Table B.

Table B: Thicknesses of various perovskite thin films prepared according to Method 3

[00110] Film thickness followed a power law relationship with increasing total precursor concentration (x + y). The SEM images show that as the films become thinner, the MGs reside on the surface of the thin films. The MG particles could no longer be buried when the film thickness was 120 nm (i.e. , for the (25, 3) film) or less. The thinnest composite film (5.0, 0.6) was too thin for thickness measurements to be obtained. However, black MGs were evident (Fig. 3E). Without being bound by theory, Fig. 3G depicts a proposal for the arrangement of the MGs in the perovskite/MG films. For the thin films, the MGs are surrounded by a relatively thin layer of perovskite that crystallizes as the MG collapsed. The morphologies of the thinnest perovskite/MG films (Fig. 3D (10, 1.2) and Fig. 3E (5, 0.6)) are profoundly different to those of control MG-free films (Fig. 3H and 3I). The control MG-free films (10, 0) and (5, 0) are irregular and have many large pinholes. In contrast the MG- containing films contain collapsed MGs as depicted in Fig. 3G, which advantageously prevent pinhole formation. Therefore, the inclusion of MGs decreases the proportion of pinholes for thin perovskite films.

[00111] Examination of the SEM image for the (5.0, 0.6) film (Fig. 3E) reveals what appear to be circular ridges around the MGs. AFM data were obtained for the same system and“volcano-like” objects were apparent (see Fig. 4). The average peak-to-peak distance was 876 ± 56 nm. This distance is sufficiently large to accommodate a flattened MG particle.

[00112] Delayed perovskite crystallization associated with the presence of MGs in the deposition mixture was confirmed by digital photographs taken of the MG-containing (20, 2.4) and the MG-free (20, 0) films on glass during annealing at 100 °C (See Fig. 5). The dark perovskite began to form at the edges of the (20, 2.4) film after 10 sec and this was complete after 17 sec. In contrast, the (20, 0) film showed formation of the perovskite phase after only 5 sec and film formation was complete after 10 sec. Therefore, MGs modulate perovskite crystallization; the inclusion of the MGs increases the time required for perovskite film formation by up to a factor of two. [00113] X-ray photoelectron spectroscopy (XPS) was used to examine the thin (10, 1.2) film. A thick MG-free perovskite film (39, 0) was used as a control it had the best coverage of the MG-free samples. A perovskite-free MG film was also used as a control. The N 1s XPS spectrum of the perovskite-free MG film (Fig. 6A) shows a single component centred at

399.5 eV binding energy (BE), which is assigned to the 0=C-NH- groups from the MG. The N 1s XPS spectra for the MG-free (39, 0) and (10, 1.2) films (Fig. 6A) have two components. A high BE peak is present in both films at 402.5 eV (39, 0) and 402.2 eV (10, 1.2) which is ascribed to the CH3NH3⁺ component of MAPI. The 0.3 eV BE shift towards low BE side suggests that the MG contributes electron density to MAPI. Neutral CH3NH2 molecules contribute to a weak shoulder for the (39, 0) film at 400.8 eV. For the (10, 1.2) film a dominant low-BE peak is present at 400.4 eV due to the MG which has shifted by 0.9 eV compared to the parent MG signal. This peak is due to MGs at the film surface and the BE shift suggests that the 0=C-NH- groups donate electron density to MAPI .

[00114] Pb 4f spectra (Fig. 6B) also show evidence that the MGs interact with the perovskite. The spectra for the (39, 0) film shows a spin-orbit split doublet from Pb²⁺ 4f_7/2 at

138.6 eV. It is apparent that in addition to the main peak there is also under-coordinated Pb° at 136.9 eV BE. Under-coordinated Pb has been reported for perovskite films [Wu et ai, Adv. Energy Mater. 2019, 1803766] Intriguingly, the signal decreases from 12% (for the (39, 0) film) to 6% (for the (10, 1.2) film) with respect to Pb. This suggests that the MG coordinates to some of this Pb species and donates electron density. Such behaviour has been reported for donor-n-acceptor molecules [Wu et ai, Adv. Energy Mater. 2019, 1803766], but not for non-conjugated polymer additives. The XPS results show that the MG particles strongly interact with the perovskite and this probably occurs through complex formation with Pb.

[00115] The XPS data also enabled the composition of the films to be probed. The N/Pb stoichiometries for the (10, 1.2) and (39, 0) films were 2.0 ± 0.2 and 0.8 ± 0.2, respectively. Using the molecular weight of the perovskite (MAPI) and that for the poly(NVF-NVEE) copolymer which comprised the MGs, this ratio corresponds to nominal MG weight fraction ( WMG ) of ~ 10.3%. The latter agrees closely to the calculated (10, 1.2) film WMG value of 10.7% used during film preparation, wherein WMG = 100y/(x + y). These data confirm the presence of the MGs in the composite perovskite/MG films. The N/Pb value for the (39, 0) film is as expected for MAPI.

[00116] Transmittance spectra for the thin films were measured (see Fig. 8A). These data show that the transmittance increased dramatically as the composition moved toward the thinnest ST systems, i.e. , as x and y decreased. Figure 8B shows the AVT values plotted as a function of x. Clearly, AVT increased as the concentration of perovskite (x) decreased. The (10, 1.2) film had an AVT of 46.8 % and is above the AVT threshold of 25% for solar window applications. This is also demonstrated by the photograph of the building viewed through the (10, 1.2) film as shown in Fig. 8B. Most remarkably, Fig. 8B shows that a cross-over of AVT occurs when x decreases to below 20%. Below this cross-over point, the AVT values of the perovskite/MG composite thin films are higher than those for the respective perovskite-only films; the inclusion of 12% microgels to the perovskite precursors resulted in the deposition of films having more than twice the AVT of the equivalent films prepared in the absence of microgels (53.4% (5.0, 0.6) vs. 26.5% (5.0, 0); 46.8% (10, 1.2) vs 16.9% (10, 0)). This is potentially due to the MG-promoted formation of thin pinhole-free ST films, resulting in lower refractive index variation and hence less light scattering.

[00117] The PL spectra (Fig. 9A) of the thin (x, y) films show interesting trends when compared to PL spectra for the MG-free films (shown in Fig. 7A.) Firstly, the PL intensities of the MG-containing thin films are about an order of magnitude higher than those for the respective MG-free films (see Fig. 7B). Secondly, the PL spectra show a large blue shift of the wavelength of maximum PL intensity (T_max) as x decreased (See Fig. 9B). There are two possible causes for these effects. The first is due to a decrease in the grain size as noted above. Such blue shifts are known for perovskites as the grain size decreases and is apparent for the MG-free films, but to a lesser extent (Fig. 9B). Additionally, it is suggested that the MG interaction with the perovskite noted above resulted in passivation. The latter is well known to cause a blue-shift in the PL spectra (Salado et ai, Nano Energy 2018, 50, 220-228) and may increase the PL intensity because of less non-radiative energy transfer.

Method 4 - Solar cell fabrication

[00118] The procedure to prepare the glass/ITO/bl-Ti0₂/(x, y) films was as described above in Method 3. Film fabrication, including Spiro deposition, was conducted inside a nitrogen-filled glovebox (humidity ~ 2%). CBZ was used as the HTM solvent at room temperature. LiTFSI (4.8 pL, 520 mg/ml) and TBP (8.0 pL) were also added to the Spiro solution. The spiro hole transfer matrix films (200 nm) were formed by spin coating at 4000 rpm for 20 s onto the perovskite films. The planar devices were coated with a gold layer (70 nm) using thermal evaporation. For the devices containing a meso-Ti0₂ layer, 18NRT was diluted 1 : 10 with EtOH and then 75 m\- was spin coated at 5000 rpm for 30 sec on the bl- T1O2 layer. The layer was then annealed at 500 °C for 30 min. The device architectures for solar cells prepared with or without the meso-Ti0₂ layer are shown in Fig. 11.

[00119] PSC devices with (10, 1.2) photoactive film layers were constructed. (10, 1.2)^meso refers to the devices constructed with a thin meso-Ti0₂ layer. Control devices were also constructed using (5, 0) photoactive layers.

[00120] Representative J-V curves are shown in Fig. 12 A & B which demonstrate that operational PSCs were prepared in all cases. The figures of merit extracted from the J-V data are shown in Table C and Figs. 12C & 12D.

Table C: PSC performance data

[00121] The average PCE for the (10, 1.2) device was 7.69%. The best cell had a PCE of 10.2% (Fig. 12D). Devices prepared using the MG-free (5.0, 0) ST film (which had an AVT of 26.5%) had an average PCE of 4.93%. The steady state current density for that system (Fig. 12C) continually increased with time, which contrasts to the (10, 1.2) film which was constant. The temperature of the (5.0, 0) film increased from 25 °C to 70 °C after 300 s during the measurement, which was the likely cause of the increase in the current density. The PCE was relatively low for (5.0, 0) devices because of the high proportion of pinholes as can be seen from Fig. 3I.

[00122] The PCE was further increased by adding a thin meso-TiC>2 layer to the (10, 1.2) system. The SEM of the (10, 1.2)^meso film and transmittance spectra for the film are shown in Figs. 10A and 10B, respectively. The morphology of the film is similar to that for (10, 1.2) from Fig. 3D. The AVT decreased to 40.7% as a consequence of the thin perovskite-filled meso-TiC>2 layer. However, the AVT remains much greater than 25%. The best (10, 1.2)^meso device had a PCE of 11.6% (Fig. 12B) and the average PCE is 9.62%, which is 25% higher than measured for (10, 1.2). The PCE values for the devices measured are compared in Fig. 12D.

[00123] The average PCE and AVT values for the (10, 1.2) device and (10, 1.2)^meso device compare favourably to those reported in the literature as shown by Table D. Specifically, the PCE is relatively high given the large AVT value, indicating the suitability of the thin perovskite films according to the present invention for use as ST PSCs.

Table D: Comparative PCE/AVT data for representative ST PSCs

aPCE obtained with illuminations from the bottom electrode and the PCE varies with glass, ITO and FTO substrates as well as the device areas. ^bThe AVT varies with the wavelength range. ^c For the (10, 1.2) and (10, i .2)^meso devices the AVT was measured without the Au top electrode and spiro layer.

[00124] Table E compares the properties and performance of exemplary films and devices prepared according to the present invention (entries 5 and 6) with MG-containing devices described in the prior art (WO2019/224550 - entries 1-3; & Dokkhan et al., Phys. Chem. Chem. Phys. 2018, 20, 27959-27969 - entries 1-2), a comparative device described in this application (entry 4) and an MG-free thin film device (entry 7).

Table E: Comparative data for (MP, MG) devices

ND = not determined; NA = not applicable; ^aPCE obtained with illuminations from the bottom electrode; ^bAVT was measured without the Au top electrode and spiro layer; ^cDevices prepared with mp-TiC>2 layer (glass/IT0/bl-Ti0₂/mp-Ti0₂/(MP, MG)/ spiro-OMeTAD/Au); ^dDevices prepared without the mp-Ti0₂ layer (glass/ITO/bl-Ti0₂/(MP, MG)/ spiro-OMeTAD/Au).

[00125] It can be seen from the data in Table E that although entries 1 & 2 comprising larger MGs (1125 nm z-average diameter in DMSO) demonstrated reasonable PCEs, reduction of the size of the MGs alone was insufficient to provide devices with AVT > 25% in order to be suitable as semi-transparent PSCs (entries 3 & 4 have approx. 2% AVT). It appears that a synergistic combination of sub-micrometer MGs and a thin capping layer (< 100 nm thickness) is required to surprisingly deliver not only good PCEs, but also high AVT values, making devices prepared with perovskite photoactive layers according to the present invention excellent candidates for application in the field of semi-transparent solar cells.

REFERENCES

1. Chen, B.-X.; Rao, H.-S.; Chen, H.-Y.; Li, W.-G.; Kuang, D.-B.; Su, C.-Y., Ordered macroporous CH3NH3Pbl3 perovskite semitransparent film for high-performance solar cells. J. Mat. Chem. A. 2016, 4, 15662-15669.

2. Eperon, G. E.; Burlakov, V. M.; Goriely, A.; Snaith, H. J., Neutral Color Semitransparent Microstructured Perovskite Solar Cells. ACS Nano 2014, 8, 591-598.

3. Della Gaspera, E.; Peng, Y.; Hou, Q.; Spiccia, L; Bach, U.; Jasieniak, J. J.; Cheng, Y.- B., Ultra-thin high efficiency semitransparent perovskite solar cells. Nano Energy 2015, 13, 249-257.

4. Ka, I.; Asuo, I. M.; Basu, S.; Fourmont, P.; Gedamu, D. M.; Pignolet, A.; Cloutier, S. G.; Nechache, R., Hysteresis- Free 1 D Network Mixed Halide-Perovskite Semitransparent Solar Cells. Small 2018, 14, 1802319. 5. Rai, M.; Rahmany, S.; Lim, S. S.; Magdassi, S.; Wong, L. H.; Etgar, L, Hot dipping post treatment for improved efficiency in micro patterned semi-transparent perovskite solar cells.

J. Mat. Chem. A. 2018, 6, 23787-23796.

6. Roldan-Carmona, C.; Malinkiewicz, O.; Betancur, R.; Longo, G.; Momblona, C.; Jaramillo, F.; Camacho, L.; Bolink, H. J., High efficiency single-junction semitransparent perovskite solar cells. Energy Env. Sci. 2014, 7, 2968-2973.

7. Zhang, S.; Lu, Y.; Lin, B.; Zhu, Y.; Zhang, K.; Yuan, N.-Y.; Ding, J.-N.; Fang, B., PVDF- HFP additive for visible-light-semitransparent perovskite films yielding enhanced photovoltaic performance. Sol. Energy Mater. Sol. Cells 2017, 170, 178-186.

8. Wang, Y.; Mahmoudi, T.; Yang, H.-Y.; Bhat, K. S.; Yoo, J.-Y.; Hahn, Y.-B., Fully- ambient-processed mesoscopic semitransparent perovskite solar cells by islands-structure- MAPbl3-xClx-NiO composite and AI203/Ni0 interface engineering. Nano Energy 2018, 49, 59-66.

9. Chang, C.-Y.; Lee, K.-T.; Huang, W.-K.; Siao, H.-Y.; Chang, Y.-C., High-Performance, Air-Stable, Low-Temperature Processed Semitransparent Perovskite Solar Cells Enabled by Atomic Layer Deposition. Chem. Mater. 2015, 27, 5122-5130.

10. Chen, S.; Chen, B.; Gao, X.; Dong, B.; Hu, H.; Yan, K.; Wen, W.; Zou, D., Neutral- colored semitransparent solar cells based on pseudohalide (SCN-)-doped perovskite. Sustainable Energy Fuels 2017, 1 , 1034-1040.

11. Horantner, M. T.; Zhang, W.; Saliba, M.; Wojciechowski, K.; Snaith, H. J., Templated microstructural growth of perovskite thin films via colloidal monolayer lithography. Energy Env. Sci. 2015, 8, 2041-2047.

12. Ramirez Quiroz, C. O.; Bronnbauer, C.; Levchuk, I.; Hou, Y.; Brabec, C. J.; Forberich,

K., Coloring Semitransparent Perovskite Solar Cells via Dielectric Mirrors. ACS Nano 2016, 10, 5104-5112.

13. Ramirez Quiroz, C. O.; Levchuk, I.; Bronnbauer, C.; Salvador, M.; Forberich, K.; Heumuller, T.; Hou, Y.; Schweizer, P.; Spiecker, E.; Brabec, C. J., Pushing efficiency limits for semitransparent perovskite solar cells. J. Mat. Chem. A. 2015, 3, 24071-24081.

14. Xue, Q.; Bai, Y.; Liu, M.; Xia, R.; Hu, Z.; Chen, Z.; Jiang, X.-F.; Huang, F.; Yang, S.; Matsuo, Y.; Yip, H.-L; Cao, Y., Dual Interfacial Modifications Enable High Performance Semitransparent Perovskite Solar Cells with Large Open Circuit Voltage and Fill Factor. Adv. Energy Mater. 2017, 7, 1602333.

15. Zhang, L.; Horantner, M. T.; Zhang, W.; Yan, Q.; Snaith, H. J., Near-neutral-colored semitransparent perovskite films using a combination of colloidal self-assembly and plasma etching. Sol. Energy Mater. Sol. Cells 2017, 160, 193-202.

16. Zhao, J.; Brinkmann, K. O.; Hu, T.; Pourdavoud, N.; Becker, T.; Gahlmann, T.; Heiderhoff, R.; Polywka, A.; Gorrn, P.; Chen, Y.; Cheng, B.; Riedl, T., Self-Encapsulating Thermostable and Air-Resilient Semitransparent Perovskite Solar Cells. Adv. Energy Mater. 2017, 7, 1602599.

17. Bag, S.; Durstock, M. F., Efficient semi-transparent planar perovskite solar cells using a ‘molecular glue’. Nano Energy 2016, 30, 542-548.

18. Heo, J. H.; Jang, M. H.; Lee, M. H.; Han, H. J.; Kang, M. G.; Lee, M. L; Im, S. H., Efficiency enhancement of semi-transparent sandwich type CH3NH3Pbl3 perovskite solar cells with island morphology perovskite film by introduction of polystyrene passivation layer. J. Mat. Chem. A. 2016, 4, 16324-16329.

19. Bu, L.; Liu, Z.; Zhang, M.; Li, W.; Zhu, A.; Cai, F.; Zhao, Z.; Zhou, Y., Semitransparent Fully Air Processed Perovskite Solar Cells. ACS Appl. Mater. Inter†. 2015, 7, 17776-17781.

Claims

Claims

1. A photoactive layer for a perovskite solar cell comprising:

• a hybrid inorganic-organic perovskite of formula ABX₃, wherein:

A is Ci-6alkyl-NH3⁺ and optionally also includes one or more of Cs⁺, Rb⁺, guanidinium and formamidinium;

B is selected from Pb²⁺ Ba²⁺ and Sn²⁺ ; and

X is selected from one or more of Br, Cl and I ;

provided that A and B balance the X charge, so that overall A is singly-charged and B is doubly-charged; and

• a plurality of microgel particles formed from a hydrophilic crosslinked polymeric material capable of swelling in polar aprotic solvents;

wherein the z-average diameter of the swollen microgel particles in a polar aprotic solvent is less than 1000 nm; and

the photoactive layer has a thickness of less than 100 nm.

2. A photoactive layer according to claim 1 , wherein the microgel particles comprise a co-polymer of monomers (I) and (II):

wherein:

Y is selected from:

Z is selected from one of the following linkers: R¹ , R², R³, R⁴, R⁵, R⁶, R⁷, R⁸ and R⁹ are independently selected from hydrogen and Ci-3alkyl;

R¹⁰ and R¹¹ are independently selected from hydrogen and Ci-3alkyl; or

R¹⁰ and R^{1 1} are taken together with the moieties to which they are attached to form a 4- to 9-membered lactam;

R¹² is selected from OH, NR¹³R¹⁴ and -(OCH₂CH₂)_P-OR¹⁵;

R¹³, R¹⁴ and R¹⁵ are independently selected from hydrogen and Ci-3alkyl;

U and L² are independently selected from divalent alkyl, divalent alkylether, divalent alkylamine, divalent alkylamide and divalent alkylester linker groups;

Ar is a divalent optionally substituted aryl or heteroaryl group;

n is 1 to 20; and

p is 2 to 20.

3. A photoactive layer according to claim 2, wherein Z is selected from:

4. A photoactive layer according to claim 1 to 3, wherein A is Ci-6alkyl-NH3⁺.

5. A photoactive layer according to claims 1 to 4, wherein B is Pb²⁺.

6. A photoactive layer according to claim 1 , wherein A is CHsNh , B is Pb²⁺ and X is a combination of Cl and I .

7. A photoactive layer according to claims 2 to 6, wherein Y is:

are both hydrogen.

8. A photoactive layer according to claims 2 to 6, wherein Y is:

are taken together with the moieties to which they are attached to form a 4- to 9-membered lactam.

9. A photoactive layer according to claim 8, wherein R¹⁰ and R¹¹ are taken together with the moieties to which they are attached to form a 7-membered lactam.

10. A photoactive layer according to claims 2 to 6, wherein Y is:

11. A photoactive layer according to claim 10, wherein R¹³ and R¹⁴ are both hydrogen.

12. A photoactive layer according to claim 10, wherein R¹³ is hydrogen and R¹⁴ is isopropyl.

13. A photoactive layer according to claims 1 to 12, wherein the z-average diameter of the swollen microgel particles in a polar aprotic solvent is less than 750 nm, such as less than 500 nm.

14. A photoactive layer according to claims 1 to 13, wherein the polar aprotic solvent is g-butyrolactone, dimethyl formamide, dimethyl sulfoxide, or a combination of these solvents, preferably. dimethyl sulfoxide.

15. A photoactive layer according to claims 1 to 14, wherein the photoactive layer has a thickness of less than 50 nm, such as less than 20 nm.

16. The photoactive layer according to claims 1 to 15, wherein the photoactive layer is suitable for incorporation in a semi-transparent solar cell.

17. The photoactive layer according to claim 16, wherein the photoactive layer has an average visible transmittance (AVT) of at least 25%.

18. A method of forming a photoactive layer according to claim 1 , comprising the steps of:

a) swelling particles of the microgel in a polar aprotic solvent so that the z- average diameter of the swollen microgel particles in the polar aprotic solvent is less than 1000 nm;

b) adding hybrid inorganic-organic perovskite precursors to the dispersion of swollen microgel particles from step a);

c) coating the dispersion from step b) onto a substrate; and

d) evaporating the solvent;

wherein the dispersion resulting from step b) comprises the microgel particles at a concentration of 0.1 to 5 % w/w and the hybrid inorganic-organic perovskite precursors at a concentration of 2 to 20 % w/w.

19. A method according to claim 18, wherein the swollen microgel particles are 1.2-100 times the size of the unswollen particles.

20. A method according to claim 18 or 19, wherein the polar aprotic solvent is selected from g-butyrolactone, dimethyl formamide, dimethyl sulfoxide, or a combination of these solvents; preferably dimethyl sulfoxide.

21. A method according to claims 18 to 20, wherein the hybrid inorganic-organic perovskite precursors are of the formula AX and BX2, wherein A, B and X are as defined in claim 1.

22. A method according to claims 18 to 21 , wherein in step c) the coating is carried out by spin coating.

23. A method according to claims 18 to 22, wherein step c) further comprises the addition of an anti-solvent (preferably selected from chlorobenzene, benzene, xylene, toluene, methanol, ethanol, ethylene glycol, 2-propanol, chloroform, THF, acetonitrile and benzonitrile) during coating.

24. A method according to claims 18 to 23, wherein step d) is carried out at 80 to 200 °C.

25. A method according to claims 18 to 24, wherein in step a) the microgel particles are provided at a concentration of 0.5 to 2.5 % w/w, and in step b) the hybrid inorganic- organic perovskite precursors are provided at a concentration of 5 to 15 % w/w.

26. A photoactive layer obtainable by a method according to claims 18 to 25.

27. A perovskite solar cell comprising a photoactive layer according to claims 1 to 17 or claim 26.

28. A perovskite solar cell according to claim 27, having an average visible transmittance (AVT) of at least 25%.

29. A perovskite solar cell according to claim 27, having an average visible transmittance (AVT) of at least 40%.

30. A perovskite solar cell according to claims 27 to 29, having a power conversion efficiency (PCE) greater than 5%.

31. A perovskite solar cell according to claims 27 to 29, having a power conversion efficiency (PCE) greater than 7.5%.