Perovskite photovoltaic device and solution-processed method for manufacture

Abstract

The invention relates to a carbon ink composition for use in printing a conductive carbon layer onto a perovskite solar device stack. Also provided is a perovskite solar device stack comprising a conductive carbon layer, and a method of manufacturing such a device by a solution processing method.

Description

Perovskite Photovoltaic Device and Solution-Processed Method for Manufacture The invention relates to a carbon ink composition for use in printing a conductive carbon layer onto a perovskite solar device stack. Also provided is a perovskite solar device stack comprising a conductive carbon layer, and a method of manufacturing such a device by a solution processing method. Background of Invention The interest in perovskite photovoltaics has been intense due to the well reported increase in lab based device performance[1-3]. Since the original reporting of perovskite based devices there has been a rapid rise in device efficiency achieving over 25% power conversion efficiency (PCE). In addition to the impressive device performance improvements in lab scale fabrication, perovskite photovoltaics as solution-processed devices have the promise of high scalability through roll to roll coating processes. Among available coating techniques, the feasibility of perovskite photovoltaic (PV) scaling-up has been mostly investigated through the slot die coating process. This industrially scalable process promises high speed roll to roll (R2R) processing of perovskite PV devices, an attractive proposition as when combined with R2R encapsulation large volumes of completed photovoltaics could be fabricated in a single process. Research in this area is growing with feasibility demonstrations conducted on benchtop deposition tools for both p-i-n and n-i-p device stacks. More recently, partly R2R coated perovskite PV devices have been demonstrated. In one study, a two-step sequential process demonstrated to fabricate slot-die coated R2R n-i-p devices with a high PCE of 13%

[0023] . Othman et al, applied a triple-cation mixed halide perovskite in an inverted p-i-n stack using R2R coating for the first time and reached a PCE of 12%

[0024] . In their work, they also modified the underlying hole transport layer with guanidinium iodide additive to form millimiter-sized perovskite clusters. The present inventors have previously demonstrated a p-i-n device structure of PET / ITO / PEDOT:PSS / MAPI / PCBM / BCP, which achieved a PCE of 12.2% with all layers sequentially coated R2R except the conductive top electrode

[0025] . Recently the potential for a laminated electrode made up of multilayer silver and carbon stack hasbeen demonstrated by applying the electrode onto a R2R coated stack using a calendar press.

[0026] These represent important advancements towards the promise of large area, high volume solution processed perovskite devices, and are summarised in Table 1. However, in almost all R2R perovskite photovoltaic research to date, the final electrode deposition is achieved through evaporation of a precious and limited- resource metal such as gold or silver. Although evaporation of gold or silver as a contact electrode has enabled researchers to fabricate efficient and consistent devices, this slow, expensive and wasteful process is a key barrier to perovskites leaving the lab and finding real world applications. It is, therefore, an object of the present invention to provide a solution to this slow and high-cost manufacturing step by preparing, optionally by roll to roll processing, a photovoltaic device comprising a printed solution based top electrode. Printed carbon electrodes have been proven to be effective in the triple mesoporous architecture, where several high temperature (>450ºC) processing steps are required which are incompatible with the polymer, e.g. polyethylene terephthalate (PET) or polyethylene napthalate (PEN), film substrates that are typically required for R2R processing

[0035] . Carbon electrodes have also been successfully utilised in planar device architectures at lab scale, however there are a few major problems inhibiting carbon electrode perovskite solar cells (PSCs) achieving high performance. Zhang et al, overcame the solvent compatibility issue through solvent exchange and hot pressing of dry carbon layers onto the completed devices

[0041] . Alternative strategies use complex or high temperature processes such as protecting the perovskite layer using Copper thiocyanate

[0042] or evaporated interlayers to enable interlayer to carbon electrode integration

[0043] . In another innovative work

[0044] a spray-coated graphene electrode was used on both sides of a modular laminated top contact PSC design, which its sheet resistance and thickness were greatly reduced under applied pressure. However, due to the complexity, high temperature or carbon electrode formulation these solutions are not scalable through R2R processing. To overcome both the scalability and compatibility issue, Gaveshana et al, introduced a novel two-component electrode including modified aluminium foil and carbon layer

[0045] . Although their electrode is compatible with layers underneath and can bepressed or laminated easily, the aluminium foil on the other hand, cannot be applied for sequential R2R purposes. It is, therefore, a further object of the invention to provide a carbon ink formulation that is suitable for depositing a carbon top electrode onto a perovskite device stack by a solution-based process such as R2R compatible slot die coating, screen printing, flexographic printing, blade coating, gravure printing, slide coating, curtain coating, inkjet printing, roll coating, curtain coating and / or wire bar coating. As will be readily appreciated by the skilled reader, formulation of the carbon ink is critical to achieve such a solution-based carbon electrode that is compatible with other perovskite device layers, with solvent orthogonality with the device stack being a limiting factor. Further, to enable coating of the full perovskite device stack including the electrode, the device architecture must be designed for process compatibility i.e. low temperature processing, layer orthogonality, interlayer alignment and solvent compatibility. We therefore herein disclose a unique carbon ink formulation and, using such a formulation to deposit a perovskite compatible back electrode in place of conventional evaporated metal contact, provide for the first time a fully roll to roll printable perovskite photovoltaic device architecture that overcomes interlayer incompatibilities and recombination losses to provide device performance comparable to those employing evaporated gold electrodes. Moreover, such roll to roll printed devices surprisingly exhibit enhanced long-term stability compared to typical PSC devices comprising conventional evaporated metal top electrodes. Statements of Invention The present invention, in its various aspects, is as set out in the accompanying claims. According to a first aspect of the invention there is provided a carbon ink composition for use in printing a conductive carbon layer onto a perovskite solar device stack, wherein the ink composition is fully compatible with the underlying layers of the device stack and is rheologically compatible with solution processing methods such as roll to roll slot die coating.The ink composition of the invention comprises: (a) one or more binders; (b) one or more populations of conductive carbon particles; and (c) one or more solvents, wherein said solvents are selected from: ethanol, ethyl acetate, 1-butanol, 2- methylanisole, xylene, toluene, cyclohexanone, indane or any isomer thereof. Each of the above solvents is expected to be compatible with perovskite materials and so can be used in the solution processing of a perovskite solar device stack without detriment to PCE. Further, each of these solvents advantageously has a suitable boiling point of between about 70 °C and about 250 °C, and preferably between about 150 °C and about 200 °C, and so permit low temperature curing whilst remaining wet through solution processed carbon deposition. In preferred embodiments, the solvent component of the ink composition is 2- methylanisole or 1-butanol, and most preferably is 2-methylanisole. This solvent has been shown to cause no photoluminescence deterioration of the perovskite layer even after following 24 hour soaking. Further 2-methylanisole has a boiling point of 170 °C and so is well suited to low temperature curing solution processing. Furthermore, there is no significant toxicity associated with 2-methylanisole, and therefore no associated workplace exposure limits are in place making it suitable for scale up and use in large volumes. It is well known that, in contrast to the above solvents, water is incompatible with perovskite materials, with perovskite materials having been shown to rapidly degrade upon exposure to water. Therefore, in order to maximise the stability, and so maintain PCE, of a solution processed perovskite solar device stack, the ink formulation of the present invention is preferably substantially free of water. As used herein, the term ‘substantially free of water’ refers to a formulation or composition that contains no more than about 10 wt.%, preferably no more than about 5 wt.%, more preferably no more than about 1% and still more preferably no more than about 0.5 wt.%, of free or unassociated water. It will be understood by a person of the art that to the extent any of the individual components of the composition are hydrophilic, they can contain water in an associated or unfree or absorbed form, and / or may absorb water from the atmosphere. In particularly preferred embodiments, such compositions contain no free or unassociated water.The ink composition of the invention comprises one or more conductive carbon particles which, following deposition and drying of the ink, form a conductive carbon layer. As will be readily appreciated by the skilled person, the type of carbon particles included in the ink are not particularly limited, and appropriate particle type(s) can be selected depending on, e.g. conductivity and viscosity requirements. However, in preferred embodiments the ink comprises or consists of a mixture of carbon black and graphite particles, and more preferably a mixture wherein the weight ratio of graphite to carbon black is from about 1.5:1 to about 4:1. An about 2.6:1 mixture of graphite to carbon black has been shown to provide excellent PCE when deposited as an electrode layer in a perovskite device stack. The ink composition of the invention comprises one or more binders. Such binders provide adhesion to the substrate and, dependent on concentration, the required rheological properties to the ink composition. As will be readily appreciated by the skilled person, the type of binder is not particularly limited, provided that said binder is soluble in the chosen solvent, and allows printing of the ink formulation onto a substrate such as a perovskite solar device stack. However, in preferred embodiments the binder comprises or consists of one or more hydrophilic polymer, examples of which include polyvinyl alcohol, polyvinyl pyrrolidone, gelatin, celluloses, pectins, polyoxazolines, polyvinylacetamides, partially hydrolyzed polyvinyl acetate / vinyl alcohol, polyacrylic acid, polyacrylamide, polyalkylene oxide, sulfonated or phosphated polyesters and polystyrenes, casein, zein, albumin, and derivatives thereof. More preferably, the binder comprises or consists of one or more cellulose. Preferably the cellulose is a cellulose ether, and still more preferably is ethyl cellulose Particularly preferred cellulose, cellulose ether and carboxymethyl cellulose binders have a Mw of from about 10,000 to about 200,000 Da. A good balance of rheological and conductive properties, which enables the inks to be used to deposit carbon electrodes in perovskite solar device stacks, is achieved using inks comprising from about 15 wt.% to about 35wt.% carbon particles, and about 5 wt.% to about 15 wt.% binder in said solvent. However, for compatibility with roll to roll processes such as slot die coating, the amount of binder and carbon particles is typically reduced in order to lower ink viscosity. For example, in a preferred embodiment, a low viscosity formulation is provided that retains excellent conductiveproperties and comprises from about 7 wt.% to about 18 wt.% carbon particles, and about 3 wt.% to about 10 wt.% binder in said solvent. As noted above, the binder and carbon particles are provided in an amount sufficient to control the rheological profile of the ink composition. In preferred examples, the ink composition is a shear thinning composition. Further, when the ink is for use in stencil or screen printing, said ink preferably has a shear viscosity of between about 3000 Pa.s and about 0.3 Pa.s, over a shear range of 0.1 to 1000s-1. However, in preferred examples, the ink is for use in a R2R compatible slot die coating process, and said ink preferably has a shear viscosity of between about 15 Pa.s and about 1.5 Pa.s, over a shear range of 0.1 to 1000 s-1. In addition, the ink formulation may be doped with one or more conventional dopants, to improve perovskite device efficiency. For example, in certain preferred embodiments, the ink may be doped with from about 0.5 mg / ml to 20 mg / ml of a dopant such as bathocuproine. Exemplary Ink compositions of the invention are set out in Tables 3 to 5. The ink of the first aspect of the invention is configured for use in depositing a perovskite compatible carbon back electrode, in place of conventional evaporated metal contact in a perovskite solar device stack, that overcomes interlayer incompatibilities and recombination losses to provide device performance comparable to those employing evaporated gold electrodes. Therefore, according to a second aspect, the invention provides a planar structural perovskite solar device stack comprising: (i) a transparent conductive layer; (ii) an electron transport layer; (iii) a perovskite layer; (iv) a hole transport layer; and (v) a conductive carbon layer. As the skilled reader will appreciate, such layers can be arranged in a n-i-p device structure, wherein the electron transport layer contacts the transparent conductive layer (in a (i)-(ii)-(iii)-(iv)-(v) arrangement), or alternatively such layers can be arranged in an inverted p-i-n device structure wherein the hole transport layer contacts the transparent conductive layer (in a (i)-(iv)-(iii)-(ii)-(v) arrangement). However, in preferred embodiments, the perovskite solar device stack has a n-i-p device structure. As will be readily appreciated by the skilled person, the choice of transparent conductive layer is not particularly limited. For example, where flexibility is not required, this layer may be formed from indium tin oxide (ITO) coated glass. However,where a flexible substrate is required, e.g. in roll to roll processing, this layer is preferably formed from an ITO coated polymer such as a PET or PEN film. Similarly, the choice of electron transport layer is also not particularly limited. However, in preferred embodiments, this layer is formed from tin oxide (SnO2), as this can be processed at low temperatures and is suitable for slot die coating. Likewise, the choice of perovskite layer is also not limited. However, in preferred embodiments, this layer comprises a methylammonium lead halide, more preferably methylammonium lead iodide (MAPI), i.e. a solid compound with perovskite structure and a chemical formula of CH3NH3PbI3. Further, the choice of hole transport material is not particularly limited and may, for example comprise Spiro-MeOTAD. However, use of such a hole transport layer in combination with a conductive carbon layer has been shown to result in a vast reduction in efficiency and depressed J-V characteristics compared with analogous stacks comprising a gold electrode. However, this reduction in efficiency is overcome by the preferred use of at least one hole transport layer comprising: nickel oxide; copper thiocyanate; copper indium sulphide; poly[bis(4-phenyl)(2,4,6- trimethylphenyl)amine (PTAA); poly(3-hexylthiophene-2,5-diyl (P3HT); a poly(3,4- ethylenedioxythiophene), i.e. PEDOT, polymer mixture; or any combination thereof. Multiple hole transport layers may be provided, and in such embodiments, the perovskite solar device stack preferably comprises two or three hole transport layers, wherein each layer comprises nickel oxide, copper thiocyanate, copper indium sulphide, PTAA, P3HT and / or a PEDOT polymer mixture. In particularly preferred embodiments, a PEDOT polymer mixture comprising a sulfonated block copolymer as the counter ionomer is used as a hole transport layer as such counter ionomers counteract any instability issues associated with, e.g. sulfonated homopolymer counter ionomers such as sodium polystyrenesulfonate (PSS), with the resultant device architecture surprisingly being found to outperform analogous evaporated gold-based devices. PEDOT polymer mixtures comprising a sulfonated block copolymer as the counter ionomer are sold commercially from various sources, e.g. CleviosTMHTL Solar 3 (Haraeus Group, Hannau, DE). Without wishing to be bound to any particular theory, it is thought that PEDOT can not only act a hole transport layer that is compatible with carbon ink solvent, but also thatits intrinsically conductive property aids charge transfer to the printed carbon electrode, thereby minimizing recombination losses. Particularly where the hole transport material comprises PEDOT, the hole transport layer preferably has a thickness from about 40 to about 200nm, and ideally is about 90 nm. At this thickness, the perovskite is effectively covered with low defect occurrence at this thickness. The conductive carbon layer is formed by printing the ink of the first aspect. Therefore, preferred features relating to this carbon component are as described above for the first aspect of the invention. Each layer of the planar structural perovskite solar device stack of the second aspect, including conductive carbon layer, can be deposited by solution processing methods such as by slot die coating. Therefore, according to a third aspect, the invention provides a method of manufacturing the planar structural solar device stack of the second aspect, wherein the electron transport layer, perovskite layer, hole transport layer, and conductive carbon layer are sequentially coated onto a transparent substrate via a solution processing method. In preferred embodiments, said layers are coated onto the transparent substrate by slot-die coating. In such processes, each of the electron transport layer, perovskite layer, hole transport layer and conductive carbon layer are suitably coated at a coating speed of from about 0.1 m / min to about 1000 m / min, preferably from about 0.2 m / min to about 500 m / min, and more preferably from 0.5 m / min to about 5 m / min, with about 1 m / min being most preferred. Additionally, following application of each layer to the stack and prior to application of any subsequent layer, the coated substrate is dried by exposure to heat, preferably between about 120 °C and about 140 °C, for a period of between about 30 seconds and 2 minutes. In preferred embodiments, the transparent substrate is preferably a flexible substrate such as PET or PEN, which may optionally be coated with a Tin doped indium oxide (ITO) film. In such embodiments, the device stack can be formed by roll to roll processing that, as shown using electroluminescence testing, is free of disruptive defects.Throughout the description and claims of this specification, the words “comprise” and “contain” and variations of the words, for example “comprising” and “comprises”, mean “including but not limited 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. Throughout the description and claims of this specification, the word "about" means ± 5 %, alternatively ± 2 % unless the context otherwise requires. All references, including any patent or patent application, cited in this specification are hereby incorporated by reference. No admission is made that any reference constitutes prior art. Further, no admission is made that any of the prior art constitutes part of the common general knowledge in the art. Preferred features of each aspect of the invention may be as described in connection with any of the other aspects. Other features of the present invention will become apparent from the following examples. Generally speaking, the invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including the accompanying claims and drawings). Thus, features, integers, characteristics, compounds or chemical moieties described in conjunction with a particular aspect, embodiment or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein, unless incompatible therewith. Moreover, unless stated otherwise, any feature disclosed herein may be replaced by an alternative feature serving the same or a similar purpose. The invention will now be described, by way of example only, with reference to the following figures and tables wherein: Figure-1. a) PL measurment- ITO / MAPI samples soaked in different solvents for a 24hr period and b) glass / MAPI / carbon with different carbon ink formulation. Carbon inks with methylanisole, 1-butanol and 1-butanol-BCP solvents had ethylcelulose as abinder and carbon ink with diacetone alcohol had tripolymer (poly(vinyl chloride-co- vinyl acetate- co-vinyl alcohol) resin; Figure-2. a) Schematic of stencil coating method and photograph of PSCs with carbon and gold top electrodes, b, c) steady-state PL spectra and J-V curves of perovskite / carbon, perovskite / spiro-MeOTAD / carbon and perovskite / spiro / gold samples, d) cross-section SEM image of stencil coated carbon, e and f) J-V curve and steady-state PL spectra of perovskite / carbon and perovskite / PEDOT / carbon; Figure-3. Statistical distribution of PCE, Voc, Jsc and FF of PSCs with MAPI / C, MAPI / spiro-MeOTAD / C and MAPI / spiro-MeOTAD / Au structures; Figure-4. Statistical distribution of PCE, Voc, Jsc and FF of flexible and rigid PSCs with MAPI / PEDOT / C and MAPI / PEDOT / Au structures; Figure-5. a) Nyquist plots and b) EQE spectra of device structures SnO2 / MAPI / Carbon, SnO2 / MAPI / PEDOT / C and SnO2 / MAPI / PEDOT / gold and c) XPS spectra mapping of rigid PSCs with different thicknesses of PEDOT; Figure-6. Statistical distribution of PCE, Voc, Jsc and FF of PSCs with different thicknesses of PEDOT layer; Figure-7. a, b) steady-state PL spectra and J-V curves of MAPI / PEDOT / C, MAPI / PEDOT / Au and MAPI / Spiro-MeOTAD / Au structures, c) photovoltaic performance development pathway of PSCs with carbon top electrode, d) humidity stability test under 70% RH at 25 ℃ and e) thermal / humidity stability test under 85% RH at 65 ℃ for carbon and gold top electrode devices; Figure-8. Schematic visualization of R2R slot-die coating method for fabrication of PSCs with carbon top electrode; Figure-9. EL emission photography of a) small-scale flexible PSCs and b) R2R carbon electrode PSCs. Images in the middle are superposition of the images on the left and on the right. Also, images on the left are captured under illumination, no-bias applied. The images in the right are taken in the dark under bias; Figure-10. a) XRD spectra of MAPI (control), carbon stencil coated and R2R coated samples, b) cross-section FE-SEM, c) XPS spectra of PSCs with R2R v stencil carbon electrodes, d, e and f) J-V and J-t curves of r2r PSCs with evaporated gold versus fully R2R structure;Figure-11. Statistical distribution of PCE, Voc, Jsc and FF of PSCs R2R slot-die coated with evaporated Au versus carbon electrode PSCs fully R2R slot-die coated; Figure-12. Photographs of perovskite samples submerged in various solvents over 24 hour period. Figure-13. a) Photoluminescence measurement of perovskite samples coated with a carbon ink formulation and b) following 24 hour solvent soaking. Figure-14. X-ray diffraction analysis of perovskite following printing with different carbon ink formulations. Figure-15. Device efficiency of cells fabricated using 1-butanol inks with varied carbon loadings. Figure-16. Device efficiency of cells fabricated using 1-butanol carbon inks with varied concentration of Bathocuproine dopant. Table 1. Photovoltaic parameters of R2R coated perovskite solar cell studies published to date; Table 2. XRD results of MAPI as control, MAPI sample with stencil coated carbon and MAPI sample with R2R carbon coated; Table 3. Optimized 1-butanol ink composition; and Table 4. Optimized 2-methylanisole compositions. Materials and Methods Unless mentioned otherwise, all chemicals and solvents including lead iodide (PbI2, TCI chemicals, 99.99%), methylammonium iodide (MAI, greatcell solar, >99% anhydrous), tin (iv) oxide colloidal dispersion liquid (Alfa Aesar, 15% in H2O), PEDOT (HTL SOLAR3, Ossila), carbon black and graphite (Imerys), 1-butanol (Scientific Laboratory Supplies, 99.8% anhydrous), methylamine solution (Fisher Scientific, 33% in ethanol), 2-Methylanisole (Alfa Aesar, >99%) were purchased and used without further purifications. Other chemicals and solvents including spiro-MeOTAD powder (HPLC, 99%), Lithium bis(trifluoromethylsulfonyl)imide salt (Li-TFSI, 99.99%), 4-tert- butylpyridine (t-BP, 96%), FK209 Co(iii) TFSI salt, ethyl cellulose, acetonitrile (99.8%anhydrous) and chlorobenzene (99.8% Anhydrous) were all purchased from Sigma- Aldrich. Characterization For full device photovoltaic performance measurements, a Keithley 2400 source meter and a solar simulator (Newport oriel Sol3A) was used. J-V and stabilized current (J-t) curves (for a period of 300 s) were measured at 1 sun after calibration. The active area was 0.09 cm2for all devices, which was defined by a mask on top of the measuring module. EQE curves were measured as a function of wavelength range 300-800 nm on a custom-built system consisting of Xenon arc and Quartz halogen white light sources and a Bentham TMc300 monochromator. Steady state PL spectra were obtained using FS5 spectrofluorometer Edinburgh instrument with an excitation wavelength of 430 nm and a 496 nm long-pass filter in the emission pathway. Thicknesses of layers were measured using a KLA Tencor D-600 profilometer. For measuring the thickness of layers on flexible substrates, an air table was used to avoid any vibration during the measurement, which in turn reduced noises in resultant profile. The morphology of films grown on a glass substrate was studied using a Hitachi TM3000 desktop SEM and JEOL-JSM-7800F field emission scanning electron microscope (FE-SEM) (5 kV acceleration voltage, a working distance of 10 mm and a magnification of x 1500). The cross-section of films grown on PET substrate were achieved in a Zeiss Crossbeam 550 FIB FEGSEM. The Ga source FIB was operated in standard configuration with the sample surface perpendicular to the Ion Beam, tilted 54 degrees to the electron beam column. A high current 30kV 15nA probe was used to rough trench the area, before the cross-section face was developed using lowering 30kV probes to 30kV 700pA and a 55.4 degree tilt. The images then were recorded with InLens secondary electron detector and backscattered electron images achieved with a filtered ESB detector. The XPS analysis was carried out with a using a Kratos axis supra XPS equipped with a minibeam 6-gas cluster ion source. XRD scans were undertaken using a Bruker D8 advance instrument using a copper source, in Bragg- Brentano geometry, between 10 and 50° with a 0.02° step size and time of 1 second per step. Electrochemical impedance spectroscopy analysis was performed using a Zahner CIMPs-X photoelectrochemical workstation to attain the Nyquist plots of the equivalent circuit used for impedance and capacitance data. Measurements wereperformed under dark situation over the frequency range 1 MHz to 1 Hz at potentiostat DC voltage of 0.8 V. The electroluminescence images were taken with a modified Canon EOS 600D camera without IR filters in the dark while a constant voltage was applied to the devices with a RS Pro Bench Power Supply. Example 1: Device Architecture To achieve Roll to Roll processed PSCs, a low-temperature solution-based process for depositing each of the device layers is required, which can be coated easily on flexible substrates. Tin oxide has been demonstrated as an effective low temperature processed electron transport layer suitable for slot die coating

[0028] in combination with perovskite and Spiro-MeOTAD layers in PSCs. Along with tin oxide, perovskite precursor dissolved in acetonitrile has been successfully used as an excellent low- toxic ink for R2R coating[25, 28]. However, the main challenge remains coating a top electrode in a continuous, sequential R2R process. Therefore, to obtain the intended carbon ink formulation for such use, the compatibility, conductivity and scalability of such carbon-containing inks was assessed. The carbon ink formulation comprises three components: solvent that will challenge compatibility with perovskite; binder that will control rheology, an important consideration for scalability; and conductive carbon, which will alter both rheology and conductivity of the ink once dried. Solvent compatibility was first assessed by soaking coated perovskite layers in a series of candidate solvents for a 24hr period, after this time the samples were visually inspected and analyzed using photoluminescence spectroscopy (Fig.1a). Quenching ability of carbon inks formulated from these solvents are shown in Fig. 1b. 2- Methylanisole showed no colorimetric change and has PL matching that of the unsoaked perovskite control, it has a suitable boiling point of 170°C allowing low temperature curing while remaining wet through processing. Further, 2-methylanisole can be considered a green solvent as there is no significant toxicity and therefore no associated workplace exposure limits making it suitable for scale up and use in large volumes. An exemplary ink was therefore formulated by combining 2-methylanisole, ethylcellulose binder, carbon black and graphite conductive components. Details onhow to prepare, develop and optimize a carbon ink formulation are given in the appendix. The potential of the optimized carbon ink formulation was then assessed, first by evaluating the performance of small-scale spin coated perovskite devices including a carbon top electrode. Devices fabricated using spin coating of planar n-i-p structures ITO / SnO2 / MAPI / Carbon and ITO / SnO2 / MAPI / Spiro / gold (as device control) where the gold electrode is evaporated, and the Carbon electrode applied by stencil coating a viscous paste. Fig.2a shows the schematic of carbon coating method and fabricated devices’ photographs. To coat the carbon by the stencil method, PET-patterned stencils were made by laser cutting with identical size and shape as the shadow-mask used for Au evaporation. The resultant devices reached champion efficiency of 18.9% for gold electrode; however, the carbon electrode devices produced only around 3% efficiency and depressed J-V characteristics (Fig. 2c and Fig. 3). Based on the results, it would appear that direct charge transfer between perovskite and carbon has serious issues. Since the carbon layer has the low sheet resistance of 5.28 Ω / square and the SEM image of its morphology (Fig.2d) does not show many voids, it is likely that the charge transfer issue was related to the interface between these two layers. On that account, spiro-MeOTAD, which is the most popular hole transporting material, was spin coated to make the structure ITO / SnO2 / MAPI / Spiro / carbon. However, due to incompatibility with the carbon ink the same result was achieved. Compared to spiro-MeOTAD / gold devices, steady-state photoluminescence (SSPL) measurements in fig. 2b show poor quenching further identifying the poor charge transfer between spiro-MeOTAD and carbon. This is an indication of poor compatibility between the hole transport layer and the carbon ink formulation. From SSPL graphs and J-V curves similarity for samples with / without spiro-MeOTAD, it was determined that carbon ink incompatibility leads to dissolving / removing spiro layer underneath the carbon. Afterwards, commercially available and easy to coat PEDOT was employed as hole transport material as a substitute for spiro-MeOTAD. The J-V measurement results of Fig.2e, shows that PEDOT replacement had a significant role on device performance due to its compatibility with the carbon ink. The device efficiency reduced for evaporated gold electrodes, however it is greatly improved for stencil printed carbon electrodes achieving 13.25% champion (14.57% stabilized), full statistical datais in Fig.4 also showing the consistency of the results. Moreover, SSPL measurement demonstrated efficient quenching (Fig.2f), suggesting the necessity of using PEDOT or any other equivalent compatible hole transporting material between perovskite and the carbon top electrode. The results confirmed that PEDOT not only acts a hole transport layer compatible with carbon ink solvent, but it is also intrinsically conductive aiding charge transfer to the printed carbon electrode minimizing recombination losses. The compatibility of the carbon electrode with the device structure is further demonstrated by Electrochemical Impedance Spectroscopy (Fig 5a), the series resistance derived from the Nyquist plots is remarkably low in value for gold and carbon electrodes (7.5 and 14.8 ohm, respectively) indicating good charge transfer at the electrode interfaces. The corresponding external quantum efficiency(EQE) spectra of the three devices are shown in Fig.5b. As expected, due to its low charge extraction the perovskite / carbon stack without interlayer has low EQE% (highest: ~30%) in the whole wavelength range. Due to the fact that increase in overall QE curve is normally caused by higher charge extraction / lower recombination, by coating PEDOT between perovskite / carbon or perovskite / gold the charge transfer from the absorbing layer toward the contact electrode is increased significantly. In the wavelength range 400-600 nm, both PEDOT / carbon and PEDOT / gold devices have relatively similar EQE%. However, from 600 to 700nm there is a reduction in EQE observed for the PEDOT / carbon stack. Since the long wavelength response is related to the rear side of the device this reduction explains the higher quality of interface and reflection between PEDOT / gold rather than PEDOT / carbon, which is to be expected. In consideration of scaling up to R2R coating of the device structure, it is important to understand the optimal thickness of the PEDOT interlayer as thin coated layers will be susceptible to defects and more robust thicker layers detrimental to series resistance. Therefore, PEDOT was coated with a thickness range from 60- 140nm with 90nm found to be optimum (Fig.6). X-Ray photoelectron spectroscopy (XPS) lead elemental mapping of 28x28mm ITO / SnO2 / MAPI / PEDOT coated substrates revealed a narrow range of atomic lead (0.83%wt maximum), suggesting the perovskite was effectively covered with low defect occurrence at 90nm thickness. Despite less perovskite beingobserved for the thicker deposition (130-140nm), the lower defect rate does not result in improved efficiency (Fig.5C). EIS has shown that charge transfer at the interface between PEDOT and carbon is comparable to PEDOT / gold, and is confirmed by both the SSPL graphs and J-V data which demonstrate similar quenching and PV performance (Fig.7a, b). This confirms that the carbon electrode used in this study is an exceptional alternative to gold, however this is only achieved following development of device structure, formulation and processing optimisation (Fig.7c). Initially device efficiency of approx. 3% was achieved with only carbon (HTL free) and spiro-MeOTAD based cells, a modest improvement was made by optimising the carbon electrode formulation but by introducing the PEDOT HTL and then optimising PEDOT thickness (Fig. S5 and S6) efficiency was increased to 12.5%, and final parameter optimisation for the stencil coated carbon electrode resulted in the optimised device efficiency of 14.57%. Typically, conventional PEDOT: PSS as an interlayer is known to be hygroscopic, and so is likely to introduce poor device stability, however stability testing demonstrates the commercially available HTL Solar 3 PEDOT polymer mixture, when used in the PEDOT / carbon system has been found to counteract any instability issue. PEDOT / carbon and PEDOT / gold devices with no encapsulation under 70%RH / 25°C conditions have long-term stability of 80% retained efficiency over 1000h compared to the conventional spiro-MeOTAD samples that lost 50% of their efficiency over the same time scale (Fig.7d). The additional protection afforded by the carbon electrode is revealed in high humidity / temperature conditions. Thermal / humidity testing at 65℃ / 85% RH under dark condition (ISOS D3) shows high stability for un-encapsulated carbon electrode devices compared to evaporated gold electrodes, Fig.7e. The device architectures with evaporated gold electrode are outperformed by the PEDOT / Carbon electrode as the spiro / Au and PEDOT / Au decline to 20% and 40% of their original PCE within the first 15 minutes (T80193s and 236s), while the PEDOT / Carbon devices retain over 85% of the original efficiency for the duration of testing (60 minutes).Example 2: Roll to Roll Processed Devices In spite of the fact that most of the studies in the field of nano-structured solar cells have been focused on performance development of the devices, in recent years a few promising breakthrough reports have centred on scaling-up of these devices, specifically through R2R coating (Table 1). However, as of the time of writing this paper, none have reported fabrication of a whole R2R structure device including the top electrode. Inclusion of a solution R2R processed electrode is an important advancement as it enables, when combined with R2R encapsulation, the fabrication of large volumes of complete perovskite photovoltaics in a single high speed process - a key enabler in transitioning the many advancements in the perovskite research from the laboratory to the factory (lab to fab). Given that the prototype carbon electrode devices of Example 1 revealed the same performance results as gold electrodes, the carbon ink was modified by lowering the viscosity for compatibility with slot die coating, and its suitability for use in preparing PSCs by R2R coating was assessed. Flexible substrate-based n-i-p devices were prepared by using a pre-patterned 100 mm wide PET roll coated with ITO (sheet resistance of 50 Ω / sq). Fig.8 represents a schematic of the R2R coating process conducted with a coatema smartcoater. The procedure for each slot die coated layer includes three continuous steps: coating, annealing and rewinding of the substrate in preparation for the next layer coating. All device layers were sequentially slot die coated at a common coating speed of 1 m / min along with passing through ovens with a total length of 1m, at completion of coating no further processing was needed to complete the device structure. The definition of coating parameters for tin oxide, perovskite, PEDOT and carbon ink can be found in the Appendix. The functionality of the R2R coated devices was examined using Electroluminescence (EL) emission tests. In the case of having a flawless structure, electrical phenomenon would convert to light emission in response to the passage of electricity through the device, with brighter areas indicating better performing zones of the device. However, having flaws on some parts / points in each layer will lead to cessation of illumination on that area or even electrical disconnection between both sides of the circuit and will result in darkness for the whole device. Fig.9a shows a flexible stencil-coated carbonelectrode device. As can be seen three out of four stripes of small-scale stencil-coated device were working well and one had issues on its corner illustrating a coating defect. Furthermore, the photograph of a random cut device sample from our slot die coated R2R solar cell is shown in Fig.9b. The illumination along the length of the electrode indicating there are no disruptive defects along them. To ensure the R2R developed carbon ink is completely compatible with the perovskite layer underneath, XRD analysis was employed. For appropriate comparison three device structures were employed, the full device stack including PEDOT and carbon applied using both stencil coating and R2R slot die coating was then compared to perovskite (MAPI) coated on Tin Oxide as a pristine control. Relative peak intensity of the XRD spectra shows that stencil coated carbon samples and R2R devices have similar PbI2 concentrations relative to the MAPI peak, slightly lower than the pristine MAPI sample (Fig.10a). This suggests that not only are the carbon inks used for stencil coating and R2R slot die both compatible with MAPI but they have also shown protective properties against its degradation (Table 2). Furthermore the sequential R2R drying processes inside the ovens has not degraded the perovskite with crystallite size of 50-90nm, which is another sign of successful coating of carbon. Cross-section FE-SEM (Fig.10b1) and FIB-SEM (Fig.10b2) images of R2R slot-die coated devices show the carbon ink forms a compact layer free of voids that may be formed by solvent entrapment, the carbon electrode makes intimate conformal contact with the underlying PEDOT layer. XPS elemental lead mapping of R2R coated SnO2 / MAPI / PEDOT (Fig.10c) indicates that the coverage is effective and covers high points so as to prevent pinhole / shorting defects. The % lead in the XPS map is lower than identified in optimum spin coated devices (Fig 5c) highlighting the increased requirement to cover defects in R2R slot die compared to spin coating. In order to test the 20m long R2R printed device, the substrate was manually diced into segments similar in size to the small-scale devices of Example 1 and randomly selected samples measured conventionally. Sections without printed carbon electrode were also collected and gold electrode evaporated, although a light soaking effect was observed for the carbon electrodes the measurement protocol was maintained for both gold and carbon electrode devices. The R2R perovskite devices using gold evaporated electrode had a high PCE of 13.21% (12.87% stabilized PCE) comparableto the highest performing evaporated electrode R2R devices reported. For fully printed carbon electrode cells a PCE of 9.79% (10.84% stabilized PCE) was achieved. Based on J-V curve of hero cells in Fig.10d and statistical data of box plots presented in Fig.11, both fully R2R and R2R / gold devices have very close values for current density and open circuit voltage. However, fully R2R devices have shown lower fill factor compared to evaporated gold samples leading to lower PCEs. Moreover, it should be noted that although the performance of devices with PEDOT / carbon compared to spiro-MeOTAD / carbon is improved, it still suffers from the lack of appropriate fill factor compared to the conventional spiro-MeOTAD / gold devices. PEDOT is not the only compatible candidate carbon electrode interlayer but it has proven effective to make the first demonstration of fully R2R coated perovskite device. This strategy in combination with the compatible carbon ink formulation can be further extended to even better hole transport / interlayer materials, in particular nickel oxide, copper thiocyanate, copper indium sulphide, PTAA and / or P3HT, in principle to achieve higher performance fully roll to roll coated perovskite. Example 3: Additional Solvent Testing The suitability of various ink formulation solvents was conducted through a variety of compatibility tests. As an initial assessment, perovskite samples were soaked in solvents of interest. Specifically, a MAPI perovskite layer was coated onto glass substrates and the coated substrates were then placed into solvent filed petri dishes to fully submerge the coated substrates and samples were left undisturbed for 24hours. The results of these initial tests are shown in Figure 12. The development of a yellow colour indicated that the perovskite has deteriorated and the original dark brown MAPI colouration has been replaced by a yellow lead iodide colour. Significant perovskite deterioration was observed in M-cresol and terpineol solvent tests. Subsequently, photoluminescence investigation was carried out on perovskite layers printed using carbon inks formulated with different solvents along with perovskite layers soaked in solvent. Quenching of the signal indicated improved charge transfer and therefore improved performance i.e. closer to the control perovskite sample with no carbon ink or solvent soaking. The photoluminescence results are shown in Figure13 a) and Figure 13 b), where carbon inks formulated from diacetone alcohol (A), 1- Butanol (Buta), BCP doped 1-Butanol (Buta BCP), Terpineol (HH), o-xylene and 2- methylanisole (MA) are employed. Further, X-Ray diffraction (XRD) analysis of inks formulated with different solvents coated onto perovskite was carried out, with results shown in Figure 14. Characteristic peaks at 14° and 12.6° represent perovskite and lead iodide respectively, and a large peak at 26.5° represents the carbon of the ink. The formulation of perovskite used includes a 6% lead iodide excess and therefore a small lead iodide peak was expected in all samples. This ratio was maintained for the 1-butanol and 2-methylanisole solvents, however this ratio was reduced for diacetone alcohol and terpineol solvents, suggesting perovskite breakdown. Example 4: 1-Butanol Ink Optimisation Preliminary optimisation studies were carried out for 1-butanol ink compositions with varied carbon loading, the results of which are shown in Figure 15. All ink formulations comprised 12.5% ethyl cellulose binder with 1-butanol. Higher loading was 29.4% weight total carbon, lower loading 21.7% weight total carbon, and the specified ratios indicate the ratio of Graphite: Carbon black in the tested ink formulation. This figure demonstrates the optimum performance of the formulation disclosed in Table 3. In addition, Figure 16 demonstrates the results of perovskite device efficiency tests wherein 1-Butanol inks were doped with Bathocuproine at values of 1mg / ml to 10mg / ml showing the optimum performance at 10mg / ml. Example 5: 2-Methylanisole Ink Optimisation Preliminary optimisation studies were also carried out for 2-methylanisole ink compositions, wherein the efficiency of an n-i-p device stack of Tin Oxide / Perovskite (MAPI) / PEDOT / Carbon was tested for formulations produced for stencil coating and for Roll to Roll slot die coating, with the optimum performance being observed using the formulations disclosed in Table 4.SUMMARY This work presents a significantly low-cost and scalable carbon ink formulation replacing the conventional expensive lab-scale metal top electrodes. Further the application demonstrates the impact of a PEDOT layer on the perovskite / HTL / carbon interface modification. Benefiting from the combination of this HTL with a low viscosity, slot-die compatible, carbon ink, we fabricated the world’s first fully roll to roll flexible perovskite solar cells. The prototype carbon electrode small-scale devices on rigid glass substrates obtained a PCE of 13.25% and then extended to a fully R2R coating system with 20m long flexible substrates to yield the stabilized PCE of 10.8%. More importantly, devices fabricated with carbon electrodes provided similar photovoltaic performance as conventional evaporated gold electrodes with the hero device PCE of 13.26% and 13.61%, respectively. Moreover, carbon electrode devices show better long-term stability rather than gold, however, these devices have shown significantly more stability under severe thermal / humidity conditions. The use of PEDOT / carbon electrodes in R2R perovskite devices has been demonstrated to be an operational alternative to conventional expensive non-scalable spiro-MeOTAD / gold electrodes. This achievement is a leap forward in reaching the promise of perovskite photovoltaics in printing millions of meters of solar cells worldwide.Table 1 PCE( Device structure R2R layers Electrode Group [Ref]PEDOT:PSS / MAPbI3−x Clx All except Ag ( screen-printed) 4.9 DTU

[0022] / PCBM / ZnO / Ag electrode PET / ITO / PEDOT:PSS / MAPbI3 / PCB PEDOT & CSIRO / Ca & Al (evaporated) 11.2 M / Ca / Al perovskite CAS

[0027] SnO2 / Cs0.15FA0.85PbI2.55Br0.45 / Spiro / SnO2 & Au (evaporated) 13.5 Solliance

[0028] Au perovskite SnO2 & IZO / SnO2 / MAPbI3 / Spiro / Au Au ( evaporated) 14.1 NREL

[0029] perovskite PEDOT:PSS / triple cation All except GIST / Ag (evaporated) 11.7 psk / PCBM / PEIE / Ag electrode CSIRO

[0021] (evaporated) 9.9 CSU

[0030] All except SnO2 / FAPbI3 / P3HT / Au Au (evaporated) 13.8 KRICT

[0031] electrodeAll except Ag (evaporated) 12.2 SPECIFIC

[0025] g electrode PEDOT:PSS / CsFAMAPb(IBr) / PEDOT:PSS Monash / Au (evaporated) 12.0 PCBM / PEIE / Au and psk CSIRO

[0024] SnO2 / FAMAPbI3 / All except Monash / C / Cu / Al (pressed foil) 10.0 PPDT2FBT / C / Cu / Al electrode CSIRO

[0032] ET / TCE / SnO2 / CFAMPb(IBr) / All except Ag & C (bar coated- CSIRO / Spiro / carbon / Ag electrode pressed) Monash

[0026] All except CSU / SnO2 / MAPbI3 / Spiro / Ag Ag (evaporated) 13.0 electrode CSIRO

[0023] OPV8 / SnO2 / FA0.91Cs0.09PbI3 / Spiro / A All except 15.4 Au (evaporated) CSIRO

[0033] u electrode 2Table 2 MAPI (Control) with stencil coated carbon with R2R coated carbon Peak Intensity % Peak Intensity Peak Intensity PbI2 12.832 2006 5.12% 12.806 931 4.09% 12.836 1082 3.98% MAPI 14.312 37203 94.88% 14.291 21834 95.91% 14.299 26071 96.02% 39209 22765 27153 Table 3 Component % Weight 1-Butanol 68.5 Ethyl cellulose 9.8 Graphite 15.7 Carbon black 6.0 5 Table 4 Component Stencil Ink Slot Die Ink 2-methylanisole 68.5% 84.2% Ethyl cellulose 9.8% 4.9% Graphite 15.7% 7.8% Carbon Black 6.0% 3.0%Appendix (A) Rigid Device Fabrication Method ITO glass substrates (sheet resistance ≈ 15 Ω / sq) were washed / cleaned in an ultrasonic cleaner for 10 min each, with hellmanex cleaning concentrate (2% in DI water), DI water, acetone and 2-propanol, respectively. After drying by N2 blowing, the substrates were treated with UV-ozone cleaner for 15 min in regards of increase in their surface energy and wetting ability. To deposit the electron transporting layer, Tin(IV) oxide 15% in H2O colloidal dispersion diluted with water (4.2%wt SnO2) and then spincoated at 4000 rpm for 30 s and annealed at 150 ℃ for 30 min. Following this, the 0.73 M perovskite solution filtered by a 0.2 µm PTFE syringe filter and deposited at 3000 rpm for 60 s, then annealed at 120 ℃ for 10 min .The preparation details of perovskite MAPI solution in acetonitrile can be found in our previous work

[0025] . In brief the perovskite inks were prepared using a methylamine gas bubbling method and acetonitrile solvent. Typically 14.23g lead iodide and 4.63g Methylammonium iodide were weight into a flask and 40ml of anhydrous acetonitrile added in a nitrogen environment. The flask is sealed and removed to ambient environment where under stirring a nitrogen gas flow is bubbled through a methylamine solution (33% in ethanol) carrying the methylamine in the gas flow through a dryerite filled drying tube into the perovskite precursor solution. Under bubbling of methylamine into the solution the precursor dissolve transforming from a black suspension to a light yellow solution. At this point the bubbling was stopped and perovskite precursor ink collected. Latterly the solution has been formulated using a commercially available ACN / MA solution (TCI chemicals) avoiding the need for bubbling the precursor solution with MA gas. In both cases the solution is refrigerated during storage and used within 8weeks of manufacture. Then, on top of perovskite absorbing layer, PEDOT solution (in toluene) spincoated at 3000 rpm for 40 s and annealed at 110 ℃ for 10 min. For devices with spiro-MeOTAD as HTL, 90 mg / ml spiro powder dissolved in chlorobenzene. Then, 20 µl LiTFSI (520 mg / ml in ACN), 30 µl 4- tert-butylpyridine and 10 µl FK209 (300 mg / ml in ACN) added as additives and the final prepared solution stirred at 60 °C. The prepared solution was then spincoated at 4000 rpm for 30 s without further annealing. Finally, Au was evaporated on the hole-transporting layer as the top electrode. Alternatively, for devices with carbon electrode, carbon ink coated by stencil coating and then dried at 110 °C for 10 min. (B) Exemplary Carbon Ink Preparation Ethyl cellulose and 2-methylanisole are weighed and mixed under high shear using a Speedmixer twice at 3500rpm for 5min, the solution is left for 24hr to ensure the binder is fully dissolved. Carbon black (0.77g Stencil ink, 0.34g slot die ink) and graphite (2g Stencil ink, 0.88g slot die ink) components are then weighed into the resin and again mixed in the Speedmixer (2x 3500rpm-5min) followed by standing for 24hr to allow solvent absorption by the carbon and achieve the final viscosity of the ink. Finally, the ink is again mixed twice at 3500rpm for 5min breaking any remaining agglomerates and ensuring a smooth and well-combined ink. (C) Roll to Roll Device Fabrication Method Device layers are sequentially slot die coated onto 50Ω / sq ITO at a common speed of 1m / min using a Cotema Smartcoater roll to roll coating system. Tin oxide (Alpha Aesar) diluted to 1.2%wt with DI water and 10% 1-Butanol was coated at 90mm width and 5-11µm (7µm optimum) wet film thickness with 1mm meniscus guide, 200µm gap and 140°C drying temperature. Next perovskite solution (as used in rigid device fabrication) is coated at 90mm width and 3-9µm (7µm optimum) wet film thickness, 1mm meniscus guide 200µm gap an airknife is employed directly following coating with 50l / min nitrogen flow then dried at 150°C oven setpoint. PEDOT (HTL solar 3) is coated as supplied at 3-6µm (4µm optimum) wet film thickness and 90mm width using a 0.25mm meniscus guide, 150µm gap and 140°C drying temperature. Carbon ink was stripe coated at 2x 3mm to define the electrode pattern, no meniscus guide was used, coating thickness was 250µm wet film and oven setpoint 140°C. Oven residence time for all coated layers was 1minute.References

[0021] J. E. Kim, S. S. Kim, C. Zuo, M. Gao, D. Vak, D. Y. Kim, Adv. Funct. Mater.2019, 29, 1809194.

[0022] T. M. Schmidt, T. T. Larsen‐Olsen, J. E. Carlé, D. Angmo, F. C. Krebs, Advanced Energy Materials 2015, 5, 1500569.

[0023] H. Li, C. Zuo, D. Angmo, H. Weerasinghe, M. Gao, J. Yang, Nano-micro letters 2022, 14, 1.

[0024] M. Othman, F. Zheng, A. Seeber, A. S. Chesman, A. D. Scully, K. P. Ghiggino, M. Gao, J. Etheridge, D. Angmo, Adv. Funct. Mater.2022, 32, 2110700.

[0025] D. Burkitt, R. Patidar, P. Greenwood, K. Hooper, J. McGettrick, S. Dimitrov, M. Colombo, V. Stoichkov, D. Richards, D. Beynon, Sustainable Energy & Fuels 2020, 4, 3340.

[0026] L. J. Sutherland, D. Vak, M. Gao, T. A. N. Peiris, J. Jasieniak, G. P. Simon, H. Weerasinghe, Advanced Energy Materials, 2202142.

[0027] C. Zuo, D. Vak, D. Angmo, L. Ding, M. Gao, Nano Energy 2018, 46, 185.

[0028] Y. Galagan, F. Di Giacomo, H. Gorter, G. Kirchner, I. de Vries, R. Andriessen, P. Groen, Advanced Energy Materials 2018, 8, 1801935.

[0029] B. Dou, J. B. Whitaker, K. Bruening, D. T. Moore, L. M. Wheeler, J. Ryter, N. J. Breslin, J. J. Berry, S. M. Garner, F. S. Barnes, ACS Energy Letters 2018, 3, 2558.

[0030] C. Gong, S. Tong, K. Huang, H. Li, H. Huang, J. Zhang, J. Yang, Solar RRL 2020, 4, 1900204.

[0031] Y. Y. Kim, T.-Y. Yang, R. Suhonen, A. Kemppainen, K. Hwang, N. J. Jeon, J. Seo, Nature communications 2020, 11, 1.

[0032] G. A. Sepalage, H. Weerasinghe, N. Rai, N. W. Duffy, S. R. Raga, Y. Hora, M. Gao, D. Vak, A. S. Chesman, U. Bach, Advanced Materials Technologies 2022, 7, 2101148.

[0033] J. F. Benitez-Rodriguez, D. Chen, A. D. Scully, C. D. Easton, D. Vak, H. Li, P. E. Shaw, P. L. Burn, R. A. Caruso, M. Gao, Sol. Energy Mater. Sol. Cells 2022, 246, 111884.

[0035] S. M. Meroni, C. Worsley, D. Raptis, T. M. Watson, Energies 2021, 14, 386.

[0041] H. Zhang, J. Xiao, J. Shi, H. Su, Y. Luo, D. Li, H. Wu, Y. B. Cheng, Q. Meng, Adv. Funct. Mater.2018, 28, 1802985.

[0042] X. Wu, L. Xie, K. Lin, J. Lu, K. Wang, W. Feng, B. Fan, P. Yin, Z. Wei, Journal of Materials Chemistry A 2019, 7, 12236.

[0043] V. Babu, R. Fuentes Pineda, T. Ahmad, A. O. Alvarez, L. A. Castriotta, A. Di Carlo, F. Fabregat-Santiago, K. Wojciechowski, ACS Applied Energy Materials 2020, 3, 5126.

[0044] C. Zhang, S. Wang, H. Zhang, Y. Feng, W. Tian, Y. Yan, J. Bian, Y. Wang, S. Jin, S. M. Zakeeruddin, Energy Environ. Sci.2019, 12, 3585.

[0045] G. A. Sepalage, H. Weerasinghe, N. Rai, N. W. Duffy, S. R. Raga, Y. Hora, M. Gao, D. Vak, A. S. Chesman, U. Bach, Advanced Materials Technologies 2021, 2101148.

Claims

Claims 1. A carbon ink formulation comprising: a. one or more binders; b. one or more populations of conductive carbon particles; and c. one or more solvents, wherein said solvent is selected from ethanol, ethyl acetate, 1-butanol, 2- methylanisole, xylene, toluene, cyclohexanone, indane or any isomer thereof.

2. The formulation according to claim 1, wherein said solvent is 2-methylanisole or 1-butanol.

3. The formulation according to claim 1 or claim 2, wherein said formulation is substantially free of water.

4. The formulation according to any of claims 1 to 3, wherein said conductive carbon particles are selected carbon black, graphite particles and mixtures thereof.

5. The formulation according to claim 4, wherein said conductive carbon particles comprise a mixture of carbon black and graphite particles, and wherein the weight ratio of graphite to carbon black is from about 1.5:1 to about 4:

1.

6. The formulation according to any of claims 1 to 5, wherein said binder comprises one or more hydrophilic polymer, and wherein said hydrophilic polymer is optionally selected from: polyvinyl alcohol, polyvinyl pyrrolidone, gelatin, celluloses, pectins, polyoxazolines, polyvinylacetamides, partially hydrolyzed polyvinyl acetate / vinyl alcohol, polyacrylic acid, polyacrylamide, polyalkylene oxide, sulfonated or phosphated polyesters and polystyrenes, casein, zein, albumin, and derivatives thereof.

7. The formulation according to claim 6, wherein said binder comprises a cellulose ether, optionally a carboxymethyl cellulose ether having a Mw of from about 10,000 to about 200,000 Da.

8. The formulation according to any of the preceding claims, wherein said formulation comprises from about 15 wt.% to about 35 wt.% carbon particles and about 5 wt.% to about 15 wt.% binder.

9. The formulation according to any of the preceding claims, wherein said formulation has a shear viscosity of between about 15 Pa.s and about 1.5 Pa.s, over a shear range of 0.1 to 1000 s-1.

10. The formulation according to any of the preceding claims, wherein said formulation is doped with one or more dopant, and optionally from about 0.5 mg / ml to about 20 mg / ml of said dopant(s).

11. The formulation according to claim 10, wherein said dopant is bathocuproine.

12. A planar structural perovskite solar device stack comprising: a. a transparent conductive layer; b. at least one electron transport layer; c. at least one perovskite layer; d. at least one hole transport layer; and e. a conductive carbon layer.

13. The device stack according to claim 12, wherein: a. said conductive layer comprises an ITO coated polymer; b. said electron transport layer comprises SnO2; c. said perovskite layer comprises a methylammonium lead halide; andd. said at least one hole transport layer comprises nickel oxide; copper thiocyanate; copper indium sulphide; poly[bis(4-phenyl)(2,4,6- trimethylphenyl)amine (PTAA); poly(3-hexylthiophene-2,5-diyl (P3HT); a poly(3,4-ethylenedioxythiophene) polymer mixture; or any combination thereof.

14. The device stack according to claim 13, wherein said hole transport layer has a thickness from about 40 to about 200 nm.

15. A method of manufacturing a device stack according to any of claims 12 to 14, wherein said electron transport layer(s), perovskite layer(s), hole transport layer(s), and conductive carbon layer(s) are sequentially coated onto a transparent substrate via a solution processing method.

16. The method according to claim 15, wherein said layers are coated onto the transparent substrate by slot-die coating, and optionally wherein each of the electron transport layer, perovskite layer, hole transport layer and conductive carbon layer are coated at a coating speed of from about 0.1 m / min to about 1000 m / min.

17. The method according to claim 15 or claim 16, wherein, following application of each layer to the stack and prior to application of any subsequent layer, the coated substrate is dried by exposure to heat, optionally a temperature of between about 120 °C and about 140 °C, and / or for a period of between about 30 seconds and 2 minutes.

18. The method according to any of claims 15 to 17, wherein: the transparent substrate is a flexible substrate such as PET or PEN, and is optionally coated with a Tin doped indium oxide (ITO) film; and wherein the device stack is formed by roll to roll processing.

Images