Among various architectures of perovskite solar cells (PSCs), the carbon-based perovskite solar cell (C-PSC) is an interesting architecture due to the low-cost, easy, scalable, and fully printable manufacturing process. The hydrophobicity of carbon materials as a top contact can significantly improve the stability of the photovoltaic device. This type of solar cell consists of a patterned transparent conductive oxide substrate (TCO), a blocking, compact layer (BL or CL), a mesoporous metal oxide as an electron transport layer (m-ETL), either a large bandgap metal oxide as an insulating separator layer or a hybrid perovskite layer, and a (mesoporous) carbon layer as a top contact; the mesoporous layers are impregnated with the hybrid perovskite absorber material. Depending on the cell architecture, one or two layers may or may not be present in the structure.
For instance, the m-ETL and/or the insulating layer are absent in the planar architecture, in which the BL acts as an ETL. Additionally, extra layers can be incorporated between the m-ETL and the perovskite layer, or between the perovskite and carbon layers to improve the device performance. C-PSCs are divided into two types: low-temperature and high-temperature. In low-temperature C-PSCs, a carbon paste with low curing temperature (<120 °C) is deposited after perovskite deposition and crystallization. In contrast, in high-temperature C-PSCs, a carbon paste is deposited and sintered at 400 °C. Then, after cooling down to room temperature, the perovskite precursor solution is deposited by drop casting and further annealed. Independent of the process fabrication temperature, the working principle of the C-PSCs is as follows: photogenerated electrons in the conduction band (CB) of the perovskite material are injected into the CB of the ETL metal oxide, while holes are extracted from the perovskite valence band (VB) by the carbon electrode. For efficient charge injection, transport, and extraction, the energy levels of the different functional layers in the cell structure must be compatible.
Therefore, one of the main objectives of research on C-PSCs is to achieve higher VOC, and to enhance charge separation, extraction, and transport properties. These goals can be achieved by using novel ETMs, modifying the existing materials or using novel inorganic HTMs and/or cost-effective HTLs between the perovskite and carbon electrode.
To fabricate C-PSCs several printing techniques such as screen printing, inkjet printing, spin coating, slot-die coating, blade coating and spray coating have been used. Each technique has its own advantages and disadvantages. For example, the most efficient PSCs have been manufactured using spin coating; however, this technique is mainly suitable for small scale and batch fabrication. In contrast, screen printing, spray, blade coating and slot-die coating have been used in the manufacture of large photovoltaic devices, although with lower performance due to the imperfection and defects of the final deposited layers. The reports on fully printed devices by only one technique are few. Usually, for the fabrication of the entire C-PSCs more than one printing technique is used. This is due to the fabrication and operational restrictions of some printing techniques. For example, preparation of aqueous binder-free pastes suitable for screen-printing is still challenging and this technique is particularly used for deposition of non-aqueous viscous pastes. Another example is the inkjet printing technique in which the size of dispersed nanoparticles of the ink is critical to avoid nozzle blockage. Carbon pastes with micron-sized graphite sheets are not suitable to be used in this technique. In spin-coated devices, commonly all layers are deposited using this technique except the carbon layer which is deposited by screen printing, blade or brush painting and spray coating. In screen-printed devices the blocking layer is mostly deposited by spray pyrolysis and the perovskite layer by drop-casting. The choice of the printing technique highly depends on the rheological properties of the original ink, paste or colloid and the solvent in which the nanoparticles are dissolved or dispersed.
2. Printing Techniques
2.1. Spin Coating
The spin coating technique produces thin, consistent coatings, which has enabled the quick manufacture of effective and reliable PSCs. An extensive variety of coating solutions and substrates can be used in this technique. The thickness of the coating can be controlled from a few nanometers to a few micrometers. The final coating properties such as thickness and homogeneity depend on the fluid rheological properties such as viscosity and thixotropy as well as surface tension. Other experimental parameters such as dispense volume, solution concentration, spin speed and time, and acceleration/deceleration rate impact the final coating properties
[30][31][31,32].
The spin coating process consists of four main steps: deposition, spin-up, spin-off, and evaporation
[32][33]. During the deposition step, a solution falls on a rotating substrate (dynamic spin coating) or fixed substrate (static spin coating) typically using a pipette or syringe. The solution spreads on the substrate due to the centrifugal or gravitational forces and, depending on its rheological properties, it may cover or wet the substrate completely. During the spin-up, the substrate is accelerated up to its desired rotational speed. At the very beginning, the substrate rotates faster than the fluid. While rotating, the fluid is radially driven out from the substrate due to the centrifugal forces. When the substrate reaches its desired speed, the fluid rotates at the same speed as the substrate and the film thickness decreases. In the third stage, the fluid rotates at a constant rate, flows out to the substrate perimeter, and continues thinning gradually. If the fluid contains a volatile solvent, usually it is possible to observe a color change in the deposited film. Finally, the fluid spin-off is stopped, and the thinning is dominated by solvent evaporation and film drying
[31][32][32,33].
To
the our
esearchers' knowledge, the highest reported efficiency for HTM-free C-PSCs fabricated by the spin coating technique (except the carbon layer) is 17.42%, and corresponds to the following architecture: BL-TiO
2/m-TiO
2/MAPI/C(commercial paste)
[33][34]. The highest efficiency for a cell with insulating layer is 9% and corresponds to BL-TiO
2/m-TiO
2/ZrO
2 (homemade paste)/MAPI/C (homemade paste) structure
[34][35].
As explained previously, fluid rheology has a strong influence on the quality of the deposited layer. For photovoltaic devices fabricated by spin coating, a choice of solvent for ink preparation is crucial and depends on the nature of the solute. For example, to deposit the blocking layer, a molecular ink is generally used in which the precursor is completely dissolved. For metal oxide mesoporous ETL or the insulator layer, a nano-ink is formulated in which the nanoparticles are dispersed into the appropriate solvent
[35][36][36,37]. In the most common spin-coated C-PSCs, the insulating layer is not present, and the structure is FTO/BL or m-ETL/perovskite/carbon. However, as mentioned before, the carbon layer is usually not deposited using this technique.
2.2. Screen Printing
Screen printing is a widely used technique for imprinting designs onto a variety of substrates, such as paper, textiles, polymers, glass, or any solid material
[37][82]. This technique involves pushing a viscous paste (or ink) through the orifices of a stencil screen with the help of a rubber blade (known as a squeegee)
[38][83]. This method has a long history, with its origins believed to be in China and Egypt
[39][84]. Nowadays, it is a particularly useful method for manufacturing scalable, high-speed printed microelectronics, as it is cost effective, easy to operate, and can be scaled up for larger production.
The key components of the screen-printing process that affect the resulting film are (i) the stencil, (ii) the squeegee, and (iii) the inks or pastes. The stencil is usually made of cotton, natural silk, plastic (nylon or polyester), or woven metal fibers; with plastic and metallic filaments being the preferred materials due to their durability and chemical resistance. The number of openings in the screen per linear inch (mesh count) also affects the quality of the pattern and the film thickness
[38][83]. To create the stencil, a light-sensitive emulsion is applied to the mesh, with the net covering only the non-printable areas
[37][82]. The squeegee is a blade made from a polymer with synthetic materials such as polyvinyl or polyurethane typically used for their sharpness and durability. The squeegee is used to control the spread of paste through the mesh and to adjust the film thickness
[37][82]. The inks or pastes need to have the correct rheology to be able to infiltrate the mesh, and are composed of a solvent or solvent mixture, material particles, and organic binders. The proportions of these components depend on the type of substrate to be imprinted. Commonly used solvents are water and organic solvents, while binders are chosen to match the desired viscosity and functionality
[40][85]. Despite the fact that a considerable number of studies have been conducted on screen printing of electronic materials, the operator must still adjust the deposition parameters and paste formulation in order to achieve the best possible coating. Here,
thwe
researchers will look back at several studies related to paste formulation in C-PSCs.
In 2013, Ku et al., reported the first mesoscopic C-PSC using MAPI perovskite
[41][86]. This triple-stack C-PSC was printed using the screen-printing technique and the layers were composed of TiO
2, ZrO
2 or alumina, and carbon. As mentioned before, researchers usually use commercially available pastes such as TiO
2 paste with particle size around 30 nm
[13][14], and it is further diluted with terpineol in 1:1 to 1:7 ratios to attain the desired thickness. Commercial ZrO
2 paste is composed of particles with diameters between 20 nm to 40 nm
[14][15] and, depending on the stencil and dilution employed, the layer thickness can be changed from 1.5 μm to 3.0 μm. Carbon paste is also available commercially
[14][15] and, as mentioned before, it consists of graphite and carbon black in different ratios. The carbon layer thickness must be in the range of 8 μm to 20 μm.
2.3. Inkjet Printing
Inkjet printing technology has been widely used in photovoltaic device fabrication since it allows to print the active materials on the solar cell stack
[42][117]. Inkjet printing provides flexibility in thickness and shape, and it is considered a material-efficient technology. The key requirements for ink materials are low viscosity, fine particle size if the ink is a colloid or a suspension, and low volatility to avoid nozzle clogging
[43][44][118,119]. Inkjet printing can be divided into two modes of operation, continuous (CIJ) and Drop-on-Demand (DOD). In a CIJ printer, charged droplets are controlled by an electrostatic field during the printing process. The imaging droplets pass through to the substrate while other droplets are deflected to an ink catcher for re-use. A DOD printer ejects ink droplets when a pulse of voltage is applied, without any drop deflecting and catching. Generally speaking, the CIJ printer has a higher jetting frequency, and the DOD printer has a much simpler inkjet head structure
[43][45][118,120]. There are two principal mechanisms of propelling ink drops in an inkjet printer head, piezoelectric propelling, and thermal bubble propelling. In the piezoelectric inkjet head, a voltage is applied to the piezoelectric pressure transducer to make it bend or change shape, which repels the ink out of the nozzle. Drop on demand inkjet printing can be used for the fabrication of C-PSCs by printing all the oxide layers in the stack as well as the organo-metal halide absorber
[46][47][48][121,122,123].
In 2007, the first organic solar cell fully fabricated by inkjet printing technique was reported with low efficiencies from 3 to 4%
[49][50][51][125,126,127]. In 2014, researchers started using inkjet printing technology to print the functional materials to fabricate PSCs
[52][53][54][128,129,130]. Even though studies on inkjet-printed PSCs are few, substantial achievements have been reported
[55][56][131,132]. Efficiencies higher than 21% have been recently reported for conventional metal electrode devices
[57][133]. Here,
thwe
researchers provide several examples of inkjet-printed C-PSCs and the effect of the ink or printing parameters on the device performance. In 2016, Hashmi et al. reported infiltration of the perovskite precursor solution using inkjet printing in the HTM-free carbon counter electrode-based PSCs
[58][134]. The inkjet infiltration of the perovskite precursor solution was performed for a triple-stack (TiO
2/ZrO
2/Carbon) mesoporous substrate using the perovskite precursor ink containing 5-AVAI. They used 5-AVAI because it significantly slows down perovskite crystal growth before and after the deposition of the precursor ink, thus preventing the inkjet printer cartridge from clogging, and also provides an opportunity for precise patterning and controlled volume dispensing of precursor ink. The fabricated devices had an efficiency of 8.1% when infiltrated with perovskite precursor in an area of 0.16 cm
2 using an inkjet printer, and after three weeks of storage in the dark under vacuum showed a high stability and a performance increase to 9.53%.
In 2020, Verma et al. achieved inkjet-printed carbon-based solar cells with four layers out of five, i.e., c-TiO
2, m-TiO
2, and m-ZrO
2, as well as the perovskite precursor ink, using environmentally friendly non-halogenated solvents
[59][135]. For the c-TiO
2, they formulated an ink from titanium diisopropoxide bis(acetylacetonate) (TAA) using an appropriate dilution with binary mixtures such as ethylene glycol: iso-propanol (IPA), tetralin: lPA, terpineol: IPA, ethyleneglycol: ethanol, tetralin: ethanol, and terpineol: ethanol. The TAA/terpineol: xylene mixture 1:16 vol/vol yielded stable jetting as well as very homogeneous wet and dry film formation. The final ink composition had a viscosity of 2.1 mPas and a surface tension of 23.1 mN/m. The m-TiO
2 ink was developed by diluting the commercial screen printing paste (Solaronix) into a terpineol: IPA mixture. For the m-ZrO
2, the Solaronix paste was diluted with a binary solvent. The ink with 6 vol of screen-printing paste in 5:5 (vol:vol) of binary solvents gave stable jetting with no satellite formation as well as homogenous layer formation. The oxide layers and the carbon electrode were first sintered at elevated temperature and then infiltrated with the standard MAPI by inkjet printing at room temperature. After infiltration and drying, the devices were post-treated in damp heat for 100 h. This treatment increases the crystallinity of the perovskite layer. They obtained solar cell devices of an active area of 1.5 cm
2 with an efficiency of 9.1%.
In 2021, Karavioti et al. provided a direct comparison between spin-coating and inkjet-printing techniques for the fabrication of fully ambient air-processed perovskite absorbent layers for C-based HTL-free PSCs
[60][136]. The results showed that the inkjet-printed perovskite layer presented a discontinuous morphology due to a “coffee-ring” effect unlike in the case of spin-coating, where a uniform morphology was obtained. This is mainly because in inkjet-printing, the applied perovskite precursor solutions should be of a much higher concentration compared to the corresponding solutions used for spin coating. In inkjet printing, the solvent evaporation rate is lower than in the case of spin coating, leading to a poor crystal structure, whereas in the case of spin-coating most of the solvent is quickly removed through centrifugal forces. Another reason is related to the wettability of the substrate by the ink, in particular, over-wetting. Moreover, the FTO/c-TiO
2/m-TiO
2/perovskite system fabricated by inkjet printing presented a lower light-harvesting efficiency (LHE) compared to cells fabricated by spin coating. These differences had an impact on the electrical characteristics of the solar cells. The devices fabricated by the inkjet printing technique achieved an efficiency of 8.40% compared to 10% for the solar cells fabricated by spin coating.
Chalkias et al. devised a strategy based on the jet-ability of the ink and on the wettability of the substrate with the ink to reduce “coffee ring defects” that appeared during ambient air inkjet printing processing of the PSCs
[61][137]. With different concentrations of the perovskite precursor ink (up to the limits determined by the solubility of the solutes in the solvent used), better charge transportation and recombination kinetics and an improvement of the external and internal quantum efficiency of the developed solar cells were achieved. C-based HTL-free PSCs using the optimized ink (1.8 M of perovskite precursor) reached an efficiency >12% for a cell with an active area of 0.34 cm
2, which is among the highest reported values for the architecture c-TiO
2, m-TiO
2 and perovskite layers deposited by inkjet printing. Finally, by the combination of inkjet printing and screen printing perovskite sub-modules of 34.2 cm
2 with an average efficiency value of 9.09% were developed to demonstrate the scalability of the technology. While the ink properties play an essential role in determining the solar cell performance, the fluid dynamics calculation of ink is crucial. However, inkjet printing of carbon and metal oxides for electrodes remains a challenge and requires further ink and process development.
2.4. Slot-Die Coating
Slot-die coating is an attractive, low-cost printing technique that is used for the deposition of the perovskite and other layers because of its high uniformity in large-scale production and thickness control over a broad range. Nevertheless, slot die coating has not been considered in the literature as a promising coating technique for C-PSCs
[62][138]. A standard slot-die coating process includes an ink reservoir and slot-die head positioned towards the substrate, then the ink is pumped into the head using a syringe pump, with the ink forced out of a narrow slit along the length of the coating head. The ink then creates a liquid bridge between the substrate and slot-die head, and deposition happens upon moving the substrate across the coating head
[62][138]. Slot-die coating covers inks within a wide viscosity window ranging from 1 to 10,000 mPa s, and final dry thickness from a few nanometers to tens of microns
[63][139]. In this technique, the determination of the operational limits to set parameters such as coating speed, flow rate and coating gap must be selected carefully to avoid coating defects
[64][140].
Khambunkoed et al. fabricated carbon-based methylammonium-free PSCs in 2021, utilizing amorphous zinc tin oxide (ZTO) as the ETL
[65][141]. The device was produced with a homemade slot-die system, and the inks used were prepared from commercial materials. Amorphous ZTO was chosen due to its beneficial optical and electronic properties, such as high electrical conductivity, high electron mobility, and high transparency. The device had an optimum PCE of 9.92% when the ZTO thickness was around 48 nm. This PCE value was comparable to that of a PSC with spin-coated ZTO. The short circuit current density, open circuit voltage, and fill factor of the device were 18.86 mA/cm
2, 0.94 V, and 56%, respectively.
2.5. Blade Coating
Doctor blade coating is one of the widely used techniques to produce thin films on large area surfaces. The operation principle ensembles the screen-printing approach. In the doctor blade process, the precursor solution is applied between the coating head and the substrate; then, the substrate is coated by moving the blade or knife across the substrate and the solvent is evaporated through heat treatment to obtain the film
[66][142]. Furthermore, the controllable substrate temperature can impact the quality of the film. The inks/pastes used in these processes usually require large amounts of binders and thickeners to produce the high viscosities (1000–10,000 mPa s) required for reproducible and reliable production of films. Viscosities can be increased with the addition of polymeric additives such as glycerol or ethylene glycol or ethyl cellulose
[67][143]. Coating thickness and speed play a vital role in determining interfaces quality. This technique is a relatively simple and scalable fabrication method for C-PSCs
[68][144].
Syrrokostas et al. reported on the fabrication of C-PSCs with double-layered ZrO
2 films using the doctor blade method
[69][145]. This included the use of commercial zirconia nanoparticles of both nano and micro sizes, which acted as spacer and light scattering material to increase the light harvesting efficiency and yield an enhanced photocurrent density. The UV-Vis absorption spectra of the N_ZrO
2 and DL_ZrO
2 films reveal an increased absorbance, covering all the visible and the near-infrared region, from 400 nm to 800 nm, in the second case. The increased absorbance for the DL_ZrO
2 film is ascribed to the higher scattering ability owing to the size of the particles. The presence of the large ZrO
2 particles in the double layered spacer film results in an increase of the optical path length, due to the improved scattering efficiency; this resulted in a 30% improvement in efficiency.
Tian et al. reported further advances by producing C-PSCs using commercial materials and incorporating fullerene C60 as the ETL, in addition to the modification of the ITO substrate with a self-assembled monolayer of 3-aminopropyl triethoxysilane (APTES)
[70][146]. This combination enabled interfacial charge extraction, ultimately resulting in reduced hysteresis in the J-V curve. They report a remarkable PCE of 18.64%, one of the highest reported for C-PSCs, which was found to be stable for over 3000 h. Bidikoudi et al. explored the use of different ammonium iodide (AI) precursors as additive and post-treatment agent in the multiple-cation mixed halide perovskite precursor solution, which enabled the deposition of homogeneous films of ZrO
2 and carbon with commercial materials
[71][147]. This led to an increase in power conversion efficiency by 33%, with further post-treatment with AI and ethyl-AI raising the efficiencies to 9.77% and 9.15%, respectively. The authors highlighted the importance of selecting the right materials for each part of the solar cell and the substantial influence of the precursor and post-treatment solutions in the C-PSCs fabrication process.
2.6. Spray Coating
Spray coating is an effective technique to produce ETL, HTL and other films, as well as the perovskite layer, on various kinds of substrates
[54][72][130,148]. It is recognized as an attractive method for fabricating large-area, high-throughput, and low-cost PSCs and modules. The spray coating technique can be classified into different methods, including ultrasonic spraying, pneumatic spraying and electrospraying. The classification is based on the droplet dispersion over the substrate
[73][149].
In 2017, Yang et al. reported a simple ultrasound spray deposition method to prepare pure MWCNT films to improve the contact at the perovskite/carbon interface
[74][151]; the c-TiO, m-TiO
2 and PbI
2 layers were deposited on FTO/glass by spin coating. Commercial MWCNTs were dispersed in chlorobenzene to form a homogeneous ink, and the MWCNT layer was deposited onto the perovskite layer using a home-made compressed air gun-based spray system. The PCE of the cell with an active area of 0.08 cm
2 was 14.07%. Moreover, they deposited a sub-monolayer of NiO nanoparticles prior to the MWCNT deposition to enhance the hole extraction efficiency and reduce the recombination rate obtaining up to 15.80% of efficiency, which is among the highest efficiency reported for C-PSCs.
Another interesting report was published by Wu et al. on the use of electrospray (ES) coating to fabricate C-PSCSs
[75][152]. This technique utilizes a high voltage to atomize a flowing solution into charged micro-droplets and can be used to deposit any layer in the perovskite solar cells. ES is distinguished from other techniques due to the advantages of versatility, scalability, designability, and low materials waste
[76][153]. Wu et al. developed an electrospray deposition technique for continuously printing C-based HTL-free PSCs. All the layers, i.e., TiO
2, perovskite and carbon were continuously printed through ES in the ambient atmosphere at low temperature. Commercial precursor solution, paste and solvents were used to prepare the dispersions. The fully ES printed carbon-based PSCs showed V
OC of 1.03 V, J
SC of 24.35 mA cm
−2, FF of 57.7% and PCE of 14.41%. Even though this technology is not widely used for perovskite fabrication, the electrospray-assisted technique is promising for C-PSCs commercial manufacture.
3. Scale-Up
Although the PCE of PSCs is high enough in small areas, it is necessary to fabricate large-scale perovskite solar modules for future commercialization of this PV technology. However, scientific challenges, technical and techno-economical issues are magnified while increasing the substrate size for the fabrication of larger devices. For a successful scale-up, novel abundant and non-toxic materials, new device architectures, easy and low-cost fabrication processes, and adequate choice of printing technique compatible with the ink formulations and substrate size are required. In addition, in a production line of large-scale devices, all manufacturing steps should be compatible and match with a continuous flow process to avoid a slow production rate and, hence, higher manufacturing cost
[77][78][154,155]
One example of a conflicting fabrication technique with large-scale substrates is spin coating. Unfortunately, spin coating technique works well for small devices with less than 1 cm
2. Another problem that is magnified when up scaling the perovskite PV technology is lead toxicity. Still, lead-based perovskite solar cells possess by far the highest efficiencies. To overcome this issue, investigation on lead-free perovskite PVs and effective encapsulation of the entire device should be the important subjects of future research. As mentioned before, the unstable metal top electrode and degradation of perovskite material, which are the main reasons of device instability, have motivated researchers to find stable top electrode alternatives, such as carbon, and new perovskite formulations. Han et al. demonstrated that the carbon-based perovskite solar cell as a fully printable architecture is a promising candidate for large-scale applications related to the easy and cost-effective manufacturing process using stable and abundant materials
[41][86]. Although the carbon-based perovskite devices have shown potential for scale-up, one barrier for large-scale application is the multiple sintering processes required in their manufacture. The sintering process helps to eliminate all organic binders from the printing pastes and results in a better interconnection between nanoparticles. However, it is costly and time-consuming, which reduces the production throughput. One solution is developing low-temperature processing materials such as aqueous-based inks and low-temperature carbon pastes. Unfortunately, the efficiency of low-temperature-based PSCs is still low. This could be due to the adhesion problems of low-temperature and aqueous-based inks to the substrate and poor particle interconnections. There are different deposition methods, including slot-die coating, screen printing, spray coating, roll-to-roll printing that can be used for the fabrication of large area C-PSCs. Rong et al. compared different fabrication methods for large-scale fabrication of PSCs and suggested that among all these methods, screen printing and slot-die coating would be the most promising methods
[79][156].