† Corresponding author. E-mail:
Perovskite solar cells (PVSCs) have attracted extensive studies due to their high power conversion efficiency (PCE) with low-cost in both raw material and processes. However, there remain obstacles that hinder the way to their commercialization. Among many drawbacks in PVSCs, we note the problems brought by the use of noble metal counter electrodes (CEs) such as gold and silver. The costly Au and Ag need high energy-consumption thermal evaporation process which can be made only with expensive evaporation equipment under vacuum. All the factors elevate the threshold of PVSCs’ commercialization. Carbon material, on the other hand, is a readily available electrode candidate for the application as CE in the PVSCs. In this review, endeavors on PVSCs with low-cost carbon materials will be comprehensively discussed based on different device structures and carbon compositions. We believe that the PVSCs with carbon-based CE hold the promise of commercialization of this new technology.
Along with the gradual exhaustion of fossil energies and the environmental pollution caused by employing them, people increasingly realize the importance of renewables. Solar energy is the most abundant energy in the world. In order to convert solar energy into electric energy which can be directly used by modern society, various kinds of solar cells were designed. Solar cell technologies are grouped into three generations by Martin Green,[1] the first generation usually refers to mono-crystalline silicon and poly-crystalline silicon solar cells,[2–4] the second generation is the family of thin film solar cells (such as cadmium sulphide, cadmium telluride, copper indium diselenide, and amorphous silicon), and the third generation includes polymer solar cell (PSC), dye-sensitized solar cell (DSSC), quantum dot solar cell (QDSC), and so on, which utilize new materials and nanotechnology. So far, the family of silicon-based solar cells dominates the world’s photovoltaic market, but they suffer from drawbacks such as high manufacturing cost, high energy-consumption processes, and severe environmental pollution during the production process. The so-called second generation solar cells, namely, thin-film solar cells, also have some shortcomings, for instance, low power conversion efficiency (PCE), scarcity of raw materials, and operational instability. The third-generation solar cells show great prospect because of their interesting features including low-cost, less energy consumption, and abundant raw materials. Therefore, the third-generation solar cells conform to the trend of the commercial market of the photovoltaic industry.
Among the third-generation solar cells, dye-sensitized solar cells (DSSCs)[5] attracted intensive research interest due to the low-cost materials and ease of fabrication. The upsurge of investigations on perovskite solar cells (PVSCs) sprang from the community of DSSCs. Since the pioneering work of Kojima et al. who introduced CH3NH3PbBr3 and CH3NH3PbBr3 as light absorbers in a liquid-state perovskite-sensitized solar cell in 2009, PVSCs have allured intensive studies both from the academic world and industrial community.[6–9] With thousands of publications and increasing attempts to commercialization, PVSCs gradually become an almost matured technology during the past eight years.[10]
PVSCs usually consist of matrix-supported transparent conductive oxide (TCO) anode, charge selective contact layer (hole/electron transport layer or H(E)TL), perovskite absorbed layer, second charge selective contact layer (electron/hole transport layer), and cathode.[11] PVSCs generate electricity according to the following mechanism (Fig.
Using Au or Ag as the CE is indispensable to fabricate state-of-the-art PVSCs, but Au is too expensive for mass industrial manufacture and evaporating Au or Ag onto the top of the cell is a high energy-consumption process which needs high vacuum and high temperature. Domanski et al. found that, when PVSCs are exposed to a temperature of 70 °C, considerable amounts of gold from the back electrode will diffuse across the HTL into the perovskite material, resulting in dramatic loss of the device performance under working conditions.[16] Ag is cheaper and works well in PVSCs with HTM, but it can become corroded when in contact with the perovskite film, which is likely to be due to the formation of silver halide.[17] Hence, replacing the noble metal cathode with low-cost back electrode materials is imperative. Through the unremitting endeavor of many research groups, carbon materials as the cost-efficient choice of back electrodes have been introduced to PVSCs to address issues above. In this review, the application of carbon materials acting as CEs in PVSCs will be exhibited. We will discuss constructions, properties, and morphologies of those materials employed as electrodes, and how these characters affect the performance of the corresponding device. A summary of almost all research works using carbon as counter electrode is shown in Table
![]() | Table 1.
A summary of almost all research works using carbon as counter electrode. . |
Carbon materials have many merits including abundant sources, low-cost, high electrical conductivity, chemical stability, diversity, and modifiability.[77] The appropriate 5.0 eV work function near to that of Au (5.1 eV) makes carbon an ideal candidate for PVSCs’ back electrode. According to the different bonded patterns of carbon atoms and different dimensions of crystalline structures, there are several allotropes with very different chemical and physical properties in the elemental carbon family, such as diamond, graphite, graphene, graphdiyne, C60, carbon nanotube, and amorphous carbon, as shown in Fig.
Carbon paste is mainly prepared by graphite and carbon black and is often formed on the top of hole-conductor-free PVSCs. Graphite is a soft, soapy, hexagonal layered crystal with some particular characteristics such as high-temperature resistance, excellent electrical conductivity and thermal conductivity, lubricity, anti-corrosion, plasticity, and thermal shock resistance. Carbon black is a light, loose, and extremely fine black powder produced by incomplete combustion or thermal decomposition of carbon aceous matter under insufficient air condition. Carbon black is one kind of amorphous carbon. The so-called ‘amorphous carbon’ does not indicate that the shapes of carbon black are confused, but it means the inner turbostratic structure of stacking graphite layers.[78] There are mainly two types of device configurations in carbon CE based PVSCs. One is called the monolithic structure which was previously reported in DSSCs featuring the loading of perovskite in the last step (Fig.
![]() | Fig. 3. (color online) Schematic structure of device configurations in carbon CE based PVSCs: (a) monolithic structure; (b) conventional structure. Reprinted from Ref. [21]. |
In 2013, Hongwei Han’s group established a hole-conductor-free CH3NH3PbI3 perovskite/TiO2 heterojunction solar cell which first introduced carbon black/spheroidal graphite as a counter electrode to substitute the noble metallic electrode, and an impressive PCE of exceeding 6.64% was obtained. This carbon black/graphite back electrode was screen-printed on the top of the TiO2 nanocrystalline layer, then the CH3NH3PbI3 precursor was dipped on top of the mesoscopic carbon layer, and finally a device configuration of FTO glass substrate/TiO2 dense layer/TiO2 mesoscopic layer/ZrO2 space layer/carbon layer was formed, which is named monolithic structure.[25,79] The device was fabricated by a simple process like screen printing and showed a long-term stability in the dark after 840 hours. This work initiated the research to low-cost carbon back electrode based PVSCs.[18] Later, they did a series of work on the carbon-based electrode PVSCs.
In 2014, they introduced a mixed-cation perovskite (5-AVA)x(MA)1−xPbI3 to replace MAPbI3 in almost the same triple-layer structure of TiO2/ZrO2/carbon, as shown in Fig.
![]() | Fig. 4. (color online) (a) The crystal structure of CH3NH3PbI3 perovskite and the corresponding energy levels of TiO2, CH3NH3PbI3, and carbon. (b) A schematic structure of the carbon-based monolithic device. Reprinted from Ref. [18]. |
It is known that the device performance depends heavily on the quality of the perovskite layer such as morphology, crystallinity, and density of defects. In the case of mesoporous device structures, higher pore filling of perovskites phase is crucial to achieving high-performance monolithic PVSCs, which generally have thick mesoporous films over
Except for the quality of the perovskite itself, they investigated the size effect of bottom TiO2 nanoparticles on their printed PVSCs. The size of TiO2 particles affects the infiltration of the precursor and the contact between the perovskite crystal and TiO2, which influences the charge transfer at the perovskite/TiO2 interface. The optimized diameter of 25 nm TiO2 nanoparticles exhibited the best PCE of 13.41%.[51]
The ZrO2 film is also very important, who separates the TiO2 film and carbon film, avoiding any direct contact which influences light harvesting and the transfer of photo-generated charge. They found that the appropriate thickness of the mesoporous ZrO2 film was around
Also, as a pivotal component, graphite decides the porosity and conductivity of back carbon electrode. Different sizes of graphite will cause differences in the filling of PbI2 and CH3NH3PbI3 precursors and also the square resistance of carbon counter electrode, which will finally affect the performance of the devices. So, they investigated the previous topic, and found that the 8 mm graphite-based carbon back electrode has a more significant average pore size, a smaller square resistance, and hence a higher PCE exceeding 11% compared with others. The square resistance of
![]() | Table 2.
The square resistance of |
Besides Han’s work, to realize more mechanization of the fabrication devices process, Hashmi et al.[45] replaced manual infiltration of perovskite precursor step used in Han’s works with inkjet printing in the same device structure of PVSCs (Figs.
![]() | Fig. 5. (color online) (a) Illustration of the scalable process to be employed for the production of HTM free carbon based PVSCs: a) fluorine-doped tin oxide (FTO) coated glass substrate, b) laser patterning of the FTO layer, c) spray pyrolysis for compact TiO2 layer deposition, d) screen printing of silver contacts, e) screen printing of a mesoporous TiO2 layer, f) screen printing of an insulating ZrO2 layer, g) screen printing of a porous carbon composite layer, h) inkjet of a controlled volume of perovskite precursor solution, which is targeted for this work. Note that only scalable processes are highlighted here. The sintering steps consecutive to screen printing are not represented for better clarity. (b) Inkjet printing of bitmap version of a digital image as an example of accurate 2D printing, image reproduced with permission from Alain Herzog/EPFL. Adapted from Ref. [18]. |
To reduce the interfacial charge recombination, Mingkui Wang’s group developed a quadruple layer of TiO2/Al2O3/NiO/carbon as scaffold infiltrated with MAPbI3, as shown in Fig.
![]() | Fig. 6. (color online) Optimized device structure. (a) Schematic representation of the TiO2/Al2O3/NiO/carbon (MAPbI3) based device. (b) Energy band diagram of the fabricated device configuration. Reprinted from Ref. [25]. |
The perovskites adopted in works mentioned above are all organic–inorganic hybrid perovskites. Recently, inorganic perovskites (e.g., CsPbI3, CsPbBr3) attracted extensive attention due to higher stability.[81–83] A printed HTM-free PVSC with all the components made of inorganic materials (porous metal oxide, inorganic perovskite, and carbon) with monolithic configuration showed increased thermal stability. Li et al.[84] fabricated lead-free inorganic perovskites CsSnX3-based PVSCs with the structure of FTO/TiO2/Al2O3/CsSnX3/C, the as-prepared all-inorganic devices showed superior thermal stability up to 200 °C, with an average PCE of 3.0% and long-term stability for 77 days without efficiency-loss.
In all the works above, the monolithic PVSCs were usually fabricated by the inexpensive and straightforward screen-printing method, of which the structure avoids costly and unstable organic HTMs and noble metal electrode like Au. These HTM-free carbon-based PVSCs boost the promise of PVSC to commercialization. The monolithic PVSCs have the advantages of ambient condition process, efficacious use of precursor, thick perovskite layer for higher stability, and better contact between perovskite and CE. Meanwhile, the thick metal oxide layers bring disadvantages like the high-temperature process and large series resistance, etc.
Regarding the carbon paste mentioned above, high-temperature sintering above 400 °C is necessary, which will hamper the large-scale commercialization of carbon-based PVSCs. So, it needs to develop the low-temperature processed CE. Apart from the lower energy consumption, low-temperature (
At almost the same time, Yang’s group reported clamping solar cells using candle soot for hole-extraction. In the fabrication process, the perovskite film is fabricated by depositing a PbI2 precursor layer, immediately followed by the rolling transfer of candle soot, and finally subjecting the sample to a CH3NH3I bath treatment, as shown in Figs.
![]() | Fig. 7. (color online) (a)–(c) Development of a candle soot/perovskite clamping solar cell. Reprinted from Ref. [62]. (d) Fabrication process flow for the instant inkjet printing of the C/CH3NH3PbI3 planar perovskite solar cells. For comparison, a different strategy was used to convert PbI2 into CH3NH3PbI3 using separate step 3 and step 4. Reprinted from Ref. [19]. |
Then, Yang’s group moved efforts towards the fabrication of high-quality perovskites by painting carbon pastes on the top of pre-prepared perovskites. They developed an electrodeposition technique to synthesize the perovskite layer. The electrodeposited perovskite layer was printed with the carbon paint as back contact. The corresponding devices achieved a PCE of 10.19%, and showed an encouraging PCE in a larger cell area up to 4 cm2.[57] Although the efficiency of the painted carbon-based-PVSCs has come up to over 10%, it is still low compared with the traditional PVSCs with Au or Ag CEs. The primary limiting factor of the efficiency of the painted-carbon-based PVSC lies in the poor contact at the perovskite/carbon interface due to the lamella feature of graphite layer and ease of producing interspace between the graphite flakes and perovskites.[20]
To improve the contact at the perovskite/carbon interface, Yang et al. designed a solvent engineering strategy based on two-step sequential method which is beneficial to an excellent pore-filling perovskite layer in a relatively thick mesoscopic TiO2 scaffold. The solvent for CH3NH3I solution at the second step is a mixed solvent of IPA/cyclohexane. This mixed solvent results in a high-quality perovskite layer with smooth surface and compact capping layer which could enhance the contact with CE for a majority of graphite flake in carbon paste. Finally, they obtained the highest PCE of 14.38% with an FF of 0.65 and a PCE of 10% for the 1 cm2 area device.[41] Moreover, through colloidal engineering to form an ultra-even perovskite layer, they boosted the efficiency of the carbon-based PVSCs to 14.58%.[43]
In the presence of high-quality perovskite layer, Sun et al. introduced CuPc nanorods as hole-transporting materials to improve the interface contact. The device displayed an impressive PCE of 16.1%, which is comparable to that of the corresponding devices with the doped spiro-OMeTAD as HTMs and noble metal Au as back electrode. CuPc nanorods induce an intimate, large interfacial-area contact with the MAPbI3 crystal grains and carbon counter electrode, therefore leading to a significant enhancement in hole-extraction, and a decrease of charge recombination (Fig.
![]() | Fig. 8. (color online) Device structure and energy-level diagram: (a) schematics of the whole device: FTO glass/compact TiO2 layer/mesoporous TiO2/CH3NH3PbI3 capping layer/CuPc/carbon and (b) the corresponding energy level diagram. Fabrication process flow for the instant inkjet printing of the C/CH3NH3PbI3 planar perovskite solar cells. For comparison, a different strategy was used to convert PbI2 into CH3NH3PbI3 using separate step 3 and step 4. Reprinted from Ref. [50]. |
To enhance the contact of CE and perovskite layer for the favorable charge transport, Qingbo Meng’s group developed two-step processed carbon CEs with different graphite flake sizes and contents of carbon black. Firstly, viscous carbon pastes were first coated on the top of the as-prepared devices. Then, a piece of graphite paper was pressed on the carbon paste as an extraction electrode, as shown in Fig.
![]() | Fig. 9. (color online) Scheme of the fabrication process of all-carbon flexible CE based HTM-free perovskite solar cells. Reprinted from Ref. [20]. |
Meng et al. also fabricated a free-standing flexible carbon electrode containing 20 wt.% polyvinyl acetates (PVAc) in the carbon film to avoid the adverse effect of solvents in carbon pastes. Carbon paste was coated on Teflon film by doctor-blading to form the carbon film after the solvent evaporation. Due to the thermoplastic PVAc, the carbon films can be directly hot-pressed onto the surface of perovskite at 85 °C. With the optimized carbon film components (graphite flakes and carbon black particles with a weight ratio of 3:1 and hot-press pressure 0.25 MPa), champion PCE of 13.53% with an average PCE of 12.03% has been achieved. Compared to PVSCs with Au electrode, this low-cost and stable carbon-based one with low-temperature-processed and flexible carbon counter electrode is promising.[27]
A low-cost carbon cloth embedded in carbon paste as CE was studied by Hagfeldt et al., as shown in Fig.
![]() | Fig. 10. (color online) (a) a) A schematic image of the carbon cloth/spiro-OMeTAD perovskite solar cell fabrication. b) Energy band alignment of a carbon cloth-based device. Reprinted from Ref. [42]. (b) Schematic representation of the packaged device. Reprinted from Ref. [33]. |
Additionally, Liao et al. reported low-temperature carbon CE based HTM-free mesoscopic PVSCs and encapsulated them by PDMS, as shown in Fig.
Carbon nanotubes, also called buckytube, have a particular 1D hollow tubular structure, which were uncovered by Iijima in 1991.[86] The tubular structure of CNTs can be considered as the result of rolling up monolayer graphite sheet, so CNTs inherit the admirable properties of graphite such as anti-corrosion, impact resistance, thermal and electrical conductivity.[87] Over the past twenty years, CNTs have attracted researchers majoring in the field of electronic and optoelectronic applications due to their exceptional charge transport property, inherent hydrophobicity, chemical and mechanical stability. CNTs have been incorporated into organic solar cells as electron acceptor successfully. Moreover, the p-type CNTs junctures with n-type silicon were demonstrated to achieve a decent PCE up to 17.0%.[88]
As we know, charge selective contacts on both sides of the perovskite absorbers are crucial elements controlling the whole performance of PVSCs. The inherent p-type nature of CNTs and the decent employment as HTM in organic solar cells make CNTs an alternative HTM for PVSCs. One advantage of using CNTs as HTM is eliminating dopants like lithium salts which are blaming to endanger the long-term stability of PVSCs.[87] Also, the mechanical resilience of CNTs may prove to be particularly beneficial for the development of flexible perovskite solar cells. In this section, some significant efforts introducing CNTs into PVSCs acting as hole conductor and extraction counter electrode will be presented.
Jang et al. developed a dripping method to drip multi-walled carbon nanotubes (MWCNTs)/chlorobenzene solution on perovskite precursor solution at the time of its spinning, as shown in Fig.
![]() | Fig. 11. (color online) Schematic illustrations of (a) the perovskite layer deposition procedure and (b) the prepared carbon electrode-based perovskite solar cells. (c) A cross-sectional SEM image of the prepared device. (d) An energy diagram of the materials used in this work. Reprinted from Ref. [89]. |
![]() | Table 3.
Summary of resistances of devices with a perovskite layer fabricated under various conditions. Adapted with permission from Ref. [89]. . |
Different from Jang’s work that dripping CNTs into the perovskite layer, Mingkui Wang et al. designed single-walled carbon nanotubes (SWCNTs)/carbon composite counters electrodes by printing carbon paste on the top of the Al2O3 layer and formed a triple layer structure of TiO2/Al2O3/C, as shown in Figs.
![]() | Fig. 12. (color online) (a) Schematic architecture of the investigated device consisting of FTO as a transparent substrate, a compact TiO2 layer, a mesoporous TiO2 layer, an Al2O3 layer coated with SWCNT added graphite/carbon black CE, (b) cross-section SEM image of a mesoscopic perovskite solar cell. Reprinted from Ref. [32]. (c) Photo of freestanding CNT film lifted by tweezers to transfer onto other substrates. (d) Schematic of CH3NH3PbI3 perovskite solar cell with CNT film electrode. (e) Top view SEM images of CH3NH3PbI3 perovskite substrate before and (f) after CNT transfer. (g) Tilted SEM image of CH3NH3PbI3 perovskite substrate (blue) partly covered by CNT film (purple). Reprinted from Ref. [66]. |
One of the initial works that employ CNTs as hole conductor in mesoscopic PVSCs was made by Mhaisalkar et al. in 2014.[66] In this work, PVSCs were fabricated by laminating freestanding films of CNTs network atop the perovskite absorbers, as shown in Figs.
Inspired by the work of Mhaisalkar et al., Wong’s group used Ti foil as substrate and TiO2 nanotube arrays as ETM to fabricate flexible PVSCs. In Ti foil-based PVSCs, the non-transparent Ti foil impedes the light incidence from the anode, so the device will not operate if using a metallic counter electrode. Therefore, a CNTs network the same as that of Mhaisalkar et al. was introduced as the counter electrode. With
Among the family of PVSCs, the clan of carbon-based HTM-free is the most promising for commercialization because of its low-cost and superior long-term stability. However, the FF is lower than that of conventional metal-based PVSCs because of the poor contact at the interface of perovskite/carbon. To address the problem of low FF which may hinder the way of commercialization, Shihe Yang’s group used MWCNT as a hole-selective counter electrode to replace conventional small organic molecule HTM and metallic electrode. With the drop-casting method, graphite, carbon black, and MWCNT electrodes were fabricated on the same as-prepared substrates. The performance of three as-fabricated carbon-based PVSCs was demonstrated by the different parameters in Table
![]() | Fig. 13. (a), (c), (e) Top view and (b), (d), (f) cross-sectional SEM images of the carbon black, MWCNT, and graphite-based PVSCs. The dashed lines represent the interface of FTO/TiO2 and CH3NH3PbI3/C, respectively. Reprinted from Ref. [72]. |
![]() | Table 4.
Solar cell parameters of the three as-fabricated carbon-based PVSCs. Reprinted from Ref. [72]. . |
After this, Yang et al. boosted their work by doping boron (B) into the MNCNTs, which increases the work function, carrier concentration, and conductivity of the MWCNTs and enhances their ability of hole-extraction and transport. Also, an insulating aluminum oxide (Al2O3) thin layer was coated on the mesoporous TiO2 film as a physical barrier to reduce the charge losses sharply. Finally, the efficiency was promoted to 15.23%.[71] In another case, Hong Lin et al. integrated a highly conductive and porous cross-stacked super-aligned carbon nanotube (CSCNT) sheet as a metal-free back contact at low temperature in HTM-free mesoscopic PVSCs, as shown in Fig.
![]() | Fig. 14. (color online) (a) Device architecture of the PVSCs tested in this study; (b) the corresponding energy levels of TiO2, CH3NH3PbI3, and CSCNTs; (c) and (d) the cross-sectional SEM images of the typical CSCNT-based PVSCs with different magnifications. Reprinted from Ref. [69]. |
The overall poor long-term thermal stability is a serious problem that hinders the commercialization of PVSCs. Under elevated temperature condition, considerable amounts of Au from the electrode diffuse across the HTM into the perovskite layer and cause a dramatic loss of the device performance.[16] Aitola et al. designed a low-cost CNT-based high temperature-stable PVSC with a configuration of FTO/compact TiO2/mesoporous TiO2/perovskite/SWCNT-spiro-OMeTAD, as shown in Fig.
![]() | Fig. 15. (a) Schematics and (b) cross-sectional SEM image of the SWCNT-contacted device. The carbon nanotube “whiskers” in front of the SEM image are due to sample preparation. (c) Schematics and (d) cross-sectional SEM image of the standard Au contacted device with spin-coated spiro-OMeTAD HTM. Reprinted from Ref. [90]. |
In 2004, Novoselov et al. successfully prepared a monolayer graphene for the first time by mechanical exfoliation.[91] Since then, graphene has attracted numerous researchers in the field of chemical and material science due to its topping mobility, absorbance, conductivity, mechanical flexibility, transparency, and specific surface area. Now graphene has played more and more important roles in solar cells, energy storage, optoelectronics, electrics, and sensing.[92–95] The first work of graphene-based PVSCs was reported by Snaith et al. who utilized nanocomposites of graphene and TiO2 nanoparticles as the ETM in PVSCs.[96] One significant application for graphene in PVSCs is as bottom electrodes which may replace conventional ITO or FTO transparent electrodes due to its high conductivity, transparency, charge mobility, chemical and mechanical robustness.[97–99] Also, several studies that introduce graphene and its derivatives as PVSCs’ HTM have been carried out.[65,100–106] Researchers found that perovskite grown on graphene oxide presents a higher orientation order than that on PEDOT: PSS, as affirmed by XRD measurement.[100] Another group found that the use of reduced graphene oxide HTMs with the inherent passivation-ability markedly prolonged the device lifetime compared to that of PEDOT: PSS based device.[101] Interestingly, Nouri et al. designed graphene oxide as HTM and Li-modified graphene oxide as ETM in PVSCs.[65]
In a word, graphene, a magical material, has given rise to many studies in PVSCs. In this paper, we focus on its role as a counter electrode. PVSCs have been exploited for their application in making transparent or semitransparent devices that can absorb light from both sides for the past few years. Those devices can be applied to the building-integrated photovoltaic system, wearable electronics, and tandem cells. The selections of transparent top electrodes are crucial to the performance of the cells. Different electrodes, such as Au,[107–109] Ni,[110] and CNTs,[66] have been studied for their performance in these semitransparent PVSCs. A desired transparent electrode should possess outstanding properties, for instance, high transparency, high conductivity, chemical stability, low cost, and charge collecting ability. Due to its excellent optical and electrical properties, graphene stands out to be an ideal candidate as a transparent electrode in PVSCs. Feng Yan et al. reported a semitransparent PVSC by laminating stacked multilayer graphene prepared by the chemical vapor deposition (CVD) method as top transparent electrodes, as showed in Fig.
![]() | Fig. 16. (color online) (a) Schematic diagram of a semitransparent perovskite solar cell. Reprinted from Ref. [111]. (b) Simplified sketch of a four-terminal tandem solar cell consisting of a graphene-based perovskite top solar cell and an amorphous/crystalline silicon bottom solar cell. Reprinted from Ref. [112]. Solar cell performance using SG and MG for hole extraction. Cross-section view of (c) SG and (d) MG based PVSC (scale bar: ![]() |
Yang et al. employed flexible reduced graphene oxide to synthesize single-layered graphene (SG) and multilayered graphene (MG) as hole-extraction electrodes, as shown in Figs.
Hu et al. reported a three-dimensional honeycomb-like structured graphene (3DHG) synthesized by the reaction of potassium with CO2.[75] HTM-free PVSCs introduced this 3DHG as counter electrode achieved a PCE of 10.06%. To complete the HTM-free device, 3DHG was coated atop a CH3NH3PbI3 layer by doctor blading. Aiming to improve the interface, the 3DHG was dispersed into the solution of the perovskite precursors, and the perovskite layer was in situ synthesized. Consequently, a transition layer, in which part of CH3NH3PbI3 perovskite molecules was highly scattered by ultrasonic on 3DHG, was formed between the perovskite layer and the 3DHG counter electrode. Common hysteresis effect was not observed for this device, as shown in Table
![]() | Table 5.
Photovoltaic performance of PVSCs with 3DHG-48 as the CE. Reprinted from Ref. [75]. . |
PVSCs with transparent contacts also used to compensate for thermalization losses of silicon solar cells in tandem devices which surpass the Schockley–Queisser limit for single-junction solar cells. PVSCs with large-area chemical vapor deposition (CVD)-graphene were fabricated by Nickel et al. for tandem solar cells, as shown in Fig.
Most of the devices above have a small active area of around 0.1 cm2. However, as a market-proof photovoltaic technology solution, one needs to demonstrate solar modules with larger area. To make solar panels with conventional structure employing layer-by-layer process, it is challenging to obtain large-area, flat, uniform, and fully covered perovskite films. To solve the problem, we deposited perovskites by slot-die coating technique assisted with our previously reported gas-pumping method (Figs.
![]() | Fig. 17. (color online) (a) Schematic diagram of slot-die coating and (b) pumping of perovskite liquid film. Reprinted from Ref. [115]. |
In 2016, Mhaisalkar et al. reported monolithic perovskite modules with active areas of 70 cm2. The module was fabricated by screen printing along with infiltration of the perovskite (5-AVA)x(MA)1−xPbI3 solution, showing a PCE of 10.74% and good ambient stability of over 2000 h with less than 5% reduction in efficiency. They pointed out that the quality of mesoscopic carbon has a significant impact on the module performance, where high Brunauer–Emmett–Teller (BET) surface area and low electrical resistance are propitious for achieving an efficient module.[116]
Later, employing the same structure, Han et al. optimized the thickness of the mesoporous TiO2 and ZrO2 layer and fabricated a large-area module (active area of 49 cm2 out of total 100 cm2) with a PCE of 10.4%. The module was stable under light soaking test with AM1.5 illumination for 1000 h and tested for a month under outdoor conditions. The shelf life under ambient conditions in the dark was over 1 year. Moreover, to assess the reproducibility of their screen-printing technique in manufacturing high-performance perovskite solar modules, a 7 m2 fully printed perovskite solar panel was also demonstrated (Figs.
![]() | Fig. 18. (color online) (a) Schematic illustration of the proposed production line of perovskite solar modules. (b) Image of 7 m2 printable perovskite solar panels. Reprinted from Ref. [117]. 2D/3D carbon based solar cell characteristics and stability. (c) J–V curve using the 2D/3D perovskite with 3%AVAI in an HTM-free 10 cm×10 cm module (device statistics and picture in the inset). (d) Typical module stability test under 1 sun AM 1.5 G conditions at stabilized temperature of ![]() |
Recently, Nazeeruddin et al. successfully fabricated 10 cm×10 cm (active area of 47.6 cm2) monolithic structure modules with a fully printable industrial-scale process, showing the efficiency of 11.2%. More importantly, the sealed module kept stable for one-year (
CE is a crucial component of PVSCs, in this review, we introduce some critical works in the field of PVSCs that focus on designing or modifying cost-effective carbon-based CEs. After summing up this work, we find that there are mainly three key factors, included the electrical conductivity, the closeness of the perovskite (or HTM)/CE interface and matching work function that determine the performance of CEs and therefore the overall performance of the devices. Carbon materials attracted quite a few researchers because of their earth-abundant, low-cost, and decent conductive properties. Carbon materials can be deposited on the top of the as-prepared substrate by different methods such as screen printing, doctor-blading, inkjet-printing, rolling transfer, hot pressing, spin coating, press transferring, and drop-casting. Although this kind of CEs often have a poor contact with perovskite or HTM, different approaches such as doping, morphology tuning, and device engineering have been used to alleviate this problem and improve the performance of carbon-PVSCs. Among the carbon materials, CNTs and graphene stand out to be potential candidate CEs for semi-transparent or transparent PVSCs, which can be employed in building-integrated photovoltaics and tandem solar cells. Moreover, device lifetime up to 3000 hours has been reported based on carbon cell. A comparison of the device performance between different types of carbon-based CEs is visualized in Fig.
It is evident that there is still a huge gap in PCEs between the carbon-based devices and noble metal-based devices. There is also an ample space for improving the device performance if we can further increase the conductivity of the carbon electrode and improve the contact between the electrode and charge-selective contacts. On the other hand, traditional PVSCs employing noble metals as CEs usually need organic HTMs such as spiro-OMeTAD, PTAA, and PEDOT: PSS, which deteriorate device stability under high humidity or high-temperature conditions. However, the carbon-based devices such as monolithic devices or devices using CNTs as HTM can avoid the use of these fragile organic compounds, and therefore the moist and thermal stability can be enhanced. More importantly, a relatively large-area module with a lifetime up to 10000 hours has been reported based on HTM-free carbon cells. Thus, given the low cost and long-term stability of the carbon materials, they can be an essential solution for the commercialization of perovskite solar cells.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] | |
[35] | |
[36] | |
[37] | |
[38] | |
[39] | |
[40] | |
[41] | |
[42] | |
[43] | |
[44] | |
[45] | |
[46] | |
[47] | |
[48] | |
[49] | |
[50] | |
[51] | |
[52] | |
[53] | |
[54] | |
[55] | |
[56] | |
[57] | |
[58] | |
[59] | |
[60] | |
[61] | |
[62] | |
[63] | |
[64] | |
[65] | |
[66] | |
[67] | |
[68] | |
[69] | |
[70] | |
[71] | |
[72] | |
[73] | |
[74] | |
[75] | |
[76] | |
[77] | |
[78] | |
[79] | |
[80] | |
[81] | |
[82] | |
[83] | |
[84] | |
[85] | |
[86] | |
[87] | |
[88] | |
[89] | |
[90] | |
[91] | |
[92] | |
[93] | |
[94] | |
[95] | |
[96] | |
[97] | |
[98] | |
[99] | |
[100] | |
[101] | |
[102] | |
[103] | |
[104] | |
[105] | |
[106] | |
[107] | |
[108] | |
[109] | |
[110] | |
[111] | |
[112] | |
[113] | |
[114] | |
[115] | |
[116] | |
[117] | |
[118] |