Recent advances on monolithic perovskite‐organic tandem solar cells

Perovskite‐organic tandem solar cells (TSCs) have emerged as a groundbreaking technology in the realm of photovoltaics, showcasing remarkable enhancements in efficiency and significant potential for practical applications. Perovskite‐organic TSCs also exhibit facile fabrication surpassing that of all‐perovskite or all‐organic TSCs, attributing to the advantageous utilization of orthogonal solvents enabling sequential solution process for each subcell. The perovskite‐organic TSCs capitalize on the complementary light absorption characteristics of perovskite and organic materials. There is a promising prospect of achieving further enhanced power conversion efficiencies by covering a broad range of the solar spectrum with optimized perovskite absorber, organic semiconductors as well as the interconnecting layer's optical and electrical properties. This review comprehensively analyzes the recent advancements in perovskite‐organic TSCs, highlighting the synergistic effects of combining perovskite with a low open‐circuit voltage deficit, organic materials with broader light absorption, and interconnecting layers with reduced optical and electrical loss. Meanwhile, the underlying device architecture design, regulation strategies, and key challenges facing the high performance of the perovskite‐organic TSCs are also discussed.


| INTRODUCTION
The escalating demand for clean and sustainable energy sources has intensified research efforts in the field of solar cell technology.Among the various emerging solar cell technologies, perovskite solar cells (PSCs) boast a remarkable power conversion efficiency (PCE) of up to 26.1%.Organic solar cells (OSCs) have also achieved an impressive PCE approaching 20%. [1]As for the PSCs (Figure 1A), the record PCE exceeds 75% of the Shockley-Queisser limit. [2,3]Nonetheless, the PCE achieved by single-junction solar cells will not surpass the theoretical limit of 33.7% as established by the Shockley-Queisser model, limiting the further enhancement of PSCs. [4,5]hile the research community continues its relentless efforts to enhance the performance of single-junction solar cells, there is also growing interest in utilizing tandem structures to further improve the PCE of perovskite-based solar cells.One of the primary challenges in fabricating perovskite-based tandem devices is to develop a process for subsequent multilayer deposition that does not damage the underlying functional layers.[8] These materials are known for their excellent resistance and structural stability, making it feasible to deposit an intermediate interconnecting layer (ICL) and front perovskite layer for tandem device fabrication.However, the fabrication of Si or CIGS devices often requires the expenditure of substantial energy, thereby prompting researchers to explore more streamlined processes for device fabrication.All-organic or all-perovskite tandem solar cells (TSCs) can be achieved through a simple solution-processing method.However, for all-organic tandem devices, organic materials exhibit limited absorption in the short-wavelength range, which hampers their ability to achieve fullspectrum absorption of solar energy. [9]Additionally, organic materials are susceptible to oxidation and photodegradation when exposed to ultraviolet (UV) light. [10]onsidering the existing photovoltaic techniques, allperovskite or perovskite-organic TSCs have the potential to be fabricated with lower energy consumption and solution process.
However, regarding a full solution process for allperovskite and perovskite-organic TSCs, polar solvents with high boiling points are used to process both the rear and front cells.A robust solvent resistance layer or ICL is typically required to shield the underlying perovskite layer from solvent damage during the deposition of the rear photoactive layer.Early works mainly used transparent metal oxide (ITO, IZO, MoO x , etc.) as the solvent resistance layer prepared by the sputtering deposition demanding a high energy and high vacuum condition. [11]herefore, the ICL plays a significant role in realizing high-performance tandems reproducibly. [12,13]While OSCs with organic narrow-bandgap (NBG) semiconductors have the advantage of orthogonal solvents complementary to perovskite layer, simplifying the fabrication process without requiring complex structures. [14]Additionally, the solution coating processing of organic semiconductor materials offers opportunities for implementing various scalable deposition techniques, including inkjet printing and blade coating that are compatible with roll-to-roll coating.This capability is essential for achieving high-throughput and cost-effective production of devices.Fabricating a tandem device composed of a front layer of wide-bandgap (WBG) perovskite and a rear layer of NBG organic is a potential approach to address the compatibility issue during solvent processing.
We have summarized and recorded the development of PSCs and OSCs, as well as all-perovskite TSCs and perovskite-organic TSCs, as shown in Figure 1B.It is the photon energy below-bandgap non-absorption and above-bandgap thermalization losses that primarily constrain the performance of single-junction solar F I G U R E 1 (A) State-of-the-art PCE of mainstream and emerging single-junction photovoltaic technologies compared to the detailedbalance limit.Reproduced with permission. [2]Copyright 2023, Elsevier.(B) The PCE evolution of single-junction PSCs, single-junction OSCs, all-perovskite TSCs, and perovskite-organic TSCs.PCE, power conversion efficiency; PSC, perovskite solar cells; TSC, tandem solar cell.
cells. [15]To be more specific, on one hand, the losses are attributed to the unavailable absorption of photon with energy below the bandgap, resulting in inefficient photon utilization.On the other hand, thermalization losses primarily occur when the energy of absorbed photons exceeds the bandgap of materials, causing the relaxation of generated electrons and holes toward the band edge and subsequent heat dissipation (Figure 2A). [15]To mitigate these losses and further enhance performance, an effective strategy is to construct tandem devices.Regarding the all-perovskite and perovskite-organic TSCs, the performance gap between them becomes minor as the development of both PSCs and OSCs, and showing great potential surpassing the subcells that are constructing the tandems (Figure 1B).
The tandem devices typically consist of two or more layers of light-absorbing materials with different bandgaps to optimize spectral absorption.For instance, two-terminal (2-T) TSCs mainly comprise a front subcell, ICL, and a rear subcell (Figure 2B), with the two subcells connecting in serial using ICL.Spectral absorption complementation and current matching between the two subcells are crucial for achieving high-efficiency tandem devices.On the one hand, the lower current of the subcell limits the output photocurrent, as both subcells must have identical currents flowing through them.On the other hand, a broader absorption for front subcells with rear subcells extended to near-infrared could further enhance the output photocurrent.As presented in Figure 2C, the obtainable maximum theoretical PCE of the two subcells in 2-T TSCs can exceed 40% through adjusting the bandgap of perovskite and organic materials.A better balance of photon absorption between the two subcells can be achieved by carefully selecting light-absorbing materials. [16]The tandem structure allows for a more efficient solar spectrum utilization, as high-energy photons are absorbed by the front subcell, while transmitted lowerenergy photons can be absorbed by the rear subcell (Figure 2D).This complementary absorption of photons significantly increases low-energy photon absorption and F I G U R E 2 (A) Schematic illustration of light absorption in dual-junction TSCs with 1.8 eV bandgap front-cell and 1.1 eV bandgap rear-cell.Reproduced with permission. [15]Copyright 2017, Wiley.(B) Diagram of the architecture of 2-T TSCs.(C) Theoretical maximum PCE as a function of the bandgap for the rear and front subcells in a 2-T architecture.Assuming negligible losses from absorption and contact, the dashed line represents the peak efficiency across a wide range of front-cell bandgaps, while the white circles indicate the maximum PCE.Reproduced with permission. [16]Copyright 2017, Springer Nature.(D) Solar irradiance spectrum depicting the spectral absorbing regions for different bandgap subcells in 2-T TSCs.Reproduced with permission. [17]Copyright 2021, Wiley.2-T, two-terminal; TSC, tandem solar cell.
reduces high-energy photon thermalization losses, ultimately leading to enhanced performance of the tandem devices.
At the state-of-the-art, perovskite materials, with tunable bandgaps and high absorption coefficients, offer a versatile platform for front subcell design, however showcasing a large energy loss considering the bandgap.Organic materials, on the other hand, provide flexibility and cost-effectiveness for the rear subcell.However, they exhibit limited light absorption compared with Si rear cell.Furthermore, achieving a suitable energy level alignment and charge transport properties at the perovskite-organic interface remains a critical focus.Moreover, the gradual maturation of ICLs has led to a harmonized balance between optical and electrical losses in perovskite-organic TSCs.To investigate the efficiency evolution of perovskite-organic TSCs, the parameters of n-i-p and p-i-n type perovskite-organic TSCs and their device structure with the WBG perovskite and NBG organic as well as the ICL are summarized in Tables 1  and 2. It is worth noting that the highest efficiency perovskite-organic TSC is realized through the combination of p-i-n type perovskite and organic subcells.This can be ascribed to that most of the high-performance WBG PSCs reported in the literature are with p-i-n type device structure, and existing n-i-p NBG OSCs still suffer more severe efficiency losses than the p-i-n OSC counterparts. [33,34]Cesium (Cs)-based all-inorganic perovskites with a WBG are commonly used as the front subcells for n-i-p type perovskite-organic TSC, while organicinorganic mixed halide perovskites are utilized for p-i-n type configurations.However, all of them encounter significant issues related to voltage loss attributed to phase segregation.The NBG organic materials employed in the rear subcells of perovskite-organic TSCs comprise donor and acceptor components.Taking advantage of the development of NBG nonfullerene acceptors, including Y6, ITIC-4Cl, BTPV-4Cl-eC9, and so on, these acceptors can be used in perovskite-organic TSCs to extend the absorption range into the near-infrared region, which significantly increases the current of the tandem devices.Recent achievements in perovskite-organic TSCs have showcased an impressive PCE of 24% that surpasses the individual single-junction OSCs and approaching the single-junction PSCs (Figure 1B).Synergistic light absorption and efficient charge extraction have collectively contributed to these advancements.This indicates that utilizing the tandem configuration for achieving further enhanced performance should focus on the perovskite front subcell, organic rear subcell, and also the ICL between them.This review concentrates on the latest progress in monolithic perovskite-organic TSCs, primarily focusing on the WBG perovskite subcells, NBG organic subcells, and ICLs.Subsequently, we delve into the challenges surrounding attaining highly efficient and stable tandem devices surpassing single-junction subcells, followed by exploring the promising prospects within the domain of perovskite-organic TSCs technology.

| LOW-V O C LOSS WBG PSCs
In recent years, WBG PSCs have drawn much attention since a high open-circuit voltage (V OC ) of WBG perovskite plays a crucial role in high-performance tandem devices.The organic-inorganic perovskites have a typical formula of ABX 3 (Figure 3A).The A site is normally F I G U R E 3 (A) Typical crystal structure of metal halide perovskites.(B) Repartition of voltage losses and V OC of PSCs: interface recombination losses (caused by the defects at the interfaces of perovskite/charge transport layer), bulk recombination losses (caused by the bulk defects), solar cell V OC .Reproduced with permission. [35]Copyright 2021, Wiley.(C) The V OC losses as a function of perovskite bandgap were recorded from the reported inverted WBG PSCs.Reproduced with permission. [36]Copyright 2023, Wiley.Photoluminescent (PL) spectra of perovskite films (D) without and (E) with synergetic treatment under 0.5 mJ laser illumination for 30 min, respectively.Reproduced with permission. [37]Copyright 2022, Wiley.Electroluminescent (EL) emission spectra of (F) BCP-based and (G) SnO x -based WBG devices with various applied voltages, respectively.Reproduced with permission. [36]Copyright 2023, Wiley.
occupied by a monovalent organic cation such as methylammonium (MA + ), formamidinium (FA + ), and cesium (Cs + ).B site is a divalent metal cation such as lead (Pb 2+ ) and tin (Sn 2+ ).X site is halide anions such as chlorine (Cl − ), bromide (Br − ), or iodide (I − ). [38,39]The bandgap of perovskite could be tuned by composition regulation.The modulation of the X-site is commonly employed to adjust the bandgap of perovskite materials effectively.For example, the bandgap of MAPbI 3 will increase to 1.80-2.30eV from 1.55 eV when the Br and Cl ions replace the I ion of the X-site. [40]Generally, the organic-inorganic mixed-halide perovskites with a bandgap of >1.65 eV are defined as the WBG perovskites, which can be acquired by mixing halide ions (Cl − , Br − , I − ). [41]he Zhu group recently summarized the lowest V OC deficit of WBG perovskites with different bandgaps.They suggested that the WBG perovskites with a bandgap of 1.75-1.83eV are subject to a large V OC deficit compared to those with a bandgap of 1.65-1.75eV. [41]The severe V OC loss of WBG perovskites caused by high defect densities and phase segregation under external stimulus deteriorates the performance of resulting tandem devices, considering that the WBG perovskites with a bandgap over 1.75 eV are customarily used to integrate the perovskite-organic TSCs owing to the current matching with OSCs.As is known, the V OC of perovskite-organic TSCs is determined to be the sum of the V OC of PSCs and OSCs, and the subcell with lower short-circuit current density (J SC ) restricts the J SC of tandem devices.Light management and optimization of interface contacts can enhance the J SC and fill factor (FF), respectively, which can be achieved by diminishing the optical loss and the resistance of carrier transport.While the V OC is associated with the quasi-Fermi level splitting (QFLS) of holes and electrons.These parameters are deeply related to the nonradiative recombination at the interface and bulk.In particular, the V OC deficit caused by interfacial recombination gradually dominates with increasing the bandgap from 1.5 eV to 2.0 eV, while the V OC losses originating from bulk recombination somewhat decrease (Figure 3B). [35]Therefore, suppressing nonradiative recombination at the surface, bulk, and grain boundaries of WBG perovskites is exceptionally prominent in reducing the V OC deficit and ameliorating the J SC and FF for perovskite-organic TSCs.
[43] For composition regulation, Qin et al. introduced the chloro-formamidinium into the perovskite precursor to reduce the iodine defects in the bulk perovskite film, yielding a V OC of 1.25 V for 1.78 eVbandgap PSCs. [29]Janssen and Sargent et al. found that the diammonium halide salt, propane-1,3-diammonium iodide, introduced during film fabrication can improve halide homogenization in WBG perovskite, leading to reduced nonradiative recombination and enhanced operating stability.A record V OC of 1.44 V was achieved in a bandgap of 1.97 eV PSCs, which is near 90% of the detailed-balance limit. [44]For interfacial passivation, the passivation of the nickel oxide with benzylphosphonic acid was employed to reduce the interfacial nonradiative recombination, achieving the V OC up to 1.26 V, which is the highest V OC for the PSCs with the bandgap of 1.79 eV at that time reported. [30]The buried interface between perovskite and charge extraction layer also plays a pivotal role in improving V OC .Particularly, the self-assembled monolayers (SAMs) with appropriate energy level alignment which can enlarge the QFLS in the perovskite have emerged as a promising alternative to the traditional HTL in PSCs. [12]Brinkmann et al. demonstrated that the perovskite on MeO-2PACz as the HTL yielded a 90 meV larger QFLS compared to their analogues on PTAA.Benefiting from the significantly decreased parasitic recombination by the introduction of SAMs as HTL, the tendency of halide segregation was substantially suppressed. [31]Very recently, Qiu's group reported an interfacial passivation strategy using Al 2 O 3 to modify the buried interface between MeO-2PACz and perovskite, and PEABr to modify top surfaces of WBG perovskite film with the bandgap of 1.77 eV, respectively. [36]The interfacial nonradiative recombination can be strikingly suppressed, resulting in the V OC of 1.33 V with the lowest losses of 0.44 V (Figure 3C).
Although the WBG perovskite devices have made impressive progress in the past decade, on the other hand, they still suffer from phase segregation due to the high Br content. [41,45]In other words, the I-rich and Br-rich domains can be formed under external stimulus, such as light, heat, and bias. [46,47]To address this issue, the lead thiocyanate and 2-thiopheneethylammonium chloride were adopted to passivate the defects originated from the bulk and interfaces of 1.79 eV-bandgap perovskites, respectively.The photoluminescent (PL) spectra results show that the control perovskite film experiences severe phase segregation under continuous illumination, while the passivated perovskite film exhibits negligible redshift, indicating the significant improvement of light stability using the synergistic passivation (Figure 3D,E). [37]Moreover, the bias-induced phase segregation in 1.77 eV-bandgap perovskites can be suppressed by substituting BCP with dense SnO x layer (Figure 3F,G). [36]Many models have been proposed to explain the phase segregation, such as polaron-induced lattice strain, [48][49][50] trapping-induced electric fields, [51][52][53] and halide oxidation. [54]Halide migration through halide vacancy defects is included in these models.For the Br-I system, the iodide oxidation process plays a more critical role in phase segregation than bromide oxidation process, considering that the iodide has a higher reducing capacity and could reduce the Br• radicals or Br 2 back to bromide, competing with the bromide being oxidized to generate Br• radicals or Br 2 under external stress conditions. [54]he iodide oxidation could trigger the formation of halide vacancy which is the pathway of ions migration.The oxidized species move across the surface and bulk of perovskite, reoccupying the perovskite lattice through a reduction reaction at vacancy sites.Ion migration can induce stoichiometric changes and element redistribution in perovskite, thereby leading to pronounced phase segregation. [55,56]The halide ions migrate to eliminate the vacancy, leading to I-rich and Br-rich domains.It is well known that degradation caused by moisture and oxygen has been a major issue for long-term stability of PSCs.Nevertheless, the significance of these factors may diminish if the device is encapsulated within a glove box protected by nitrogen gas.Furthermore, operational conditions under light illumination exacerbate ion migration.As reported, the I-rich domains can accelerate undesired carriers recombination, resulting in the V OC losses and poor operational stability. [35,54]In other words, the migration of ions resulted in the formation of bromide-enriched majority domains and iodide-rich minority domains, thereby establishing carrier traps that induce nonradiative charge recombination. [57]This phase segregation can be attributed to the localized strain resulting from the interaction between a single photoexcited charge and the soft and ionic perovskite lattice. [48]Therefore, reducing defect densities and suppressing the ions migration would be an effective method to further enhance the V OC and stability of WBG perovskite.

| BROAD-SPECTRUM ABSORPTION NBG OSCs
To further improve the PCE of perovskite-organic TSCs and optimize WBG perovskites with complementary absorption spectra of OSCs, it is also necessary to design NBG OSCs with extended near-infrared light absorption. [58]In the early stage, due to the restricted tunability of fullerene's energy levels, advancements in OSCs were hindered, reaching efficiency of approximately 11%. [59]The conventional fullerene acceptors exhibit inadequate light absorption and limited structural diversity, thereby impeding the fabrication of more efficient devices.On the contrary, various novel nonfullerene acceptors with diverse bandgaps and energy levels have been developed (Figure 4A). [60]For example, A-D-A type ITIC series (D = donor and A = acceptor) have F I G U R E 4 (A) Highest occupied molecular orbital/lowest unoccupied molecular orbital (HOMO/LUMO) energy levels of fullerene and nonfullerene acceptors.Reproduced with permission. [60]Copyright 2019, IOP Publishing Ltd.(B) Molecular structures of commonly used donor and acceptor for OSCs.Reproduced with permission. [61]Copyright 2021, Springer Nature.(C) Absorption spectra of PM6 and Y6 neat films.Reproduced with permission. [62]Copyright 2020, Elsevier.(D) Absorption spectra of typical nonfullerene acceptors.Reproduced with permission. [63]Copyright 2023, Springer Nature.(E) Complementary absorption spectra for ternary OSCs.Reproduced with permission. [64]opyright 2019, Wiley.OSC, organic solar cell.demonstrated significant potential as acceptors for fabricating high-performance devices. [65]But the trade-off between achievable V OC and charge generation efficiency in these devices hinders their overall performance.Subsequently, numerous additional studies have been published on single-junction A-DA'D-A-type Y6 seriesbased devices, demonstrating exceptional PCE exceeding 19%. [66]The commonly applied molecular structures of the adopted donor and acceptor materials in perovskiteorganic TSCs are shown in Figure 4B. [61]The nonfullerene acceptors are expected to have distinct advantages over fullerene acceptors, such as tunable bandgaps for exceptional near-infrared absorption, low energy loss, reduced driving force loss for charge separation, good compatibility, and blend morphology with donors.Developing new nonfullerene acceptors, advanced small molecules and polymeric donors would be a promising and fundamental solution to further improve OSC performance.Figure 4C displays the absorption spectra of donor PM6 and nonfullerene acceptor Y6.The ideal donor/acceptor mixture should have significant absorption coefficients encompassing the 400-950 nm wavelength or even beyond to amplify the electric current density. [62]Although the absorption band edge of the nonfullerene acceptor Y6 not being as red-shifted as some near-infrared molecules like ITIC-4Cl, BTPV-4Cl-eC9, and so on, Y6-based OSCs demonstrated more efficient exciton dissociation, enhanced charge transport, and minimal voltage loss. [67]As shown in Figure 2C, the theoretical PCE of 40% is achieved with the top cell bandgap of 1.6-1.7 eV and the bottom cell bandgap of ~1.0 eV.Assuming a tandem FF of 85% and a V OC loss of 0.5 V, the combination of a NBG OSC with an energy gap of 1.15 eV and a WBG PSC with bandgap of 1.75 eV offers the potential to achieve a PCE as high as 31.3%. [31]However, so far, efficient OSCs with a bandgap of ~1.0 eV have not been developed.Thus, as reported in the previous publication, [27] theoretically, a Y6-based rear subcell with a bandgap of ~1.41 eV would be more compatible with a WBG PSC having a bandgap of 1.95 eV.The challenge with lower bandgap acceptors lies not only in achieving extended absorption but also in maintaining efficient charge transport and minimizing losses associated with energy level mismatches. [27]igure 4D shows the absorption spectra of nonfullerene acceptors with various bandgap and corresponding absorption regions. [63]Li et al. developed an ultra-narrow bandgap nonfullerene acceptor BTPSeV-4F through the replacement of terminal thiophene by selenophene in the central fused ring of BTPSV-4F. [29]However, a single acceptor's absorption range is limited, which may limit the photon harvesting in active layers with one donor and one acceptor.In addition, implementing smaller bandgap nonfullerene acceptors as near-infrared absorbers leads to a decreased device voltage. [64]The ternary strategy has been proven to be a practical approach to enhance performance by merging three materials with complementary absorption spectra into a single active layer (Figure 4E). [68,69]In ternary OSCs, the incorporation of three distinct components into the photoactive layer offers opportunities for enhancing PCE.This can be achieved through strategies such as expanding the absorption spectrum, optimizing blend morphology, or fine-tuning exciton splitting and charge extraction mechanisms.Numerous works have been reported with the number of publications increasing speedily in the past decade (Figure 1B). [70]The increasing improvement in device performance is anticipated to lead to widespread adoption and commercialization of OSCs.Furthermore, the emergence of novel nonfullerene acceptors has redirected research focus towards TSCs.

| INTERCONNECTING LAYER IN PEROVSKITE-ORGANIC TSCs
The fundamental architecture of perovskite-organic TSCs involves stacking two subcells with complementary absorption properties.Besides perovskite and organic subcells, the ICL has a notable impact on the performance of perovskite-organic TSCs in terms of device efficiency, operational/mechanical stability, and scalability.Typically, ICL comprises a charge transport layer (CTL) with an electron transport layer (ETL), a hole transport layer (HTL), and a middle recombination layer.The effective ICL should have the following characteristics: (1) high optical transparency to suppress the parasitic absorption; ][73] ETL/HTL requires properties of high transmittance of light, low sheet resistance, and chemical stability against perovskite.Besides, the ETL and HTL also act as a barrier layer and solvent-resistant layer that goes against oxygen moisture and solvent damage for deposition of rear subcells, respectively. [74]Therefore, inorganic buffer layers are commonly employed, such as NiO x , MoO x , CuO x , and their doped variants for p-type buffers, along with ZnO, TiO x , SnO x , WO x , and their doped counterparts for n-type buffers.The recombination layer of the monolithic perovskite-organic TSCs has evolved from the solution process with poor reproducibility to ultra-thin metal (Figure 5).It should be pointed out that the present recombination layer materials used in perovskiteorganic TSCs mainly include ultra-thin metal layers (e.g., Ag and Au) or transparent conducting oxides, such as InO x and IZO.The high-performance ICL design aims to minimize parasitic light absorption, V OC loss, and carrier resistance to optimize the photovoltaic parameters of tandem devices.Advanced optical and electrical design strategies, such as charge transport characteristics and intermediate recombination effect, are employed to maximize photon distribution in the subcells and electron-hole extraction. [16]

| Charge transport layer in interconnecting layer
The HTL and ETL are deposited on the front of the WBG perovskite, thus requiring their fabrication at a low temperature (i.e., below 100-150°C) to mitigate the potential damage to the underlying subcell.In the n-i-p structure of monolithic perovskite-organic TSCs, the ICL involves HTL, recombination layer, and ETL.The HTL and ETL extract holes and electron carriers from the corresponding subcells, respectively.Developing CTLs, including HTL and ETL, with precisely aligned energy levels and inherently low energy disorder is crucial for eliminating the charge extraction barrier, thereby maximizing the QFLS in the tandem devices to increase the V OC .Kyaw et al. adopted MoO 3 as the HTL and PFN-Br as the ETL to establish ICL, exhibiting the highest vertical conductivity among the various interfacial materials for connecting two subcells, as shown in Figure 6A.High conductivity ICL is crucial for achieving  [24] Copyright 2022, Wiley.(I) Schematic illustration of the carrier transport properties of the tandem devices with a thin PCBM layer thickness.(J) The corresponding EQE performance of PCBM layer-based tandem device.Reproduced with permission. [72]Copyright 2022, Elsevier.(K) ICLs within different thicknesses of MoO x layers.Reproduced with permission. [27]Copyright 2020, Elsevier.HTL, hole transport layer; ICL, interconnecting layer; PSC, perovskite solar cell; TSC, tandem solar cell.efficient recombination in tandem devices.Meanwhile, the ICL based on PFN-Br exhibits a transmittance of over 90% in the wavelength range of 650-1000 nm, ensuring effective light absorption for the rear organic subcell (Figure 6B).They discovered that the Ag/PFN-Br interface reduced contact resistance due to efficient charge transport and restrained the nonradiative recombination.Eventually, a high-efficiency n-i-p perovskite-organic TSC with the ICL of MoO 3 /Ag/PFN-Br was successfully fabricated (Figure 6C), achieving a PCE of 20.6% (with a FF of 75.1%). [75]The imbalanced carrier distribution and transportation pose a significant challenge in impeding performance enhancement in tandem devices.As depicted in Figure 6D,E, the U-shaped distribution of charge carriers illustrates that these carriers were primarily concentrated at the interfaces between the active layer and the ETL/HTL.Their concentration gradually diminished within the inner layers of the active material.The hole density at the CsPbI 1.9 Br 1.1 perovskite/HTL interface was significantly lower than the electron density at the ETL/heterojunction interface.Such a significant offset would result in imbalanced charge carrier recombination within the ICL of the TSC when the two subcells are connected in series.
To further overcome this issue, the D18-Cl HTL, with high hole mobility and excellent hole extraction capabilities, established a quasi-Ohmic contact within the ICL (Figure 6F).This design mitigated hole losses during the extraction and transport, balancing charge carrier recombination within the ICL.In addition, the increased built-in potential and smaller surface contact potential difference (CPD) value observed in the D18-Cl-based TSCs confirms that the optimized electrical contact in the ICL can effectively extract holes to the Au recombination center and suppress the accumulated charge carriers (Figure 6G,H). [24]Tan et al. fabricated an ICL consisting of PM6/MoO 3 /Ag/PFM-Br, achieving a record-breaking PCE of 24.07% in inorganic perovskite-organic TSCs, which is among the best results for these types of solar cell. [76]Wang et al. employed PBDB-T-Si, D18, and polyTPD as HTL within the ICL to demonstrate the critical role of HTL. [23]The HTL of polyTPD was proved to be the optimal choice due to its minimal parasitic absorption, moderate hole mobility, and quasi-Ohmic contact properties, significantly suppressing the charge accumulation and voltage loss within the tandem device.
In the p-i-n structure of monolithic perovskite-organic TSCs, the ICL is fabricated with the structure of ETL/ recombination layer/HTL.Generally, the ETLs can be fullerene C 60 /BCP or PCBM/BCP layers, and the HTL is a thermal evaporated MoO x layer.Xie et al. developed a perovskite-organic TSC based on the PCBM/BCP/Ag/ MoO x ICL to investigate the carrier balance properties between ICLs architecture and 2-T perovskite-organic TSCs performance. [72]The hole could tunnel to the ETL side when the thickness of the ETL is thin (Figure 6I).The thin ETL could compromise the carrier modulating capability of ICLs due to inefficient holeblocking, deteriorating the performance of the tandem device.On the other hand, an extremely thick ETL can lead to inefficiency in extracting electrons to the ICLs and recombination with the holes extracted from the other side, resulting in a simultaneous reduction of external quantum efficiency (EQE) in the subcells (Figure 6J).For the HTL in ICL, a MoO x layer with a minimum thickness of 35 nm is necessary to prevent chloroform from damaging either C 60 or BCP (Figure 6K).These results show that the CTLs have a dual function: serving as both charge transport and protective barriers against solventinduced damage to the front layers.

| Metal recombination layer
There are multiple recombination layer options for perovskite-organic TSCs.Ultra-thin metal recombination layers are commonly employed to facilitate carrier recombination due to their simple vacuum-evaporated process (Figure 7A). [24]Figure 7B illustrates the distribution of absorbed photons for each subcell simulated using the transfer matrix method. [27]The front cell primarily captures visible photons within the range of 350-700 nm, whereas the rear cell absorbs near-infrared photons spanning from 700 to 900 nm.There was an overlapping region between 525 and 675 nm.The thin layer of Ag aids in reducing electrical losses by facilitating balanced and efficient carrier recombination.The optimal device shows a remarkable PCE of 20.6% in the reverse scan with a V OC of 1.902 V, a J SC of 13.05 mA cm −2 and a considerable FF of 83.1%.The high FF can be ascribed to the effective charge recombination in the ICL.Ultrathin Au, selected for its superior conductivity and lower reactivity than commonly used Ag, was employed as the charge recombination center. [22,37,78]1 nm thick Au was deposited in a vacuum onto the PBDB-T/MoO 3 HTL, followed by spin-coating of ZnO ETL to form an ICL (Figure 7C).The ideal energy level improves carrier concentration in tandem devices, thereby promoting the extraction of charges from ICL and ultimately facilitating charge recombination in the ICL. [22,79]he combination of Ag and PEDOT:PSS for hole extraction from the rear cell may lead to corrosion and device instability. [80,81]Baran et al. investigated the operational stability of tandem devices based on three types of ICL, without Ag, Ag 0.5 , and Ag 1 (0.5 and 1 nm of Ag remarked as Ag 0.5 and Ag 1 , respectively) (Figure 7D).Within the first 100 h, all tandems undergo a declining period.However, the device with an Ag-free ICL displays a slower degradation of PCE compared to the Ag-containing devices. [77]Notably, an extra SnO x layer fabricated through atomic layer deposition (ALD) at a low temperature is incorporated as the ETL, which can block the diffusion of the metal to the perovskite subcell. [82,83]In state-of-the-art all-perovskite TSCs, the ICLs typically consist of SnO x and evaporated thin metals or sputtered thick (typically 10-100 nm) transparent conducting oxides. [84,85]The ICLs structure including SnO x as ETL and Au as recombination layer were also incorporated to construct the perovskiteorganic TSCs (Figure 7E,F).Au as the recombination center was preferred over Ag because the ultrathin Ag film is susceptible to oxidation during aqueous PEDOT:PSS deposition in the air. [37,78]

| Transparent conducting recombination layer
Although ICLs based on ultra-thin metal layer (Ag or Au) can function as adequate recombination layers, the high degree of optical loss of these metal layers may limit the J SC of the rear subcell, thereby reducing the PCE of the perovskite-organic TSCs.Therefore, designing new ICLs is crucial to ensure the effective recombination of carriers while minimizing optical or electrical losses.It was found that transparent conducting metal oxide recombination layer, such as ITO, ZnO, and IZO, have higher transparency than metal recombination layer (Au and Ag).This is because metallic materials have high parasitic absorptivity, suggesting that metal oxide recombination layers are advantageous in improving the PCE of tandem devices.He et al. optimized an ICL structure by inserting a 4-nm-thick sputtered layer of IZO between organic C 60 /BCP and MoO x to achieve improved electrical properties and transmittance in the nearinfrared region (Figure 8A). [30]As delineated in Figure 8B, the near-infrared transmittance of IZO-based ICLs was significantly higher than that of Ag-based ICLs.EQE measurements demonstrate the superiority of replacing a thin metal-based ICL with a thin IZO-based ICL.The decrease in optical loss and current loss within the 700 to 900 nm range led to a substantial increase in F I G U R E 7 (A) Device structure of the perovskite-organic TSC highlighting the structure of the ICL.Reproduced with permission. [24]opyright 2022, Wiley.(B) Tandem cell structure with a highlight on ICL design and distributions of photon absorptions in subcells, simulated by transfer matrix method.Reproduced with permission. [27]Copyright 2020, Elsevier.(C) The energy-level diagrams of the 2-T TSC with shown charge recombination process in the ICL.Reproduced with permission. [22]Copyright 2021, Wiley.(D) Photostability test of the TSCs with three different ICLs (without Ag, Ag 0.5 , and Ag 1 ).Reproduced with permission. [77]Copyright 2022, American Chemical Society.(E) Schematic representation of the perovskite-organic tandem device structure with ALD SnO x /Au functioning as recombination center.Reproduced with permission. [78]Copyright 2023, Wiley.(F) The structure of 2-T TSCs with synergetic treatment and corresponding cross-section SEM image and EQE of devices.Reproduced with permission. [37]Copyright 2022, Wiley.2-T, two-terminal; EQE, external quantum efficiency; ICL, interconnecting layer; TSC, tandem solar cell.
J SC of 1.46 mA cm −2 (Figure 8C).The device with IZObased ICLs showed excellent current matching between the rear and front subcells, enabling perovskite-organic TSCs with a high J SC of 14.87 mA cm −2 and a PCE of 23.60% (certified 22.95%).However, the high energy of the sputtered transparent conducting metal oxide can potentially cause damage to the underlying layer and lead to a degradation in device performance.To mitigate the potential damage caused by high-energy sputtering, a protective layer is typically applied before initiating the sputtering process.Tan et al. demonstrated the utilization of cross-linkable organic molecules as a protective layer during sputtering ITO for the recombination layer in perovskite-organic TSCs.By combining a 1.77 eV bandgap perovskite with a 1.35 eV bandgap organic active layer, a perovskite-organic TSCs was successfully demonstrated, exhibiting an impressive 24.07%PCE and exceptional stability. [86]rinkmann et al. fabricated perovskite-organic tandem devices with SnO x /recombination layer/MoO x as ICL, as shown in Figure 8D.A prominent S-shaped J-V characteristic was found, indicating the formation of the Schottky barrier at the SnO x /MoO x interface (Figure 8E).A thin layer of metal (~1 nm Ag or Au) is normally incorporated into the interface between the SnO x and MoO x to make the interconnection ohmic contact. [87,88]owever, even an Ag layer as thin as 1 nm introduces significant optical losses that reduce the rear cell's EQE and the tandem device's overall J SC , as shown in Figure 8F,G.An ultrathin ALD-grown InO x layer (~1.5 nm) was developed as a recombination layer to replace the metal layer.The insertion of InO x between SnO x and MoO x showed outstanding electrical and optical properties, significantly enhancing the EQE of the organic rear subcell and increasing the overall J SC by approximately 1.5 mA cm −2 for the tandem compared to the scenario using a 1 nm-thick Ag-based ICL.Consequently, a champion tandem cell with a stabilized PCE of 24.0% has been attained.In addition, ALD enables largearea, high-throughput processing and conformal deposition of textured surfaces commonly used in light trapping concepts. [89,90]This compact layer acts as a permeation barrier to enhance long-term stability and safeguards the underlying layers against chemical attack from solvents used in subsequent processes. [91,92]Thus, such an ICL structure is not confined to perovskite-organic tandem devices.It can also be exploited for good applications in other tandem devices.Reproduced with permission. [30]Copyright 2022, Springer Nature.(D) Schematic of perovskite-organic TSCs with InO x or Ag as interconnect and (E) J-V characteristics of SnO x /(InO x )/MoO x junction with and without InO x .(F) Optical transmittance of an interconnect based on approximately 1.5 nm of InO x , bare and sandwiched between SnO x and MoO x .For comparison, the InO x has been replaced by a nominally 1 nm thick layer of Ag. (G) The resulting EQE spectra of the organic rear cell with InO x or Ag as an interconnect demonstrate the notable current losses induced by only 1 nm Ag.Reproduced with permission. [31]Copyright 2022, Springer Nature.EQE, external quantum efficiency; ICL, interconnecting layer; TSC, tandem solar cell.
Although a relatively small V OC loss and a high FF can be achieved with both transparent conducting oxides and ultrathin metal recombination layer, these ICLs structures are still complicated and the relatively high parasitic absorption could not be avoided.ALD-deposited thick SnO x and sputtered ITO have been normally used to protect the underlying cells from damage during the solution processing or sputtering of front cells. [84,85,91]For all-perovskite tandem cells with the incorporation of SnO x , the PCE could be initially achieved to 17.0% when two perovskite cells with the bandgap of 1.20 and 1.80 eV were integrated. [93]Recently, the PCE of all-perovskite tandem devices has been incredibly enhanced to 29.1% through optimizing the ICLs, WBG, and NBG perovskite. [1]owever, the presence of the ITO or metal leads to parasitic loss, limiting the J SC of tandem devices.On the other hand, the sputtering process with high-energy ions has an unfavorable effect on the underlying layers, reducing the FF of device.To overcome these limitations, a simple ICL structure of C 60 and SnO 1.76 was reported for obtaining a high PCE of all-perovskite tandem devices. [94]hey pointed out that C 60 is n-doped by iodide ions from underneath perovskite, and the SnO 1.76 has an ambipolar transport property.Figure 9A-D show that similar photovoltaic performance was obtained when SnO 1.76 serves as HTL or ETL compared to PEDOT:PSS or C 60 /BCP, respectively.The combination of doped-C 60 and ambipolar SnO 1.76 could produce desirable ohmic contact, effectively balancing the transportation of charge carriers.As a result, a high PCE of 24.4% was achieved for all-perovskite tandem devices.
The same issues also happen in the perovskite-organic tandem devices.The metal Ag causes corrosion when combined with PEDOT:PSS extracting hole carries from the organic subcells. [95,96]To solve this issue, Baran et al. employed the ICL of C 60 , ALD-deposited SnO x and PEDOT:PSS to investigate the influence of metal-free ICL on the performance of perovskite-organic tandem devices (Figure 9E).The perovskite with the bandgap of 1.78 eV exhibits a V OC of 1.18 V. Consequently, a PCE of 17.6% for the tandem device was obtained using metal-free ICL, which is comparable to the Ag-based device (Figure 9F). [77]Very recently, Qiu et al. optimized the WBG perovskite subcell through interfacial passivation to fabricate high-performance perovskite-organic tandem devices.The champion V OC of 1.33 eV could be obtained for the perovskite with bandgap of 1.77 eV.Besides, they simplified the ICL structure by substituting Au with ALDdeposited SnO x , as shown in Figure 9G.A high PCE of over 22% was obtained for perovskite-organic tandem devices based on the metal-free ICL of PCBM/SnO x / PEDOT:PSS (Figure 9H). [36]Nevertheless, the PCE of perovskite-organic tandem devices based on metal-free ICL is still so far behind that of metal-based perovskiteorganic and all-perovskite tandem devices.The simplified ICL provides a unique platform to further increase the J SC of perovskite-organic tandem devices.Numerous efforts will be made to further enhance the PCE of perovskiteorganic tandem devices based on metal-free ICL and close the PCE gap between perovskite-organic and other tandem devices.

| CONCLUSIONS AND PERSPECTIVES
The progress in perovskite-organic TSCs is a testament to the ongoing advancements in photovoltaic technology.By addressing the limitations of single-junction cells, researchers are inching closer to realizing highly efficient solar energy conversion by constructing tandem devices.The low V OC of WBG PSCs and NBG OSCs limits the V OC of perovskite-organic TSCs, which in turn limits the PCE of tandem devices.Consequently, high-efficiency WBG PSCs and NBG OSCs with minimized V OC losses must be fabricated to increase the perovskite-organic TSCs efficiency.On the other hand, current matching is another crucial factor in fabricating efficient monolithic 2-T TSCs.The currently used NBG organic materials in perovskite-organic TSCs exhibit limited near-infrared light absorption compared to Pb/Sn perovskite, which is the main reason for the lower J SC in perovskite-organic TSCs.Synthesis of novel organic materials with exceptional broad-wavelength light conversion ability is highly anticipated.The ICL has a significant impact on the FF, J SC and V OC of a monolithically tandem device.The ICLs connecting the rear and front subcells should possess the properties of high light transmittance, ohmic contact with adjacent layers, and protection barrier.The excess charge carrier could be recombined in the ICLs to guarantee the balance of charges transportation.
In this review, we summarized the recent achievements of perovskite-organic TSCs.Specifically, the regulation strategies of WBG perovskites and NBG organic semiconductors as well as the developments of ICLs were discussed.The V OC losses of WBG perovskite devices are the main factor that impedes the further enhancement of V OC in perovskite-organic TSCs.It can be mitigated by strategies such as additive engineering and interfacial passivation, which could significantly reduce the imperfections in the bulk and grain boundaries, thus suppressing the nonradiative recombination and decreasing the V OC deficit.On the other hand, integrating various strategies has achieved single-junction OSCs with PCE exceeding 20%, such as the utilization of novel nonfullerene electron acceptors with strong absorption in the solar spectrum, minimal loss in driving force during charge separation, film morphology optimization, effective interfacial layers, and precise device engineering.We highlight the design of new nonfullerene acceptors with high light absorption and feature adding a third component in the active layer to achieve spectral complementation.Furthermore, we reviewed three types of ICLs, namely metal-based, transparent conducting oxide-based, and metal-free ICLs.
Although considerable progress has been made in TSC technology, various challenges persist.Especially the long-term stability of perovskite materials is of particular concern.One of the essential reasons is that WBG PSC with high Br content is prone to suffer from phase separation and V OC loss caused by ion migration within perovskite.To better comprehend phase segregation and achieve high-efficiency PSCs with enhanced intrinsic stability, we present a concise overview of several research directions aimed at suppressing ion migration.
(1) Innovation in component design and device structure.Introducing the functional buffer layer atop or bottom of perovskite to form a permeation barrier that can prevent organic species from escaping, reduce bulk and surface defects, and mitigate phase separation.(2) Developing a universally applicable technique to obstruct ion migration F I G U R E 9 Schematic diagram of (A) NBG perovskite and (C) WBG perovskite single-junction solar cells, respectively.J-V curves of the corresponding (B) NBG and (D) WBG single-junction solar cells, respectively.Reproduced with permission.Copyright 2020, Springer Nature.(E) Perovskite-organic monolithic tandem stack.(F) J-V curves of the champion devices for each tandem stack.Reproduced with permission. [77]Copyright 2022, American Chemical Society.(G) Design of perovskite-organic tandem cell structure with a highlight on a simplified ICL design and (H) J-V curves of the champion tandem device with different scan directions.Reproduced with permission. [36]opyright 2023, Wiley.NBG, narrow-bandgap; WBG, wide-bandgap.
pathways.Perovskite films contain various imperfections that may be introduced during the preparation process.These defects and grain boundaries act as pathways for ion migration.Thus, reducing the number of defects and grain boundaries is essential for suppressing ion migration.(3) The phase segregation is likely to occur in mixedhalide perovskites that include Br atoms.This creates an obstacle to the stability of PSCs.Besides, replacing the A-site can be an appropriate method to mitigate ion diffusion and improve stability, for example, incorporating a large organic cation spacer (ethane-1, 2-diammonium dihydroiodide, Phenylethylamine Hydroiodide, etc.) to form 2D/3D perovskites typically results in an increased bandgap. [97]In that case, it is not necessary to incorporate abundant Br to increase the bandgap, which is beneficial in suppressing the phase segregation.
The low bandgap nonfullerene-based OSCs can absorb broader spectra, while they often exhibit considerable losses in their V OC values.Therefore, it is imperative to develop novel organic materials with lower V OC loss to enhance the performance of OSCs, thereby increasing the performance of resulting perovskite-organic TSCs.Future investigations could focus on synthesizing new donors and acceptors with high solubility and aggregation properties to enhance exciton generation and optimize the ternary structure to overcome the limitations undermining binary OSCs' efficiency.Especially the ternary structure comprising three different components offers the advantage of facile implementation and has provided impressive results in terms of efficiencies.Elucidating the morphology of ternary mixtures and its impact on working mechanisms poses a formidable challenge, requiring advanced analytical techniques.Additionally, employing systematic theoretical approaches to manipulate ternary mixture components and composition is recommended for establishing correlations between molecular design and device performance.The exploration of developing ternary OSCs with ratiointensitive donor-to-acceptor to improve reproducibility is warranted.Additionally, the enhancement of efficiency can be further achieved through the development of novel donor materials that exhibit excellent compatibility with Y-series acceptors.Eventually, the stability of OSCs, such as their resistance to oxygen, moisture, and heat, must be systematically studied by examining their chemical structures, photophysical properties, and interface modifications.
Currently, the utilization of interconnecting layers in TSCs, such as embedded metal thin layers (Au, Ag), transition metal oxides (ITO, IZO, MoO x ), and ALD SnO x often necessitates high evaporation temperatures and prolonged deposition times.However, these requirements inevitably result in detrimental effects on the underlying perovskite layer, including volatilization of organic molecules and lattice deformation, thereby compromising the stability of the device.To effectively address this issue, it is imperative to identify materials with lower evaporation temperatures and develop simplified preparation processes that enhance the reproducibility of tandem device fabrication.Moreover, substituting the commonly used HTL (MoO x , PEDOT:PSS) with newly developed hole extraction materials in ICL could potentially enhance the efficiency of perovskite-organic TSCs. [2]For example, the SAMs, such as 2PACz, MeO-2PACz, and Me-4PACz, [98][99][100][101][102][103][104][105] have the superiorities of ultra-thin thickness and high transparency, which can significantly enhance the transmittance of light.
Additionally, it is crucial to develop cost-effective fabrication techniques and scalable manufacturing processes to ensure practical deployment.Although the process is more facile than standard all-perovskite tandems, we should minimize the change between solution and nonsolution processing for easier manufacturing.This can be achieved through the development of linear thermal evaporation or fully solution-processed systems. [27]

F
I G U R E 5 The evolution of the interconnecting layer for monolithic perovskite-organic tandem solar cells.F I G U R E 6 (A) J-V curves of various ICL structures: MoO 3 /Ag/PFN-Br, MoO 3 /Ag/PDINN, MoO 3 /Ag/PDINO, and MoO 3 /Ag/ZnO NPs.(B) Transmittance spectrum of MoO 3 /Ag/PFN-Br ICL.(C) The cross-section SEM image of TSC.Reproduced with permission. [75]Copyright 2022, Wiley.Drive-level capacitance profiling results of (D) the single-junction PSC with a configuration of ITO/SnO 2 /CsPbI 1.9 Br 1.1 /PM6/ MoO 3 /Al, and (E) those of the single-junction OSC with a structure of ITO/ZnO/PFN/D18-Cl:N3:PC 61 BM/MoO 3 /Al, measured at a high alternating current frequency of 100 kHz.(F) The conductivity of ICL with ITO/HTL/MoO 3 /Au/ZnO/PFN/Al structure.(G) Mott-Schottky plots of the TSCs with different HTLs obtained under dark conditions.(H) Schematic of the Kelvin probe force microscopy (KPFM) measurement, KPFM images of the interface between HTL and MoO 3 /Au along the white dotted line, and CPD values obtained from the abrupt change in the surface potential.Reproduced with permission.

F
I G U R E 8 (A) Design of ICLs with either transparent conducting oxide or metal in perovskite-organic TSCs.(B) Transmittance spectra of the perovskite subcells with 4 nm IZO-based and 1 nm Ag-based ICLs.(C) EQE spectra of the TSCs with IZO-based and Ag-based ICLs.
Summary of n-i-p type perovskite-organic tandem solar cells.Summary of p-i-n type perovskite-organic tandem solar cells.