A Comprehensive Review of Tandem Solar Cells Integrated on Silicon Substrate: III/V vs Perovskite

High-efficiency solar cells with low manufacturing costs have been recently accomplished utilizing different technologies. III-V-based tandem solar cells have exhibited performance enhancement with a recent efficiency of greater than 39% under AM1.5G and 47% under concentration. Integration of such III-V materials on a relatively cheap Silicon (Si) substrate is a potential pathway to fabricate high-efficient low-cost tandem solar cells. Besides, perovskite solar cells, as third-generation thin film photovoltaics (PV), have been meteorically developed at a reasonable cost. At present, there are still questions for cost reduction of perovskite materials and solar cell modules because of their limited commercialization. In this review, stacking Si solar cells with III-V material to form Si-based III-V tandem solar cells is presented with different integration technological routes. Also, perovskite/Si tandem solar cells have been reviewed alongside their main engineering challenges introduced through the fabrication of perovskite-based tandem solar cells. Finally, a comparison between III-V tandem solar cells, Si-based III-V tandem solar cells, and perovskite-based tandem solar cells is introduced so that the best technology for a specific application could be determined. The review provides a comprehensive study of two different technologies (III/V and Perovskite) to demonstrate the most valuable cost reduction availability for each.


Introduction
The conversion efficiencies of conventional solar cells were intensely optimized but, at the same time, they cannot exceed a specific limit.According to Goetzberger et al., the limiting efficiencies of several types of solar cells can be predicted using extrapolating the past development of each type [1].This extrapolation method shows that the progress of the efficiency will congregate soon, which is primarily constrained by the Shockley-Queisser limit [2].To increase the efficiency of a solar cell (SC), a strategy of stacking different semiconductor materials, each with a different bandgap, has been utilized to absorb a wide band of the solar spectrum.Such architecture is called a tandem solar cell by which the thermalization loss is reduced due to the good absorption of the near-infrared region by small-bandgap materials [3].The III-V SCs have been successfully implemented via this approach since III-V semiconductors have a proper range of bandgaps and lattice constants to choose from.NREL achieved 32.9% efficiency using GaInP/GaAs two-junction cell [4].Great attempts have been done to increase the efficiency of a tandem III-V SC.Efficiencies of 37.9% (GaInP/ GaAs/GaInAs 3-junction) and 39.5% (GaInP/GaAs(mqw)/ GaInAs 3-junction), as well as 39.2% with 6-junction under 1-sun AM1.5G were achieved [4][5][6].Although III-V solar cells accomplish the ultimate efficiencies amid all photovoltaic technologies, their market share in electricity generation is very small due to their high manufacturing cost.The most considerable cost contributor of III-V solar cells is the substrate.Ge or GaAs substrates are mostly utilized for the growth of III-V SCs.Replacing such substrates with Si is a promising technique to decrease the manufacturing cost [7,8] paving the way to a relatively low-cost solar cell [9].As Si has a narrower bandgap than some III-V materials, adding it to the III-V can extend the absorption edge toward the lower wavelength part of the solar spectrum.In addition, Si offers higher thermal conductivity in comparison to Ge or GaAs [10].However, the main challenge lies in the larger Silicon (2023) 15:6329-6347 1 3 differences in thermal expansion coefficients as well as lattice constants between Si and III-V materials.Some research studies on integrating III-V semiconductors on Si substrate for photovoltaic applications were first introduced in the 1980s [11].In the last few years, these studies have been retrieved relating to the research on wafer-bonding [12][13][14], mechanical stacking [15][16][17] techniques, and new metamorphic buffer approaches [18][19][20].Recently, the highest achievable certified efficiency for a Si-based III-V is 35.9% by mechanically stacking a Si cell to a GaInP/GaAs cell measured under one-sun AM1.5G spectrum [4,17].
Furthermore, the interest in tandem SCs has increased due to the dramatic advances in perovskite, a new class of thin-film PV materials, first discovered in 2009 [21].Perovskites (PVKs) are considered a perfect partner in tandem solar cells because of their direct bandgap with excellent absorption coefficient, bandgap tunability, and high radiative recombination efficiencies [22][23][24].According to the lowcost fabrication processes and high efficiency, perovskite solar cells (PSCs) drew huge interest and rapid development.So far, at the end of 2022, the highest certified efficiency of PVK/Si is greater than 32% [25] In this review, different technologies to fabricate tandem SCs are reviewed.Firstly, we review the latest progress in Si-based III-V tandem solar cells along with their advantages and challenges.Next, the recent research progress, in order to enhance the efficiency of PVK/Si is presented.Thus, the most important structural, electrical and optical characteristics of the PVK/Si are investigated.This review is structured as follows, Sect. 2 discusses several technological techniques for integrating III-V SCs over Si substrate through heteroepitaxial integration, mechanical-stacking, wafer-bonding, and smart-stacking techniques.In Sect.3, we review the progress toward the present status of PVK/Si tandems in terms of efficiency and device design in addition to their main challenges.Next, in Sect.4, all reviewed tandem technologies are compared regarding efficiency, cost, and longterm stability.Also, advances in technologies to meet the addressed challenges are introduced.Finally, we draw the conclusions from this work in Sect. 5.

III-V/Si Tandem Solar Cells
As aforementioned, III-V solar cells slightly share the market due to their high manufacturing cost.Substituting the heteroepitaxial growth substrate with Si is a potential route so that the production cost can be reduced [2,26].The key issues are the lattice mismatch and thermal expansion coefficient difference between Si substrate and the III-V semiconductors [27].The monolithic III-V tandem cells need lattice matching between all epitaxial layers to sustain good crystallinity, which is necessary to achieve high-ranking efficiency [2].This section presents the latest development of Si-based III-V tandem SCs.The epitaxial growth of different III-V materials such as AlGaAs, GaAsP, GaInP and on Silicon to fabricate two-junction or more than two-junction solar cells are reviewed.Also, the fabrication of the Si-based tandem solar cell using smart-stacking method is introduced.

Two-Junction Tandem Solar Cells
In this subsection, we review the III-V/Si two-junction tandem cells.The fabrication of III-V/Si is extremely difficult due to the distinction in both lattice constant and thermal expansion between III-V material and Si.These differences cause a high density of dislocation [15,28] which results in recombination centers, i.e., a reduction in efficiency.Much research presented different methods and approaches to advance this challenge using different III-V materials such as GaAs [17,29,30], AlGaAs [31][32][33], GaInP [15][16][17], and GaAsP [34-36].

GaAs
GaAs solar cells have the advantage of high efficiency potential, possibility of thin-film, radiation resistance and potential of multijunction applications [37].The high efficiency of 29.1% had been obtained for GaAs single junction solar cell [4].Due to the limiting efficiency of GaAs single-junction solar cell, the lack of the multijunction cell is significant for high-efficiency solar cell.In this section, we present the development and progress of GaAs/Si tandem cell.In 2002, TAGUCHI et al. fabricated GaAs/Si tandem cell by epitaxial lift-off (ELO) technique without degrading the crystal quality [29].They obtained 19.4% efficiency of tandem cell which is relatively low.Further, in 2017, Essig et al. achieved a 32.8% efficiency using mechanically-stacked GaAs//Si tandem cell under one-sun illumination using AM1.5G spectrum [17].In contrast, VanSant et al. introduced a GaAs//Si tandem cell by replacing traditional metalorganic vapor phase epitaxy (MOVPE) with a lower-cost hydride vapor phase epitaxy (HVPE) [30].Their study reported on the performance of an HVPE-grown GaAs top cell incorporated into a GaAs//Si tandem device with an efficiency of 29%.Also, they followed a structural optimization technique to increase the efficiency to 31.4%.

AlGaAs
The first AlGaAs/Si monolithic two-terminal (2-T) tandem SC was fabricated in 1991 by Umeno et al. using thermal cycle annealing [32].The device consists of two sub-cells (Al 0.1 Ga 0.9 As and Si), Al 0.8 Ga 0.2 As as a window layer, and GaAs as a buffer and contact layer between the two sub-cells (Fig. 1a).The authors reported that thermal cycle annealing is efficient in reducing the density of dislocations, which results in higher efficiency.This cell exhibited 12.8% efficiency.The improvement of the quality of AlGaAs grown on Si by Metal Organic Chemical Vapor Epitaxy (MOCVD) using a Graded Band Emitter Layer is reported in [38].The Graded band emitter layer is used to boost the carrier collection efficiency which in turn increases the efficiency by the internal field.The tandem cell efficiency was measured at 1-sun AM0 and recorded an increase to 19.9% and 20.6% for 2-T and 4-T cells, respectively.Further, Veinberg-Vidal et al. presented a fabrication process of a monolithic 2-T Al 0.2 Ga 0.8 As//Si tandem solar cell using surface-activated direct wafer-bonding which results in an amorphous layer between the GaAs layer and the Si cell [39].
A unique method was established to achieve an III-V/ Si two-junction cell by merging both ELO and print-transfer-assisted bonding techniques.This method is applied to Al 0.3 Ga 0.7 As (1.8 eV) which is the best top subcell bandgap corresponding to the Shockley-Queisser efficiency limit [33].Figure 1b shows a detailed schematic layer structure of a two-junction Al 0.3 Ga 0.7 As/Si SC.This tandem cell gives open-circuit voltage (V OC ) = 1.63 V, short-circuit current density (J SC ) = 11.9 mA/cm 2 , fill factor (FF) = 65%, and efficiency = 12.7%.Also, tests of Al 0.3 Ga 0.7 As/Si cell under range (0.25-sun to 2.5-sun) incident light power densities were carried out.However, AlGaAs suffers from oxygen-related defects which acts as very active defect center because it has higher capture cross section.Ando et al. confirmed that the oxygen-related defects act as a non-radiative recombination center [40].The non-radiative recombination centers decrease the External Radiative Efficiency (ERE) which directly reduce the efficiency of the multijunction solar cell [41].Table 1 summarizes the development in the fabrication of AlGaAs/Si tandem solar cells showing the performance parameters of the solar cell (if available) and the interconnect method for each.

GaInP
It has been demonstrated that single-junction SCs, based on GaInP, attained efficiencies over 20% at 1-sun illumination intensity [42], which is a good condition to reach 30% efficiency for a Si-based two-junction SC [15].The fabrication of a monolithic GaInP/Si tandem cell [7] is a challenging process because the material quality of the front subcell deteriorates from the thermal expansion coefficient discrepancies and lattice constants mismatch between Si and GaInP [43].Another approach to producing a GaInP/Si solar cell is wafer-bonding [44,45], by which two mirror-polished surfaces are made in contact and stick to each other perpetually [46].This approach has the advantage of fabricating III-V and Si sub-cells separately and to be joined at low temperatures.A third approach to fabricating the GaInP/Si tandem solar cell is mechanical stacking, in which either a direct metal interconnect or transparent adhesive between the sub-cells is used [16,17].
In 2015, Essig et al. fabricated and modeled a mechanical-stacked GaInP/Si tandem SC (Fig. 2a) using two different conditions, a series connection, and an independent connection and an efficiency of (27.1 ± 0.8)% has been reached.The work reported that the efficiency can be increased by Fig. 1 a Cross-Sectional illustration of AlGaAs/Si tandem SC reported by [32], and b layer structure of the two-junction tandem Al 0.3 Ga 0.7 As//Si SC.Reprinted from [33]  enhancing the red response of the Si rear subcell and utilizing light trapping techniques [15,16].Such performance has been improved by using a-Si:H/c-Si heterojunction SCs (SHJ) [17].A summary of the performance key factors of some GaInP/Si tandem SCs is listed in Table 2.The performance measure includes V OC , J SC , FF, and efficiency.In addition, the interconnection technique is presented.

GaAsP
GaAsP semiconductor material is another candidate to be applied as a front subcell in the Si-based III-V tandem solar cells.According to the detailed-balance theoretical modeling, GaAsP/Si (~ 1.7/1.12eV) has an ideal bandgap profile which is remarkably desirable thanks to its lower complexity and decreased quantity of required epitaxy [19].Monolithic GaAs 0.75 P 0.25 /Si SCs, grown through both MOCVD and Molecular Beam Epitaxy (MBE), are reported for the first time by Grassman et al. in 2015.Later in 2019, Fan et al. presented an NREL-certified efficiency of 20.0% for an epitaxial GaAs 0.75 P 0.25 /Si tandem cell [47].This efficiency was obtained using the structure shown in Fig. 2b.The structure consists of GaAsP top cell, Si bottom cell, AlGaAsP/GaAs tunnel-junction, and GaAsP layers as Compositionally Graded Buffer (CGB) with a specific thickness.
The transparent CGB layers are so important that the threading dislocation density (TDD) could be reduced which in turn reduces the current mismatch between the front and rear subcells [48].
Later, in 2021, Gen-5 GaAsP/Si tandem SCs were fabricated by III-V MOCVD growth technique.This cell generation (Fig. 2c), utilizing an ex-situ produced Si sub-cell with B-diffused BSF, an Al 2 O 3 /TiO x ARC, and in the combination of the GaAsP top cell optimizations, yielded remarkable enhancements in both V OC and J SC in excess of the earlier generations [49].Using the aforementioned design results in a 23.4% NREL-certified AM1.5G efficient GaAsP/Si tandem solar cell [4].Furthermore, the same research group has improved the low-TDD GaAsP/Si structure of [49] that enhanced the J SC and FF without sacrificing the V OC .These improvements enable about 27% efficient GaAsP/Si tandem solar cells [50].

Three-Junction GaInP/GaAs/Si Cell
The integration of III-V multi-junction and Si SCs by either direct epitaxial growth, wafer-bonding, or mechanical stacking have been extensively studied to develop PV devices with high conversion efficiencies.In this sub-section, we present the development in the performance of the GaInP/ GaAs/Si SCs.In 2014, Dimroth et al. presented a GaInP/GaAs twojunction SC on Si using two different process technologies to be fabricated, direct growth and wafer-bonding [7].The SC configuration is displayed in Fig. 3a.It consists of GaInP/ GaAs two-junction grown by MOVPE technique and connected to Si using n + GaAs/p + AlGaAs tunnel diode.The direct growth process produced 16.4% efficiency while utilizing the wafer-bonding process, an efficiency of 26.0% could be achieved.This relatively higher efficiency is due to low TDD and good crystal quality.Further, a 2-T GaInP/ AlGaAs/Si three-junction SC was realized using surfaceactivated wafer bonding [51].In this realization, the Aluminum composition was varied to be optimized.Additionally, an improved Si back-side passivation was presented.Efficiencies of 24.7% and 26.2% were obtained when using Al content of 7.5% and 3.5%, respectively.Cariou et al.
achieved efficiencies up to 33.3% for a 2-T GaInP/GaAs// Si device using the approach of direct wafer-bonding [52].They introduced two innovative features, Si passivating contacts and diffraction grating.In 2017, Essig et al. achieved a 35.9% efficiency using mechanically stacked GaInP/GaAs/ Si 3 − junction tandem cell under one-sun illumination using AM1.5G spectrum [17].Later, in 2018, Feifel et al. presented a direct growth of III-V/Si three-junction SCs having 19.7% efficiency [18].In their work, SiN x was used as a diffusion barrier on the rear surface of the wafer to prevent deterioration of the minority carrier lifetime in the Si subcell.Also, a metamorphic GaAs y P (1-y) buffer layer lets rise the lattice constant from Si to GaAs (Fig. 3b).They improved this structure using improved nucleation conditions proposed for the first GaP layer (Fig. 3c) [8].This improvement led to an efficiency of 22.3%.Ray-Hua Horng Fig. 3 a Representation of two approaches for the integration of highefficiency GaInP/GaAs on silicon.Reprinted from [7].b The main structure of the devices by Markus Feifel et al. [18], c improved structure by Markus Feifel et al. [8], d schematic architecture of GaInP/ GaAs//poly-Si triple-junction SCs in [51], and e Schematic representation of layer stack of the III-V//Si triple-junction SC design used by [55] et al. present a new mechanical-stacking fabrication method to bond the GaInP/GaAs and polycrystalline silicon SCs (Fig. 3d).GaInP/GaAs/poly-Si solar cells were observed to reach a Power Conversion Efficiency (PCE) of 24.5%, with V OC of 2.68 V, J SC of 12.39 mA/cm 2 , and FF of 73.8% [53].
Further, transparent Al x Ga 1-x As y P 1-y step-graded metamorphic buffers were explored to enhance the transmittance significantly, while there is no remarkable change in the TDD is observed.This leads to a record efficiency of 25.9% under AM1.5G illumination [54].By the Fraunhofer Institute for Solar Energy Systems, a 2-T GaInP/(Al)GaAs/Si SC has been enhanced with an efficiency of 34.1% [55].The improvement comes from using Tunnel Oxide Passivated Contact (TOPcon) on the front and back sides of the Si subcell.In addition, replacing the inverted-grown GaInP homojunction to uprightgrown back heterojunction as well as changing the composition of Al in the middle subcell not only increase V OC but, in addition, depressed the absorption loss in the GaAs bonding layer.By optimizing the III-V top structure, the PCE of a 2-T waferbonded III-V//Si triple-junction SC is boosted from 34.1% in 2020 [55] to 35.9% in 2022 [56].This efficiency enhancement was achieved for two major reasons.The first factor is that the integration of a GaInAsP absorber (Fig. 3e) in the middle subcell raised V OC by 51 mV.The second aspect is that a better current matching of all subcells improved J SC by 0.7 mA/cm 2 .This efficiency is considered the greatest recorded to date for Si-based multijunction SC technologies [56].

Smart Stacking
Smart-stacking technique is a unique bonding technique in which a metal (Pd) nanoparticles (NP) array is used as a bonding mediator between the front and rear subcells (Fig. 4a) to satisfy both electrical conduction and optical transparency at the interface [57][58][59].Mizuno et al. studied the electrical behavior of the Pd NP array-mediated interconnection on a sample that consists of GaAs/Pd/InP and compared the results with those obtained from directly GaAs/InP sample [58].It was confirmed that using Pd nanoparticles reduces the total resistance of the GaAs/InP sample from ~ 100 Ω-cm 2 to 7.5 Ω-cm 2 at 15 mA/ cm 2 .In addition, Mizuno et al. fabricated InGaP/GaAs/Si threejunction cell using a Pd NP array between GaAs and Si to limit the light illumination within the grid electrode of the front cell [59].Such cell produces V OC = 2.81 V, J SC = 10.46 mA/ cm 2 , FF = 79%, and efficiency of 23.2% (1-sun) and 23.74% (8-sun).In the same manner, Sugaya et al. fabricated InGaP/ GaAs/GaAs/Si 4-junction cell with 2-T configuration using the Pd NP array [60].The structure of the such cell (Fig. 4b) demonstrated an efficiency of 18.5% with V OC of 3.3 V which is the first experiment of a four-junction SC with a Si bottom sub-cell.Later in August 2021, Yasushi Shoji et al. developed a GaAs/ InGaAs solar cell by the integration of a low-cost deposition HVPE and low-cost bonding (smart stacking utilizing Pd NP as the bonding mediator) technologies [61] where an efficiency of 22.6% has been achieved.
Recently, Hagar et al. achieved a new low-temperature processing approach to join two standalone different solar cells using Intermetallic Bonding (IMB) [62].The tandem cell structure by the IMB process is shown in Fig. 4c in which two different independent sub-cells are collected and Indium layers are deposited on the metallization grid such that both top and bottom Indium pads are either aligning or crossing each other to minimize the optical shadowing.The I-V curve (Fig. 4d) for bonded and unbonded structures indicates that the IMB approach presents, almost, no extra resistance due to the bonding between the two subcells.That means IMB offers good interconnectivity and no losses due to the bonding [63][64][65].Thus, IMB can be considered a relevant substitute to direct wafer-bonding and tunnel junctions [63].

Perovskite Tandem Solar Cell
According to the development in solar cell efficiency, PSCs have been established faster than any other PV technology in history [66][67][68].PSCs demonstrate certified efficiencies that competing all other thin-film PV technologies, reaching above 25.5% and 32.5% for single-junction and tandem with silicon SCs, respectively [69].These PSCs are based on inorganic-organic metal halide perovskite materials.Inorganic-organic perovskite materials are compounds with ABX 3 chemical formula where A and B are two cations while X is the anion.The A cation contains Formamidinium Iodide (FAI), Methylammonium Iodide (MAI), Cs or their mixture, and Pb, Sn, or a mixture of them as B cation, while I and/or Br as X anion [70].In addition, these materials ensure long carrier lifetimes due to reduced recombination defects which facilitates the charges diffusion toward selective contact layers and to be extracted efficiently [71][72][73].Furthermore, these materials have the advantages of ease of fabrication and several processing techniques [74][75][76].However, PSCs suffer from recombination losses at the interfaces to the charge-selective layers [77][78][79].These losses can be reduced by employing surface modifications, interlayers, or suitable contact materials so that optimal efficiency could be reached [80,81].
In this section, we discuss the status of the PVK/Si tandem SCs.A summary of different structures and electrical characterization techniques.Finally, the current and future challenges of perovskite tandem SCs are summarized.

Perovskite/Si Tandem Solar Cell
Herein, we represent the perovskite tandems based on Si-homojunction and SHJ SCs.The first PVK/Si tandem 1 3 SC was achieved on a Si-homojunction cell [76].Instead, researchers developed perovskite tandems based on SHJ cells.One benefit of the SHJ over Si-homojunction is the higher V OC (0.74 V) which leads to higher efficiency [82,83].However, in homojunction cells, the hole contacts are usually planar because of the challenges to attaining good quality passivation of textured hole contacts [84].Besides, only n-i-p configuration can be used while SHJ provides more flexibility and can be used in either n-i-p or p-i-n configuration [85].

Perovskite/Si-Homojunction Cells
In 2015, the first PVK/Si monolithic tandem SC was constructed on a Si-homojunction cell using a Si-based tunnel junction that interconnects the front PVK and rear Si subcells [83].Such a cell achieved a stable 13.7% efficiency which is limited strongly by the J SC of 11.5 mA/cm 2 .This limited current is due to the thick spiro-OMeTAD Hole Transport Layer (HTL) which led to a parasitic absorption equal to the loss in photo-current density.In 2016, Werner et al. reported 16.3% efficiency cells using Zinc Tin Oxide (ZTO) as a recombination layer, two layers of mesoporous TiO 2 and TiO 2 as Electron Transport Layer (ETL) and CH 3 NH 3 PbI 3 as an absorber layer [86].In 2017, Wu et al. presented advances in designing PVK/Si tandem SCs that include a mesoscopic PVK front subcell and a high-temperature Si-homojunction bottom sub-cell [87].The Si-homojunction top and back surfaces were passivated by Al 2 O 3 /SiN x and SiN x films, respectively.The structure of this design is shown in Fig. 5a.This device improved the V OC and J SC compared to that of [89] which in turn increased the efficiency to 22.5%.
Later in 2018, Zheng J. demonstrated an efficiency of 21.8% on 16 cm 2 with (FAPbI 3 ) 0.83 (MAPbBr 3 ) 0.17 as a top absorber which resulted in a higher V OC [88].In the same  5c) to improve V OC [93].They fabricated the device without the deposition of a recombination layer on a large area.The fabricated device demonstrated an efficiency of 17.3% and a V OC of 1.784 V on 25 cm 2 active area.Table 3 summarizes the progress in the performance of perovskite/Si-homojunction tandem cells.

Perovskite/Silicon-Heterojunction Cells
To evade the degradation of the SHJ SC, the PVK/SHJ tandem should be processed at a low temperature (< 120 ο C) [97].So, the high-temperature mesoporous TiO 2 or a stack of TiO 2 /  mesoporous TiO 2 layer should be substituted by another ETL processed at low temperatures [98].The first 2-T PVK/SHJ tandem cells (Fig. 6a) were introduced in 2015 by Albrecht et al. [99].The perovskite absorber was FAMAPb(I 0.83 Br 0.17 ) 3 and a stack of spiro-OMeTAD/ thermally-evaporated-MoO 3 / sputtered-ITO acts as the front contact.On an area of 0.16 cm 2 , they demonstrated a stabilized efficiency of 18.1% where V OC = 1.79 V, J SC = 14 mA/cm 2 , and FF = 79.5%.Then, Werner et al. showed better efficiency using the same SHJ, with an emitter coated by a sputtered Indium Zinc Oxide (IZO) as a recombination layer [100].The PSC, in n-i-p structure, comprised of (PCPM/PEIE) as an electron contact, MAPbI 3 as an active layer, and an HTL of spiro-OMeTAD.This tandem cell achieved PCEs of 19.2% and 21.2% for cell areas of 1.22 cm 2 and 0.17 cm 2 , respectively.Further, the mechanically-stacking process enabled the fabrication of a 1.43 cm 2 large PVK/SHJ of the monolithic PVK/Si tandem SC used in [112], (e) Cross-sectional SEM image illustrating the full device structure of PVK/Si tandem device [113], (f) Schematic architecture of CsPbIxBr3-x PSCs [116], (g) Schematic device structure of PVK/Si SCs used in [117] tandem cell, utilizing a back-side textured rear subcell to widen its near-infrared response and thus enhance the current density of the bottom sub-cell.Such cells recorded 20.5% efficiency [101].In 2017, Fan et al. employed a solution process to fabricate the tunneling junction and the PVK absorber in the 2-T monolithic PVK/SHJ tandem SC [102].By engineering the bandgap of the PVK subcell from 1.55 to 1.69 eV, they achieved an improved current matching between the two subcells and monitored the change in V OC of the tandem cell.Thus, the efficiency was improved to 23 To achieve better performance for PVK/Si tandems than those aforementioned cells, researchers adopted the inverted configuration of the perovskite subcell [38,105].Bush et al. realized, for the first time, a 23.6% efficient p-i-n perovskite/textured SHJ (Fig. 6b) with improved stability [106].The parasitic absorption in the window and absorber layers was decreased resulting in J SC = 18.1 mA/ cm 2 , but V OC is only 1.65 V.In 2018, Bush et al. presented improvements to their work to optimize light harvesting (Fig. 6c) [107].In the same year, Sahli et al. reported a certified 25.5% efficient cell on an active area of 1.42 cm 2 [108].The key to their improvement lies on texturing the front and rear sides of the SHJ which manages to achieve a high J SC of 19.1 mA/cm 2 .
Further, in 2019, a 25.4% efficient 2-T PVK/SHJ SC was reported [69].The wide bandgap triple cation (Cs 0.15 (FA 0.83 MA 0.17 ) 0.85 Pb(I 0.75 Br 0.25 ) 3 (1.64 eV) was used as a perovskite absorber to match the photo-current between the subcells in a PVK/Si tandem SC which resulted in a high V OC of 1.80 V. Furthermore, Kohnen et al. executed a transparent n-type front contact layer stack in monolithic PVK/Si tandem SCs.The results showed an efficiency of 26% [109].By achieving high quality PV films, a 2-T PVK/ Si tandem solar cell was manufactured with an n-i-p perovskite front subcell.The resulting efficiency of the tandem SC was 22.80% [110].McGehee's research group reported an efficient 1.67 eV wide-bandgap PVK top sub-cell using triple-halide alloys of chlorine, bromine, and iodine to stabilize the semiconductor under irradiance [111].They fabricated a 2-T PVK/Si tandem with an area of 1.0 cm 2 that demonstrated a stabilized PCE of 27.04%.In 2021, Kohnen et al. established a 2-T PVK/Si tandem SC based on a 100 µm thin industrially Czochralski (CZ) silicon wafer (Fig. 6d) [112].
Most of the recent studies focus on the cation component, Kim et al. developed a stable PSC by anion engineering a wide-bandgap (1.68 eV) perovskite (Fig. 6e) [113].Using this approach, they extended the light stability as well as achieved a performance improvement that a PCE of 26.7% had been recorded.Hou et al. represented a PVK/ Si tandem SC that combined solution-processed µm-thick PVK front subcell with fully textured SHJ back cells [114].A PCE of 25.7% was achieved.Oxford PV, on the other hand, has shattered the global record with a newly certified efficiency of 29.5% at the end of 2020, but no specifics about the device structure or photovoltaic characteristics have been released yet [69].He et al. demonstrated the importance of adding methylammonium chloride (MACl) in bandgap tuning, defect passivation, and crystal grain growth to realize a high V OC [115].An efficiency of 25.9% for PVK/Si heterojunction tandem SCs had been attained with a V OC of 1.74 V.
In 2022, Wang Sanlong et al. provided an encouraging strategy for utilizing inorganic passivation materials to reach high efficiency and stability (Fig. 6f) [116].They demonstrated that NiI 2 treatment can considerably improve V OC and enhance the stability of inorganic CsPbI 3-x Br x single junction PSCs.Accordingly, an efficiency of 19.53% and a V OC of 1.36 V were reported.Such efficient single junction inorganic perovskite SC contributes with an SHJ bottom cell in a high-efficiency PVK/Si tandem SC.They recorded an efficiency of 22.95% and a high V OC of 2.04 V.Moreover, Jiafan Zhang et al. successfully constructed a monolithic PVK/Si tandem SC based on the industrial textured Si heterojunction SCs with front transparent conductive electrodes of 60 nm and 110 nm thick IZO thin films (Fig. 6g) [117].In July 2022, Researchers from the Swiss Center for Electronics and Microtechnology (CSEM) and the École polytechnique fédérale de Lausanne (EPFL) reported a PCE that exceeds 30% for a 1 cm 2 PVK/Si tandem SC.This pathbreaking milestone was individually certified by National Renewable Energy Laboratory.Particularly, an efficiency of 30.93% for a 1 cm 2 was obtained.The used solar cells were based on high-quality PVK films on a planar Si surface.Besides, an efficiency of 31.25% was reported for a cell of the same size but fabricated with a hybrid vapor/ solution processing technique [118].Most recently, a group of researchers from Helmholtz-Zentrum Berlin (HZB) has achieved a new world efficiency record for a PVK/Si tandem solar cell, with a certified efficiency of 32.5% [25,119].Table 4 lists the aforementioned development in the performance of the perovskite/SHJ tandem cells along with the utilized recombination layer.

Current and Future Challenges
From the above review of the perovskite solar cell technologies, we can summarize their current and future challenges in some points, the structure of the tandem cell, stability, large-scale fabrication, efficiency, and toxicity.

Structure of Tandem Cell
In 2-T PVK/Si, as the sub-cells are grown epitaxially, the tandem cell will be damaged if there is any failure in one subcell or recombination layer.Good current matching between all subcells as well as low energy loss are mandatory conditions.Therefore, the bandgap of both perovskites and the recombination layer must be selected carefully in order to reduce energy loss and well prevent ion migration.Also, the recombination layer should be well optimized to protect the underlying bottom sub-cell from damage as the top sub-cell is grown by solution deposition.Regarding PVK/Si technology, silicon surfaces are sensitive to impurity densities and might be contaminated by halide ions [106].

Stability
Achieving long-term stability is a major target for perovskite tandem solar cell development.Currently, most stability studies are performed on small-area cells.It is challenging and critical to realize the deterioration associated with different factors such as moisture, heat, or Oxygen.To enhance the stability of perovskite-based tandem SCs, highquality perovskites, i.e., lower defects and grain boundaries, must be prepared [125] by reducing oxygen and water corrosion.As the top perovskite material are prepared by inorganic-organic halide, the halide separation may be a crucial issue affecting the device stability To overcome the halide separation and hence improve long-term stability, some studies reported increasing the crystallinity of perovskites and passivating grain boundaries [127].In addition, it is reported that the stability is affected by dopant HTL which can destroy the perovskite tandem cell [128].So, developing dopant-free HTL with high mobilities may become a research trend.Besides, the stability of PVK tandems could also be expanded by compositional engineering of the thin layers [129,130].Furthermore, Encapsulation methods applying glass sealing or laminate plastic layers to enhance device stability for over 125 days at 60°C were demonstrated [131].In general, most active materials in perovskite solar cells are susceptible to environmental factors such as oxygen and moisture, Ultraviolet radiation, and heat.Thus, the active layers must be protected from these phenomena through appropriate encapsulation [132].That it is, the encapsulation layer acts as a barrier layer by restricting the diffusion of moisture or oxygen through the encapsulation material.Once the stability of perovskite solar cells and encapsulation issues could be resolved successfully, the perovskite-based tandem solar cell would be willing for commercialization [106].

Large-Scale Fabrication
Scaling up the active area of a perovskite tandem cell is important for the industrialization and commercialization of such cells.The dominant high-efficient perovskite tandems are still realized with an active area of smaller than 1 cm 2 .
To fabricate highly efficient large-scale perovskite tandem SCs, it is necessary to build up high-quality perovskites with a large area.Challenges in upscaling the perovskite solar cell involve developing scalable printing techniques for all cell layers such as ETL, absorber, and HTL as well as a reliable module design to obtain high quality films and excellent device properties [133,134].Solution processing represents a major advantage, and slot-die and ambient spray coating have emerged as scalable processes.Therefore, the selection of compatible materials and processes is important for PSC commercialization.In recent years, significant progress has been achieved in the quality of perovskite films as well as the efficiency of large-area Perovskite tandems [135][136][137][138][139][140].Higuchi et al. developed a large-scale (203 mm × 203 mm) PVK Module (35 series cells) with an efficiency of 12.6% [140].Up to today, the largest area for such perovskite module (55 series cells) is recorded to be 30 cm × 30 cm with aperture area of 804 cm 2 .The efficiency of that PVK cell is 17.9% [4].Further, upscaling the area of PVK/Si tandem cells is discussed by Kamino et al. using low temperature screen-printed metallization to achieve a steady-state efficiency of 22.6% over an aperture area of 57.4 cm 2 [139].
With the ongoing efforts from the research groups and industrial communities toward upscaling the PVK cells, it is expected that the efficiency gap between lab-scale cells and industrial modules will be narrowed and will achieve a level comparable to that of other PV technologies.

Efficiency
The effective method to enhance the performance of the PVK/ Si cell is to improve the performance of the perovskite cell itself.Currently, the champion single-junction perovskite cell achieves 25.5% efficiency based on 1.5 eV bandgap material and is certified by NREL [69].However, Korean researchers attained 25.8% efficiency for single-junction PSCs due to an interlayer placed between the ETL and the absorber layer, which excluded the requirement for passivation [138].Most studies of single-junction perovskite solar cells rely on this bandgap rather than any other bandgap, i.e., either wide-bandgap or narrow-bandgap.These wide-bandgap and narrow-bandgap are essential to develop perovskite tandem SCs with excellent performance.Yet, wide-bandgap perovskite suffers from enormous energy loss due to nonradiative recombination, as well as the narrow-bandgap Sn-Pb based perovskite solar cell, suffers from some stability issues which are much more challenging.

Toxicity
All high-efficiency perovskite solar cells may contain toxic Lead (Pb).Although the amount of Pb is small, it presents a challenge for perovskite solar cells to get commercialized.A lot of research groups are currently investigating lead-free perovskites with Sn and other components [141][142][143].Among the different elements, Tin-based perovskite solar cell exhibits the highest efficiency.Although much research has been accomplished to increase the tandem cell performance, the efficiency is much smaller than the efficiency of Pb-based PVK tandem SCs because of the oxidation of Tin.So, an encapsulation to Pb-based perovskite tandem cells can be a solution to prevent Pb leakage.In addition, more works regarding the advancement of highly efficient Sn-based PVK tandem SCs are important via suppressing oxidation and controlling crystallization.

Comparison Between Different Solar Cell Technologies
As aforementioned, to increase solar cell efficiency, tandem structures are utilized.We reviewed two different solar cell technologies, III-V/Si and PVK/Si.In this section, we provide a concise comparison between these different tandem solar cell technologies.While the comparison provides a current perspective, we also explore the future potential for the next few upcoming years.

Comparison Criteria
The criteria for the comparison are defined such that to achieve the lowest levelized cost of electricity (LCOE) as well as the adjustability of electric parameters (i.e., voltage and current).The criteria that influence the LCOE are efficiency, cost, and long-term stability.Herein, a brief comparison between the aforementioned reviewed technologies is introduced.Eventually, any tandem solar cell has a more cost than a Si single-junction solar cell, which has a tremendous market share solar cell (> 90% of market share) [144], thus efficiencies higher than the efficiency of Si single-junction cell (25.2%) are needed to validate the tandem approach.
The tandem SCs that have encountered this metric are III-V solar cells, III-V/Si, PVK/PVK, and PVK/Si.Therefore, the comparison is essentially performed between these different four technologies.

Efficiency
Historically, tandem solar cells have been used in the application of space and concentrator SCs.However, the progress of new materials makes them compete with single-junction cells [145].The III-V tandem solar cells provide the best-ever efficiencies of up to 39.2% under 1-sun concentration (6 junctions) and 47.1% under concentrated irradiation (6 junctions).
On the other hand, the highest certified efficiency recorded for all-perovskite (PVK/PVK) tandem SCs is 28% to date [4].However, there is still space for efficiency advancement for PVK/PVK tandems.Tandem SCs using Si as a bottom cell, having high efficiencies, have been reviewed as III-V/Si and PVK/Si SCs.The III-V/Si SCs suffer from a degradation in efficiency comparable with III-V tandem SCs because of the current mismatching between the III-V and Si materials.The efficiency reached 32.8% and 35.9% for two-junction GaAs/ Si and 3-junction GaInP/GaAs//Si solar cells, respectively.Instead, PVKs have the benefit of much lower cost, and rapid improvements have been accomplished [146,147].The highest attainable lab record performance of perovskite/Si solar cells is 32.5% by HZB in just a few years of development.Figure 7a illustrates the progress in the efficiency of perovskite/ Si for both homojunction and heterojunction silicon bottom cells.and III-V/Si tandem SCs.It is clear that the heterojunction silicon contributes more efficiency than the homojunction silicon cell.Further, Fig. 7b shows an efficiency map for different Si-based III-V tandem SCs.Direct growth, mechanical stacking, wafer bonding, and smart stacking are four different technologies utilized to realize two-junction or three-junction tandem solar cells.From Fig. 7b, mechanical stacking and wafer bonding technologies result in higher efficiencies than that in other technologies.However, new research studies show a great improvement in the efficiency of the twojunction and three-junction solar cells fabricated using direct growth technology.Furthermore, smart stacking technology is a promising technology that records a considerably high efficiency.In addition, high-efficiency potential of more than 42% with 3 − junction Si-based tandem solar cells (III-V/Si 3-junction and PVK/Si 3-junction) has been addressed in the literature [2,148,149].

Cost
One of the most significant criteria to lower LCOE is the cost.Although III-V tandem solar cells achieve the best performance among all solar cell technologies, this significant performance comes at a cost.The cost per Watt Peak ($/Wp) of a III-V tandem solar cell is 1-2 orders of magnitude higher than Si [146].This high cost lies in the substrate for epitaxial growth, deposition, and back-end processing.If these needs are reduced, the total cost is to be lowered.The substrate cost could be decreased by replacing the Ge or GaAs substrate with the Si substrate.Regarding semiconductor deposition, recent research has significantly boosted growth rates and shows much employing of both MOVPE and HVPE [146].
That is why the cost of a Si-based III-V tandem solar cell is lower than that of an III-V tandem SC.On the other side, most highly efficient perovskite tandem cells are manufactured by solution processing which makes them desirable as a low-cost solar cell technology.The estimated perovskite solar cell cost is one-third of the cost of the Si solar cell [106].This cost can be further reduced by developing low-cost HTL material to replace the expensive PTAA and spiro-OMeTAD.Therefore, compared with the high production cost of multi-junction SCs made of III-V semiconductors, PVK/PVK, and PVK/Si multijunction SCs offer an efficient and low-cost choice for the terrestrial application.Song et al. proposed a feasible manufacturing flow for single-junction perovskite PV fabrication [150].They estimated the LCOE to be 5.82 c$/kWh in Wichita, Kansas, if a system lifetime of 30 years can be achieved.Later, Li et al. performed cost analysis for different PV modules.They concluded that the single-junction perovskite solar cells are promising to achieve a competitive LCOE in the PV market.Meanwhile, perovskite-based tandems are capable of lowering the LCOE effectively due to the extremely low manufacturing cost and high efficiency of perovskite sub-cells [151].

Long-Term Stability
The long-term stability considers how fast degradation occurs under regular conditions such as heat, moisture, oxygen, and UV radiation.III-V tandem SCs are very stable but again come to cost.Also, Si-based III-V tandem SC is one of the PV technologies that have a reliability track record.The long-term stability of PVK/Si is comparably low since moisture has a deteriorating effect on cell performance.Therefore, additional efforts have to be performed in order to prevent this degradation.The stability needs for PVK/Si tandem SCs have been studied thoroughly and it is observed that the maximum allowable PVK front cell degradation rates are around 1-2% per year [146].PSCs have shorter life ranging from a few days to months and a maximum of up to a year [152].So, an important area of focus now is to develop and test a long-life perovskite tandem cell.In conclusion, Table 5 provides a concise comparison between the main two tandem solar cell technologies, III-V/Si and PVK/ Si, along with III-V tandems and all-perovskite tandems.

Future Potential
Tandem technology can be utilized directly to Si.Current advancements in Si-based tandem solar cells are reported.Option for low-resistance interconnection between the subcells includes tunnel junctions and TCO as a recombination layer.For some materials (III-V) of deposited top sub-cell directly on Si, advances are required in heteroepitaxy and nucleation to improve the performance, reduce manufacturing cost, and avoid Si damage.So, additional work is needed to merge these constraints.Also, innovations in top cell materials, deposition techniques, and designs are required.Regarding the III-V top sub-cell, cost reduction is the crucial factor but as discussed previously to achieve the cost target, advances in deposition techniques, substrate, and metallization are required.For perovskite top sub-cell, rises in efficiency and long-term stability alongside cost reduction are needed.The most significant factor in lowering the processing cost is to replace expensive materials such as PTAA, spiro-OMeTAD, and gold.Some promising approaches can achieve higher efficiency and stability simultaneously.These approaches may include inorganic transport layers and new perovskite formulations.
The future potential for the upcoming few years is much more open and difficult to access.Recently, perovskite-based tandem solar cell technologies demonstrated tremendous development toward higher efficiencies and low cost.Therefore, these technologies are actively developed by the research community.The Si-based tandem solar cell could replace the Si technology for the PV market but there are still further developments necessary to reach this point and the Si-based tandems are still improving.

Conclusion
In this review, the present and future status of different tandem solar cell technologies: III-V/Si tandems and PVK/Si are presented along with the scientific and engineering challenges for each.A comparison between these different technologies is summarized based on important criteria such as efficiency, cost, and long-term stability.III-V tandem SCs introduce the highest efficiency and reliability among all other technologies, but it comes at a cost.To lower the cost, the III-V substrate is replaced with a cheaper Si substrate.Several ways to combine III-V material with Si substrate are discussed.For potential high-performance and low-cost III-V/Si tandems, careful analysis of the design challenges is crucial.
In addition, perovskite/Si tandem solar cell is introduced alongside the present status for each.As well, five major challenges of perovskite tandem cells, cell structure, efficiency, long-term stability, large-scale fabrication, and toxicity, are highlighted.PVK/Si tandem cells gained the greatest interest among these distinct perovskite tandem SCs thanks to their high performance and advanced manufacturing technology.Additionally, due to their exceptional performance and low cost, perovskite-based tandem SCs have the potential to be the current pioneer of all other solar cell technologies.So, these devices attract more and more interest as the research activity for such technology have been increased in the last five years.Rapid advancement in perovskite solar cells to overcome critical challenges is monitored and expected to be continuously increased in the near future.

Fig. 2 a
Fig. 2 a Schematic layer structure of the two-junction tandem GaInP/Si SC, Reprinted from [15], b GaAsP/Si two-junction cell structure.Reprinted from[47], and c Schematic diagram of GaAsP/Si tandem SC[49]

Representative 2 -
T PVK/Si-homojunction tandem SC based on a p-i-n structured perovskite front subcell is as follows.Kanda et al. fabricated individual PVK and Si solar cells and connected each other by depositing an ITO film on evaporated Au on the back surface of the PSC and directly on top of the c-Si SC [94].For this design, an efficiency of 13.7% was reported.The prospect of commercializing n-type Si SCs utilized in tandem SCs that are based on Si, according to Hoye et al., accounts for just 5% of the worldwide SC industry [95].They built a 2-T PVK/Si tandem SC utilizing a p-type Si-based rear subcell and a p-i-n structured PVK front subcell, with a recorded PCE of 16.2%.Further, Kanda et al. fabricated a facile processed small-sized single c-Si solar cell (25 mm 2 ) by a non-vacuum process as an energy supply and processed for a 2-T PVK/Si tandem SC [96].They presented a conversion efficiency of 15.5% utilizing a new buffer layer of MoO x .

Fig. 7
Fig. 7 Efficiency Record Map for (a) Perovskite/Si and (b)Sibased III-V tandem solar cells

Table 1
Advances in AlGaAs/Si tandem solar cells

Table 2
Summary of the performance of some GaInP/Si tandem solar cell

Table 4
Summary of the reported status for PVK/SHJ tandem SCs

Table 5
Comparison of the different tandem solar cell technologies regarding the defined criteria of this section.A plus sign ( +) denotes a comparable positive characteristic for the respective criteria while a Minus sign (-) denotes a comparable lesser fulfillment of the respective criteria.(U.D) denotes a very strong development in this category