Inclusion of 2D Transition Metal Dichalcogenides in Perovskite Inks and Their Influence on Solar Cell Performance

Organic–inorganic hybrid perovskite materials have raised great interest in recent years due to their excellent optoelectronic properties, which promise stunning improvements in photovoltaic technologies. Moreover, two-dimensional layered materials such as graphene, its derivatives, and transition metal dichalcogenides have been extensively investigated for a wide range of electronic and optoelectronic applications and have recently shown a synergistic effect in combination with hybrid perovskite materials. Here, we report on the inclusion of liquid-phase exfoliated molybdenum disulfide nanosheets into different perovskite precursor solutions, exploring their influence on final device performance. We compared the effect of such additives upon the growth of diverse perovskites, namely CH3NH3PbI3 (MAPbI3) and triple-cation with mixed halides Csx (MA0.17FA0.83)(1−x)Pb (I0.83Br0.17)3 perovskite. We show how for the referential MAPbI3 materials the addition of the MoS2 additive leads to the formation of larger, highly crystalline grains, which result in a remarkable 15% relative improvement in power conversion efficiency. On the other hand, for the mixed cation–halide perovskite no improvements were observed, confirming that the nucleation process for the two materials is differently influenced by the presence of MoS2.


Introduction
Extensive, ongoing research has been performed to date towards the development of highly efficient and stable perovskite solar cells (PSC) [1]. Organic-inorganic metal halide perovskites, with the general formula ABX 3 (where A is a cation, B is a divalent metal cation and X is a halide), are a class of semiconductors that have the potential to deliver high-efficiency photovoltaic devices at mild temperature processing conditions and low cost [2]. These materials feature unique optical and electronic properties, such as high carrier mobility, long charge diffusion length, low trap-state densities and intense broadband absorption [3].
Perovskite crystalline structure, morphology, dimension, and distribution of the poly crystallite perovskite thin films are critical factors that affect the optoelectronic and photovoltaic properties [4]. Currently, one of the main challenges for PSC development is the deposition of high-quality films with optimized morphology, large grains and which are pinhole-free. Many different approaches have been investigated to improve the film quality: annealing conditions [5], compositional [6] and solvent engineering [7], interfacial modification [8], deposition methods [9], and, finally, additive engineering [10][11][12][13]. The inclusion of additives into the perovskite precursor solution is widely adopted, as their presence impacts the final morphology of perovskite films [14], stabilizes the active crystalline phase [15], tunes the energy level alignment between material constituents, and suppresses non-radiative recombination in perovskite materials [16]. Furthermore, additives play an important role in perovskite crystal growth and can lead to a stronger control over nucleation and crystallization kinetics [11], or activate surface passivation mechanisms, achieving highly crystalline perovskite films that are suitable for device integration. Several types of additives have been added into perovskite precursor solutions, including polymers [13], molecules [17], organic [18,19] and inorganic [20] halide salts, inorganic acids [21], carbon-based materials and nanoparticles [22]. Within this collection, interesting perspectives involve the use of systems characterized by extended surfaces, such as two-dimensional (2D) materials, which could extend the interaction over a region of large perovskite grains [23][24][25].
Two-dimensional transition metal dichalcogenides (2D-TMD), obtained by liquidphase exfoliation, are characterized by high electrical conductivity and charge carrier mobility, tunable optoelectronic properties, and offer the possibility of engineering their surfaces by physical or chemical methods. Therefore, with PSC advancement in mind, 2D-TMDs offer the possibility of integration into multiple features: as electrodes, as charge transporting layers, and/or as an additive to be used in the precursor solutions of perovskite active layers [26].
Two-dimensional flakes of MoS 2 , MoSe 2 and TiS 2 have recently been used as an interlayer between perovskite and either the hole transporting layer (HTL) or electron transporting layer (ETL) [27][28][29]. This prevents the formation of shunt contacts between the perovskite and the electrode and provides a more suitable energy band alignment between the perovskite active layer and the transporting layer. Finally, chemically exfoliated MoS 2 was employed both as an additional component of the perovskite precursor solution and an interfacial layer to enhance PSC [30] efficiency in the p-i-n architecture with PEDOT:PSS as the HTL [30]. The inclusion of MoS 2 enlarged the grain size and improved the crystalline quality of MAPbI 3 perovskite films. Furthermore, due to the incorporation of an MoS 2 interlayer, the direct contact between hydrophilic PEDOT:PSS and MAPbI 3 film is avoided, preventing early deterioration of device performance.
In this work, we focused on the inclusion of MoS 2 TMD sheets into a perovskite precursor solution of two representative benchmark materials, MAPbI 3 and mixed cationhalide CsMAFAPbIBr (MA = methylammonium, FA = formamidinium), with the aim of improving our knowledge of the role of such an additive in diverse perovskite formulations. The solar cell architecture was an inverted p-i-n structure employing organic transporting layers as sketched in Figure 1a,b. Our findings show that the integration of MoS 2 into the MAPbI 3 perovskite active layer leads to larger grain size, resulting in improved device performance. In fact, the PSCs show a 15% improvement in power conversion efficiency (PCE) through embedding TMDs in the perovskite active layer. Conversely, identical additive engineering for CsMAFAPbIBr perovskite did not show similar device improvements, suggesting that the polycrystalline perovskite film formation is influenced differently by the presence of 2D MoS 2 depending on the different ions involved in solution. Nanomaterials 2021, 11, x FOR PEER REVIEW 3 of 16

Preparation of 2D MoS2 Nanosheets
2 g of MoS2 powder were added to 40 mL of N-methyl-2-pyrrolidone (NMP), and a probe sonic tip was used to sonicate the solution for a certain number of hours (power: 60% amplitude). The sonic tip was pulsed for 6 s on and 2 s off to avoid damage to the processor and reduce solvent heating, and thus, degradation. The beaker was connected to a cooling system that allowed for cold water (under 5 °C) to flow around the dispersion during sonication. Then, exfoliated MoS2 nanosheets in NMP solvent were centrifuged at high spin rate (10,000 rpm) to obtain the MoS2 nanoflakes. Then, these MoS2 nanosheets were redispersed in DMF through bath sonication in a concentration of 0.08 mg mL −1 . This solution was then used for the different additions to perovskite precursor solutions.

Preparation of 2D MoS 2 Nanosheets
2 g of MoS 2 powder were added to 40 mL of N-methyl-2-pyrrolidone (NMP), and a probe sonic tip was used to sonicate the solution for a certain number of hours (power: 60% amplitude). The sonic tip was pulsed for 6 s on and 2 s off to avoid damage to the processor and reduce solvent heating, and thus, degradation. The beaker was connected to a cooling system that allowed for cold water (under 5 • C) to flow around the dispersion during sonication. Then, exfoliated MoS 2 nanosheets in NMP solvent were centrifuged at high spin rate (10,000 rpm) to obtain the MoS 2 nanoflakes. Then, these MoS 2 nanosheets were redispersed in DMF through bath sonication in a concentration of 0.08 mg mL −1 . This solution was then used for the different additions to perovskite precursor solutions.

Photovoltaic Device Fabrication
Perovskite solar cell architectures are shown in Figure 1a,b. ITO-coated glass substrates were cleaned by ultrasonication in a deionized water, acetone, 2-propanol. Polytriarylamine (PTAA) layer (1.5 mg mL −1 in toluene) was deposited by spin coating at 6000 rpm for 30 s and annealed at 100 • C for 10 min. The rest of the processes were performed in an N 2 -filled glovebox. The MAPbI 3 perovskite precursor solution was spin coated onto PTAA coated ITO substrates at 4000 rpm for 25 s; for solvent dripping, 200 µL of toluene was dropped onto the film 15 s prior to the end. The CsMAFAPbIBr perovskite precursor solution was spin coated onto PTAA coated ITO substrates by a consecutive two-step spin coating process at 1000 rpm and 6000 rpm for 10 s and 20 s, respectively; for solvent dripping, 200 µL of chlorobenzene was dropped onto the film 5 s prior to the end. The perovskite film was then annealed for 10 min at 100 • C and then allowed to cool down to room temperature. Then, a PCBM solution in chlorobenzene (25 mg mL −1 ) was spin coated onto the perovskite layer at 1000 rpm for 60 s. Finally, a BCP solution in 2-propanol (0.5 mg mL −1 ) was spin coated at 6000 rpm for 20 s. Solar cell devices are completed by thermal evaporation of 100 nm Al electrodes.

Materials Characterization
The scanning electron microscopy (SEM) imaging was performed by a MERLIN Zeiss SEM FEG instrument at an accelerating voltage of 5 kV, using an in-lens detector. The particle size distribution was estimated using the open-source ImageJ software, by measuring the major axis of 100 perovskite grains for each sample. HRTEM images were acquired on an FEI Titan (Thermo Fisher Scientific Inc., Waltham, MA, USA) operating at an acceleration voltage of 300 keV. The EDX spectrum of the MoS 2 dispersion was acquired on Jeol2100 (Jeol Ltd., Akishima, Tokyo, Japan) operating at 200 keV and equipped with an Oxford Instruments 80mm 2 silicon drift detector with a 10 • holder tilt. Ultraviolet-visible (UV-Vis) absorption spectra were measured on PerkinElmer (Lambda 1050, Waltham, MA, USA) spectrophotometer in the 300-800 nm wavelength range at room temperature. Steadystate and time-resolved photoluminescence (PL) was performed using an Edinburgh FLS920 spectrometer (Edinburgh Instruments, Scotland, UK) equipped with a Peltiercooled Hamamatsu R928 photomultiplier tube (185-850 nm). An Edinburgh Xe900 450 W Xenon arc lamp (Edinburgh Instruments, Scotland, UK) was employed as the exciting light source. Emission lifetimes were determined using the single photon counting technique by means of the same Edinburgh FLS980 spectrometer (Edinburgh Instruments, Scotland, UK) using a laser diode as the excitation source and a Hamamatsu MCP R3809U-50 as detector.

Photovoltaic Device Characterization
The devices were characterized in an N 2 atmosphere by using a Keithley 2400 Source Measure Unit (Tektronix, Berkshire, UK) and AirMass 1.5 Global (AM1.5G) solar simulator (Newport 91160A, Irvine, CA, USA) under an irradiation intensity of 100 mW cm −2 . Current-voltage characteristics were acquired at voltages ranging from 1.2V to −0.2 V. The step voltage is fixed at 10 mV and the delay time to 100 ms.

Properties and Characterizations of Materials
The MoS 2 nanosheets used in this work as an additive in hybrid halide perovskite solutions were prepared by liquid-phase exfoliation (see experimental section). Firstly, the 2D material was characterized in order to confirm an effective exfoliation. The scanning electron microscopy (SEM, Figure 2a) image shows few-layered MoS 2 nanoflakes. Highresolution TEM (HRTEM, Figure 2b) and its corresponding fast Fourier transform (FFT, inset of Figure 2b) pattern reveal the highly crystalline structure of the as exfoliated MoS 2 material. Moreover, the exfoliated flakes reveal well-defined edges and no apparent damage in the basal planes, suggesting that the obtained material is of high quality.  Figure 2b) and its corresponding fast Fourier transform (FFT, inset of Figure 2b) pattern reveal the highly crystalline structure of the as exfoliated MoS2 material. Moreover, the exfoliated flakes reveal well-defined edges and no apparent damage in the basal planes, suggesting that the obtained material is of high quality. In the EDX spectrum (Figure 2c), the molybdenum (Mo) and sulphur (S) peaks are clearly present confirming the identity of the nanosheets. The carbon (C) peak is also clearly identifiable and attributable to both the lacey carbon grid, as well as to possible residual DMF solvent. Additional peaks can be attributed to the background scattered signal (iron (Fe) and chromium (Cr)) from the TEM polepieces. Copper (Cu) peaks are present due to the copper TEM grid. The silicon (Si) peak is due to an EDX spectral artefact, being most probably attributable to an internal fluorescence peak from the silicon window on the drift detector.
The as-obtained MoS2 sheets were dispersed in N,N-dimethylformamide and employed as an additive for MAPbI3 perovskite precursor solution at different volume ratios: 5, 10, 20 (%, V/V). A preliminary check on the compatibility between MoS2 TMD sheets and the reaction environment in which the nucleation and growth processes of the perovskite material take place confirmed the possibility of using this material as an additive for perovskite precursor solutions. In this environment, reasonably good solubility of the TMD is observed, and no precipitation or phase separation occur during the spin coating process. This high solubility can be explained by the coordinated equilibrium of lead cations and iodide ions present in the perovskite precursor solution along with MoS2 [31]. Indeed, the S nonbonding lone-pair orbital can function as an electron donor (Lewis base), interacting with lead cations (Lewis acids) in solution, while halide ions can strongly coordinate molybdenum [32]. The resulting films are spin coated on glass/ITO/polytriarylamine (PTAA), which is the hole transporting material employed in the solar cell architecture. These films are smooth, particularly homogeneous and pinhole-free (insets of Figure  3). The thickness of the perovskite films was found to slightly increase upon the addition of MoS2 sheets (Table 1). Small differences among the three samples are within experimental error. In the EDX spectrum (Figure 2c), the molybdenum (Mo) and sulphur (S) peaks are clearly present confirming the identity of the nanosheets. The carbon (C) peak is also clearly identifiable and attributable to both the lacey carbon grid, as well as to possible residual DMF solvent. Additional peaks can be attributed to the background scattered signal (iron (Fe) and chromium (Cr)) from the TEM polepieces. Copper (Cu) peaks are present due to the copper TEM grid. The silicon (Si) peak is due to an EDX spectral artefact, being most probably attributable to an internal fluorescence peak from the silicon window on the drift detector.
The as-obtained MoS 2 sheets were dispersed in N,N-dimethylformamide and employed as an additive for MAPbI 3 perovskite precursor solution at different volume ratios: 5, 10, 20 (%, V/V). A preliminary check on the compatibility between MoS 2 TMD sheets and the reaction environment in which the nucleation and growth processes of the perovskite material take place confirmed the possibility of using this material as an additive for perovskite precursor solutions. In this environment, reasonably good solubility of the TMD is observed, and no precipitation or phase separation occur during the spin coating process. This high solubility can be explained by the coordinated equilibrium of lead cations and iodide ions present in the perovskite precursor solution along with MoS 2 [31]. Indeed, the S nonbonding lone-pair orbital can function as an electron donor (Lewis base), interacting with lead cations (Lewis acids) in solution, while halide ions can strongly coordinate molybdenum [32]. The resulting films are spin coated on glass/ITO/polytriarylamine (PTAA), which is the hole transporting material employed in the solar cell architecture. These films are smooth, particularly homogeneous and pinhole-free (insets of Figure 3). The thickness of the perovskite films was found to slightly increase upon the addition of MoS 2 sheets (Table 1). Small differences among the three samples are within experimental error.
SEM images (Figure 3b,c) show the MoS 2 additive's influence on the MAPbI 3 perovskite grain size: the addition of the additive into the perovskite precursor solutions, at low-intermediate concentrations (5% and 10%), results in the formation of larger grains. The particle size analysis of these images ( Figure 4, Table 2) reveals a significant increase indeed in the average grain size, from 116 nm for the pristine sample to 177 nm and 187 nm for the MoS 2 (5%) and MoS 2 (10%) perovskite samples, respectively. Moreover, the 10% MoS 2 additive sample has the narrowest grain size distribution, with more than 50% of particles located in a small range from 160 nm to 220 nm. The implication of this observation is a reduced extension of the grain boundary region. This prefigures a reduced charge loss, typically occurring at the grain boundary region, giving rise to an improvement in final device performance.   Table 2) reveals a significant increase indeed in the average grain size, from 116 nm for the pristine sample to 177 nm and 187 nm for the MoS2 (5%) and MoS2 (10%) perovskite samples, respectively. Moreover, the 10% MoS2 additive sample has the narrowest grain size distribution, with more than 50% of particles located in a small range from 160 nm to 220 nm. The implication of this observation is a reduced extension of the grain boundary region. This prefigures a reduced charge loss, typically occurring at the grain boundary region, giving rise to an improvement in final device performance.  There are several factors that can affect the perovskite's morphology, including the intrinsic properties of the material itself, the deposition method, the presence of impurities, the surface energy of the substrate, and the application of post-treatments [31]. The MoS 2 additive seems to be able to regulate film morphology by acting upon the crystal growth and by altering the colloid distribution in the perovskite precursors [31]. This results in high-quality, pinhole-free perovskite films with larger grain size and filled grain boundaries [33,34]. Since the crystal growth rate is relatively fast and is a function of solution supersaturation, reaching a high nucleation rate before the onset of crystal growth is required to improve perovskite film coverage. The heterogeneous nucleation mechanism in addition provides fewer nucleation sites, when compared to homogeneous nucleation, leading to the formation of larger crystalline domains [34,35]. This interesting observation is extended here to the deposition of perovskite onto organic PTAA substrates, showing that TMDs can also affect this kind of device, which would be ideally suited for flexible, light, and portable PSCs. Noticeably, at 20% MoS 2 additive concentration, the effect on the grain size is lost as these films exhibit similar grain sizes and particle size distribution to pristine perovskite ( Figure 4, Table 2). The effect of such high MoS 2 concentration is ascribed to an increased number of heterogeneous nucleation sites that would impair the formation of large grains during the perovskite crystal growth. This is suggested by the SEM image in Figure 3d. A morphology characterized by smaller grains, such as the one recorded for the 20% MoS 2 additive perovskite film, is in general associated with poor charge transport and collection in PSCs, which are limited by inter-grain boundary recombination losses [36].  There are several factors that can affect the perovskite's morphology, including the intrinsic properties of the material itself, the deposition method, the presence of impurities, the surface energy of the substrate, and the application of post-treatments [31]. The MoS2 additive seems to be able to regulate film morphology by acting upon the crystal growth and by altering the colloid distribution in the perovskite precursors [31]. This results in high-quality, pinhole-free perovskite films with larger grain size and filled grain boundaries [33,34]. Since the crystal growth rate is relatively fast and is a function of solution supersaturation, reaching a high nucleation rate before the onset of crystal growth is required to improve perovskite film coverage. The heterogeneous nucleation mechanism in addition provides fewer nucleation sites, when compared to homogeneous nucleation, leading to the formation of larger crystalline domains [34,35]. This interesting observation is extended here to the deposition of perovskite onto organic PTAA substrates, showing that TMDs can also affect this kind of device, which would be ideally suited for  The approach reported in this work was extended to mixed cation-halide CsMAFAP-bIBr perovskite, which is among the best-performing perovskite material for solar cells. Surprisingly, we found that the inclusion of MoS 2 additive into the triple-cation did not significantly affect the final morphology of the perovskite films ( Figure 5), as all the samples exhibited a uniform and compact surface with similar grain size distribution ( Figure 6, Table 3). This can mainly be ascribed to the marginal role of the additive in influencing and controlling the crystallization dynamics of the CsMAFAPbIBr perovskite material. We suppose that in this case, the lower-solubility cesium salts act as heterogeneous nucleation seeds, promoting heterogenic crystal growth [35].
bIBr perovskite, which is among the best-performing perovskite material for solar cells. Surprisingly, we found that the inclusion of MoS2 additive into the triple-cation did not significantly affect the final morphology of the perovskite films ( Figure 5), as all the samples exhibited a uniform and compact surface with similar grain size distribution ( Figure  6, Table 3). This can mainly be ascribed to the marginal role of the additive in influencing and controlling the crystallization dynamics of the CsMAFAPbIBr perovskite material. We suppose that in this case, the lower-solubility cesium salts act as heterogeneous nucleation seeds, promoting heterogenic crystal growth [35].     UV-Vis absorption spectra of the perovskite thin films are shown in Figure 7a. Both pristine MAPbI 3 and MAPbI 3 :MoS 2 films show the same absorption onset. MoS 2 TMD additive, as expected, does not contribute to the absorption spectra due to the high extinction coefficient of perovskite material and due to low additive concentration. Conversely, steady-state photoluminescence (PL, Figure 7b) revealed that the MoS 2 inclusion causes PL quenching, and this is more pronounced at high additive concentrations. We also observed a very small blue shift of the PL band upon additive addition, probably due to a surface and/or grain boundary trap-states passivation effect. The PL quenching observed is a consequence of an enhanced charge transfer from the perovskite structure to the MoS 2 TMD sheets. This is in line with what has previously been observed on alternative systems [37,38] and is related to the optimal band gap alignment and carrier dynamics for these two materials. Moreover, the mechanism of high carrier transfer efficiency among these heterostructures involves both holes and electrons, leading to an improvement in device performance [38] through an intelligently designed charge funneling mechanism. The time-resolved photoluminescence (TRPL) decay (Figure 7c) is consistent with this picture, as the charge carrier lifetime (τ av ) [39] decreases with increasing MoS 2 content, from 120 ns for the pristine MAPbI 3 perovskite to 9 ns for the 20% MoS 2 additive perovskite.

Photovoltaic Performances
Perovskite films with and without MoS 2 additive were included in p-i-n solar cell architectures, as shown in Figure 1. Perovskite solar cells bearing this layout have advantages over n-i-p ones because of the possibility of low-temperature preparation and negligible J-V hysteresis effects [40]. Additionally, the reasons for the selection of PTAA as the hole transporting layer include increased resilience to the negative effects of oxygen and moisture due to its hydrophobicity, and remarkable intrinsic hole mobility [41]. Figure 8 and Table 4 display the J-V curves and the photovoltaic performances of the MAPbI 3 material. The J-V characteristics indicate that MoS 2 sheets as an additive improve device properties. The best PCE of 17.4% was recorded for the 10% MoS 2 additive PSC (FF = 68.9%, V OC = 1.02 V, J SC = 24.76 mA/cm 2 ); however, even at low concentration (5%), the inclusion of MoS 2 additive leads to better performance (16.0%) with respect to the pristine MAPbI 3 device (15.1%). These findings are consistent with the morphological characterization, for which low and intermediate additive concentrations enable us to obtain larger grain sizes, and therefore reduced grain boundary recombination. The J SC shows no significant change with additive modification, and thus the increment in PCE is mainly related to the enhanced V OC . The V OC improvement could be a consequence of suppressed non-radiative recombination processes within the structure, which occur in the absorber layer and at the interfaces between the perovskite and the transporting layers [42]. On the other hand, for higher MoS 2 additive concentration (20%), the PCE significantly drops to 13.3%. Adding an excess of MoS 2 TMD sheets can lead to the formation of a poor-quality perovskite layer, which can be attributed to aggregated MoS 2 nanosheets impairing perovskite growth [43], resulting in shorter diffusion lengths of photogenerated carriers, higher density of trap states, and high carrier recombination rates. This is an indirect proof of how important it is for MoS 2 sheets to be well exfoliated, and thus well isolated, rather than in a bulky or aggregated phase. Decreased performance of the 20% MoS 2 PSC is in good agreement with the photoluminescence measurements that showed the almost completely quenched PL signal at higher additive concentrations (20%). performance [38] through an intelligently designed charge funneling mechanism. The time-resolved photoluminescence (TRPL) decay (Figure 7c) is consistent with this picture, as the charge carrier lifetime (τav) [39] decreases with increasing MoS2 content, from 120 ns for the pristine MAPbI3 perovskite to 9 ns for the 20% MoS2 additive perovskite.  On the other hand, the photovoltaic performance of mixed cation-halide CsMAFAP-bIBr PSCs (Figure 9, Table 5) implemented in an identical device architecture, reported in Figure 1b, did not show any significant change following the MoS2 inclusion. PCE of the best device was above 18%, VOC was between 1.01 and 1.03 V, FF was above 78%, and JSC was above 23 mA/cm 2 for the pristine perovskite and CsMAFAPbIBr:MoS2 (5%). For MoS2  On the other hand, the photovoltaic performance of mixed cation-halide CsMAFAP-bIBr PSCs (Figure 9, Table 5) implemented in an identical device architecture, reported in Figure 1b, did not show any significant change following the MoS 2 inclusion. PCE of the best device was above 18%, V OC was between 1.01 and 1.03 V, FF was above 78%, and J SC was above 23 mA/cm 2 for the pristine perovskite and CsMAFAPbIBr:MoS 2 (5%). For MoS 2 (10%) and MoS 2 (20%), J SC slightly decreases to 22.62 and 22.08 mA/cm 2 , respectively, possibly due to the occurrence of some nanosheet aggregation as observed for MAPbI 3 .
It should be noted, however, that these differences are negligible if we consider a similar mean value and statistical distribution, which reflects the negligible differences in film morphologies ( Figure 5) observed with and without MoS 2 .
Nanomaterials 2021, 11, x FOR PEER REVIEW 13 of 16 (10%) and MoS2 (20%), JSC slightly decreases to 22.62 and 22.08 mA/cm 2 , respectively, possibly due to the occurrence of some nanosheet aggregation as observed for MAPbI3. It should be noted, however, that these differences are negligible if we consider a similar mean value and statistical distribution, which reflects the negligible differences in film morphologies ( Figure 5) observed with and without MoS2.

Conclusions
In summary, we demonstrated an effective additive-engineering approach to improve the morphology of perovskite films grown on an organic substrate, leading to superior device performance. Liquid-phase exfoliated molybdenum disulfide sheets were employed as an additive in MAPbI 3 perovskite precursor solution, to measure their effect at diverse concentrations and upon diverse precursor formulations. Our findings suggest that MoS 2 incorporation (10% V/V) into a MAPbI 3 perovskite photoactive layer results in high-quality films with larger grains and optimized morphology, suitable for device integration. As a result, the champion device yields a remarkable PCE of 17.4%, 15% higher than the undoped device. The beneficial impact of 2D layered materials originates from seed induced heterogeneous nucleation, resulting in superior perovskite film morphology, and therefore in high-performance PSCs. Conversely, for the mixed cation-halide perovskite, no improvements were observed with the addition of the additive to the precursor ink. Our results confirm that the nucleation process differs for distinct perovskite precursor compositions, and so, that the process is influenced differently by the presence of additives. These results contribute to the development of perovskite-based solar cells and, more specifically, to the wide research front focused on controlling photoactive film deposition and growth.