Strong Photocurrent from Solution‐Processed Ruddlesden–Popper 2D Perovskite–MoS2 Hybrid Heterojunctions

This study reports overall improvement in structural, morphological, and optoelectronic properties of Ruddlesden–Popper (RP) perovskites of type (BA)2(MA)n−1PbnI3n+1 by forming their bulk heterojunction hybrids with few‐layer MoS2 nanoflakes. RP perovskite–MoS2 hybrid thin films have shown significantly improved packing and crystallinity compared to pristine perovskites. The presence of MoS2 at RP perovskite interface has improved the quantum confinement effects and transport of photogenerated charge carriers from perovskite to MoS2, due to suitable conduction band of MoS2 and more number of decay channels. The optoelectronic properties of RP perovskite–MoS2 hybrids are studied for various MoS2 concentrations (4.2–25.6 × 10−3 m) and at optimum concentration (12.8 × 10−3 m) the photodetectors (n = 2, 4) have shown strong, sharp, and highly stable photocurrent response. At 0.0 V bias, the RP perovskite (n = 4) and MoS2 (12.8 × 10−3 m) hybrid‐based photodetectors, prepared without any encapsulation, have shown strong photocurrent density of ≈9.8 µA cm−2 under 1 sun illumination, which is ≈17 times higher compared to the pristine RP perovskites‐based photodetector (0.6 µA cm−2). Further transient photocurrent, performed over 200 cycles for hybrid (n = 4+MoS2) thin film photodetector under laser (λex ≈ 405 nm, ≈630 mW cm−2) illumination and ambient air conditions has shown highly stable photocurrent with only ≈9.6% reduction in the peak photocurrent density.


Strong Photocurrent from Solution-Processed Ruddlesden-Popper 2D Perovskite-MoS 2 Hybrid Heterojunctions
Rashid M. Ansari, Akshaykumar D. Salunke, Mohammad Rahil, and Shahab Ahmad* DOI: 10.1002/admi.202202170 their comparable absorption coefficients (≈10 4 cm −1 ), [3] charge carrier lifetimes (≈30 ns), [4] mobilities (≈9 cm 2 V −1 s −1 ), [4] and diffusion lengths (≈7-14 µm) [5] make them potential material for various optoelectronic and photo-electrochemical applications such as solar cells, [6,7] photo detectors, [8][9][10][11] light-emitting diodes (LEDs), [12] lasers, [13] solar-water splitting, [14,15] photobatteries, [16] and so on. In short, the RP perovskites have a generic formula of (R-NH 3 ) 2 (MA) n−1 M n X 3n+1 where R-NH 3 is the alkyl spacer cation (n-C 4 H 9 NH 3 + , iso-C 4 H 9 NH 3 + , C 6 H 5 C 2 H 4 NH 3 + , etc.), M 2+ is the divalent metal (Sn +2 , Pb +2 , Ge +2 , etc.), X − is the halide (Cl − , Br − , I − ), and n denotes the number of inorganic layers between the organic moieties. [17] For n = 1 the perovskite structure represents a pure 2D material structure where inorganic metal halide layers and organic moiety layers are alternatively stacked, thus forming a perfect multiple quantum well type structure. However, for higher members (n > 2), a quasi-2D perovskite structure is formed, where the stacked inorganic layers are still separated by the organic moieties but their metal halide octahedrons (MX 6 4− ) are connected to each other via terminal and bridging halides, thus forming a n-value limited 3D perovskite structure as well. For n∼∞, the perovskite structure approaches an extended 3D perovskite structure. [17] Thus, the structural stability and optoelectronic properties of these RP perovskites have direct dependence on their dimensionalities, where lower n-value members offer better stabilities and higher n-value members show improved optoelectronic properties. [17] Various reports also show that the perovskite photodetectors are fabricated using crystalline perovskite of different morphologies such as thin films, nanocrystals, nanowires, nanoflakes, etc., which showed varied performances due to different surface area offered by the perovskites for interaction with light as well as due to the variation in the density of states. [18][19][20][21][22][23][24][25] However, the overall performance of these photodetectors also depends upon the electron transport layers (ETLs) and hole transport layers (HTLs). Attempts are made to improve the efficiencies of bulk and quasi-2D perovskite-based photodetectors using various ETLs/HTLs such Spiro-OMeTAD, [26] PEDOT:PSS, [27] P3HT, [28] and PTAA [29] as HTLs as well as PCBM [27] as ETLs. Although with the use of organic charge transport materials the performance of perovskite photodetectors has been improved, but at the same time the fabrication cost increased due to use of expensive organic materials-based charge transport layers. [30] Therefore, in order to get a stable RP perovskite-based photodetector it is important to explore the suitable inorganic charge transport materials which may improve the performance of the device as well as bring down the overall cost, which is desirable toward commercialization of these devices.
In the recent past various 3D perovskite-based photodetectors are demonstrated where different 2D inorganic materials such as transition metal dichalcogenides (TMDs), [31,32] graphene, [33][34][35] reduced graphene oxide (rGO), [36] etc. are used, due to the various reasons mentioned below, to improve the extraction of photogenerated charge carriers from the perovskite film. In these photodetectors different HTMs such as few-layer MoS 2 or graphene are introduced in the planar device configuration to make interface with perovskite film whereas rGO is blended as ETM with the 3D perovskite. [37] Among many 2D inorganic materials, MoS 2 (mono to few layer) has potential to serve as an ETL, [38] HTL [39] as well as buffer layer (BL), [40,41] because of its tunable bandgap from 1.2 to 1.8 eV, [42,43] which can be achieved by adjusting the number of layers. Moreover, few-layer MoS 2 offers fast transport of charge carriers in vertical direction, [44] fewer traps as it has no dangling bonds on the surface, [38] high carrier mobility of nearly 500 cm 2 V −1 s −1 [45] which makes them a potential charge transport material for optoelectronic applications. In 2017 Wang et al. fabricated a 3D perovskite and 2D MoS 2 -based photodetector and demonstrated that the ON/OFF ratio of a photodetector can be improved in the photodetectors by including the 2H phase into the device rather than 1T phase of 2D MoS 2 . [37] In 2019 Ranbir et al. studied the 3D perovskite MAPbI 3 solar cell with CVD-grown MoS 2 as the ETL and Spiro-OMeTAD as the HTL and reported 13.2% PCE, comparable to the PCEs achieved from SnO 2 and compact TiO 2 ETL-based PSCs, [38] which was attributed to the wide transmission in the visible spectrum from 400 to 780 nm, [46] with absorption efficiencies above 77% and maximum of 88% of MoS 2 . [38] In 2020 Zhiyong et al. investigated the PSC with MoS 2 serving as both an additive as well as BL between PEDOT:PSS and perovskite. This BL prevents the decomposition of perovskite film due to aqueous PEDOT:PSS solution. An improvement in PCE from 15.29% to 18.31% of PSC is reported with active layer of MAPbI 3 :MoS 2 (10 v%) and two-layer MoS 2 as BL, moreover 87% of the initial PCE was preserved after 20 d. [40] In 2018 Saman et al. investigated the PCE of a 3D perovskite MAPbI 3 -based solar cell with variable thickness of MoS 2 functioning as the HTL. An increase in PCE is observed with increase in thickness of MoS 2 , reaching 20.53% for 1.34 nm thick MoS 2 film and as the thickness of MoS 2 film increased to 21.44 nm PCE reduced to 16.21%. This fall in PCE was attributed to the rise of resistance in the device due to the multilayer MoS 2 and reduced light transmission through MoS 2 . [39] These recent reports demonstrate that monolayer as well as few-layer MoS 2 can be a promising additive, buffer layer as well as charge transport inorganic material which can efficiently improve the performance of perovskite-based optoelectronics due to their suitable band alignment with perovskite, improved interface area for better charge transport and can offer a stable heterojunction, solvent compatibility, etc. However, the effect of these TMDs is not explored on the performance of RP perovskite-based optoelectronic devices, which is very important in order to achieve stable and efficient operation of such devices.
In this work, we report the fabrication of RP perovskite and MoS 2 -based hybrids by synthesizing n-butylamine (BA, CH 3 (CH 3 ) 3 NH 3 + ) based RP perovskite series (CH 3 (CH 3 ) 3 NH 3 ) 2 (CH 3 NH 3 ) n−1 Pb n I 3n+1 , n = 1-4 (hereafter (BA) 2 (MA) n−1 Pb n I 3n+1 , n = 1-4, where MA is CH 3 NH 3 + ) and few layer (around five layers) exfoliated MoS 2 nanoflakes (2H phase). 2H phase MoS 2 are obtained by liquid phase exfoliation techniques (LPET), [47] which are blended with the RP perovskite photoactive layer to form bulk heterojunction (BHJ). The effect of presence of MoS 2 on the structural, morphological, and optical properties of RP perovskite is investigated. Furthermore, the optoelectronic properties of the proposed perovskite-MoS 2 hybrids are demonstrated by using vertical configuration photodetectors with different concentrations of MoS 2 in the RP perovskite photoactive layer. The MoS 2 , which worked as an ETM for (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4), at the perovskite interface, has affected the transport of photogenerated charge carriers as observed in the PL quenching, performed for the perovskite-MoS 2 hybrid thin films. [37,38] After substantial optimization for the composition of MoS 2 (12.8 × 10 −3 m), the RP perovskite-hybrid photodetector (n = 4+MoS 2 ) has shown significant enhancement in the photocurrent density (9.8 µA cm −2 ) and responsivity (98 µA W −1 ) compared to the reference device without MoS 2 (0.6 µA cm −2 , 6 µA W −1 ) under Xe lamp standard 1 sun (100 mW cm −2 ) illumination. We also report a highly stable 200 cycles of transient photocurrent performed under the ambient atmosphere for the fabricated RP perovskite-MoS 2 hybrid (n = 4+MoS 2 ) photo detector without using any encapsulation. Such RP perovskite-MoS 2 hybrids in the BHJ configuration are not explored till now for the photodetectors applications, moreover this work opens up the opportunity window to investigate various RP perovskites and TMDs-based hybrids for real-life optoelectronic device applications.

RP Perovskite-MoS 2 Hybrids
(BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) perovskite materials are prepared by adapting reported synthesis route, [17] using PbO, BA, MA as precursors and H 3 PO 2 as reducing agent and HI (57 wt% in H 2 O, distilled, stabilized, 99.95%) as reaction medium. The number of layers, represented by n-value, in the RP perovskites is controlled by changing the molar ratio of BA, MA, and H 3 PO 2 (See the Experimental Section for detailed synthesis procedure). The schematic crystal structures of prepared (BA) 2 (MA) n−1 Pb n I 3n+1 perovskite for n = 1-4 are shown in Figure 1a, where the number of inorganic metal halide octahedron [PbI 6 ] 4− layers is defined by the n-value. Organic spacers BA are sandwiched between varying thickness of octahedrons by weak intermolecular interactions and hydrophobic forces during self-assembling process, whereas MA cations are positioned in between the octahedrons thus forming an orthorhombic crystal structure. [17] These alternatively stacked and confined organic-inorganic layers enable the dielectric and quantum confinement effects which results in excellent www.advmatinterfaces.de structural and optical properties, which are discussed below. [48] The optical band tunability, from yellow (n = 1) to black (n = 4), can be observed in the digital camera images of pristine RP perovskite spin-coated thin film (Figure 1b).
A stable MoS 2 dispersion is prepared from bulk MoS 2 flakes by using LPET in NMP (see the Experimental Section) as the surface tension of NMP at ≈40 mJ m −2 closely matches with the surface energy of many layered materials. [49][50][51] Moreover, LPET is cost-effective, environmentally benign, and contaminationfree technique to produce MoS 2 dispersion. The RP perovskite-MoS 2 hybrid for n = 1-4 are prepared by blending its 0.5 m molar mass (which correspond to 25.9 mg for n = 1, 44.4 mg for n = 2, 63.0 mg for n = 3, and 81.6 mg for n = 4 of RP perovskite) into 60 µL MoS 2 solution with six different millimolar concentration of 4.2, 8.5, 12.8, 17.0, 21.3, and 25.6 × 10 −3 m followed by ultrasonication and stirring for uniformly distribution. As the size of MoS 2 nanoflakes vary from 30 to 200 nm (discussed below), it is expected that when the solution of hybrids are casted, then during crystallization the perovskite crystals undergo heterogeneous nucleation [40] triggered by the presence of MoS 2 heterogeneous phase in perovskite precursors, as shown in close view schematic representation (Figure 1d).

Effect on Structural and Surface Properties
Thin film X-ray diffraction (XRD) measurements are carried out to understand the crystal structures of these RP perovskite materials, MoS 2 and RP perovskite-MoS 2 hybrids. XRD plots of RP perovskite materials and RP perovskite-MoS 2 hybrid thin films, shown in Figure 1e, confirm the highly crystalline nature of the films. As evident from the crystal structure of n = 1, these pure 2D materials favor growth perpendicular to the substrate, i.e., along the c-axis, thus show characteristic (00l) reflections. [48] The inclusion of MA cations and a change in the stoichiometric formula result in n = 2 RP perovskite. Figure 1. a) Schematic illustration of the crystal structure of (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) and b) the corresponding digital camera image of thin films. c) The digital camera image of MoS 2 dispersion prepared by LPET in NMP, d) schematic illustration of (BA) 2 PbI 4 perovskite crystals assembled on MoS 2 to form heterostructure, e) X-ray diffraction (XRD) patterns of pristine (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) spin-coated thin films, MoS 2 drop-cast film, and (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 2 and 4) + MoS 2 (12.8 × 10 −3 m) hybrid spin-coated thin films. All the films are fabricated on the glass substrate.

www.advmatinterfaces.de
The existence of (151), (190), and (113) reflections indicates that the MA cations were inserted between octahedron layers. These reflections confirm the progressive growth of the unit cell along c-axis. [48] The XRD plot of n = 3 and n = 4 has shown the (111) and low intensity (222) reflections peaks which are attributed to the characteristic 3D perovskite phase. For the MoS 2 drop-casted film, XRD plot has shown increased FWHM and reduced intensity for the characteristic (002) peak, which confirmed the presence of few-layer MoS 2 ( Figure 1e and Figure S1, Supporting Information). [49] Moreover, the characteristic (002) peak at 14.4° confirmed the presence of the 2H phase (trigonal prismatic coordinated crystal structure) of MoS 2 , [52] which is suitable for optoelectronic applications because of its semiconducting nature. [53][54][55][56] The XRD plot for (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 2 and 4)+MoS 2 (12.8 × 10 −3 m) hybrid thin films, showed that plane diffraction angles of the perovskite did not change due to the presence of MoS 2 . Moreover substantial increase in the intensity of reflected X-rays are observed which shows that the presence of MoS 2 has significantly improved the crystallinity of RP perovskites. [40] This is attributed to the fact that presence of MoS 2 improves the packing of the perovskite crystals as a results of heterogeneous nucleation of perovskite caused by the presence of heterogeneous MoS 2 phase in the perovskite precursors which promotes the crystallization during annealing (100 °C). [40,57] Surface morphology and size of MoS 2 (Figure 2a) are investigated using FESEM, which shows that after exfoliation the micrometer size bulk MoS 2 flakes [58] are reduced to nanoflake structures of ≈60 nm average size. AFM is used to find out the number of MoS 2 atomic layers present in the nanoflakes by measuring the height of the nanostructures on glass substrate ( Figure 2b). The AFM line profile measurements revealed the height of MoS 2 is ≈3.712 nm (see Figure S2, Supporting Information), which corresponds to five atomic layers of MoS 2 , [59] assuming single MoS 2 atomic layer is 0.65 nm thick. [60] The presence of MoS 2 (4.2 and 12.8 × 10 −3 m) in the RP perovskite (n = 3 and 4, respectively) thin films has improved the morphology of the RP perovskite-MoS 2 hybrid thin films by significantly reducing the pinholes and improving the coverage of the perovskite films (Figure 2c-f). Figure S3 (Supporting Information) shows the SEM images of drop-casted and spin-coated films of n = 3 RP perovskites and its hybrid. Such improved morphology and high crystallinity of the perovskite-MoS 2 hybrids can result in large light absorption, reduced defects, and series resistance, as well as reduced trap-assisted recombinations and thus improve the device performance significantly. [48] The observed improvement in perovskite-MoS 2 film morphology and reduced pinholes are attributed to heterogeneous nucleation of perovskite crystals triggered by the presence of MoS 2 as discussed above. [40] The interface of perovskite crystals and MoS 2 , as shown in the schematic representation Figure 1d, results in the formation type-II (staggered) semiconductor heterojunction.
The elemental composition and their spatial distribution in the hybrid, n = 3+MoS 2 (12.8 × 10 −3 m) drop-casted film are verified by the energy-dispersive X-ray (EDAX) analysis (see Figure S4, Supporting Information). The corresponding elemental composition spectra and mapping confirm the presence of carbon, iodine, lead, molybdenum, and sulfur, uniformly distributed on the surface of the film.

Optical Characterizations of RP Perovskite-MoS 2 Hybrids
The thin film optical absorbance measurements of pristine as well as MoS 2 -doped RP perovskite has shown strong exciton absorption peaks and uniform bandgap tunability in the visible region of spectra ranging from 513 nm for n = 1 to 644 nm for n = 4 (Figure 3a,b). The presence of multiple exciton peaks for n = 2-4 confirms the presence of mixed phases; [61,62] however, the optical bandgaps are measured using Tauc's plot (see Figure S5a, Supporting Information) for the characteristic absorption peaks at 513 nm (λ 1 , n = 1), 567 nm (λ 2 , n = 2), 607 nm (λ 2 , n = 3), and 644 nm (λ 3 , n = 4) and have shown decrease in energy band from 2.48 to 1.93 eV, respectively. [63,64] The absorbance of MoS 2 is performed on drop-casted film to further confirm  (Figure 3b). The appearance of peaks between 600 and 700 nm revealed the presence of a 2H phase, which rendered them semiconducting. These peaks are ascribed to A and B excitons originating from the K point of the Brillouin zone, which are responsible for the semiconducting behavior. [37,65] The characteristic absorption peak (669 nm) confirms the presence of few-layer MoS 2 , which corresponds to the optical bandgap of 1.58 eV (see Figure S5b, Supporting Information). Absorbance spectra of (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4)+MoS 2 (12.8 × 10 −3 m) thin films have shown that the exciton absorption peaks became more intense and narrow, with no deviation in the peak wavelengths (Figure 3b), which is attributed to enhanced light absorption due to improved surface morphology and crystallinity of the perovskite films (see Figure S5c, Supporting Information). [38] The full-width at halfmaximum (FWHM) values of (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) are 27.37, 26.13, 19.58, and 18.03 nm, respectively, where (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4)+MoS 2 have 26.28, 20.93, 18.89, and 17.93 nm (Figure 3c). The reduced FWHM for the MoS 2 doped perovskite film shows that the confinement effects are more pronounced due to improved stacking of organic and inorganic layers (Figure 3c). [18,66] PL measurements of pristine (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) thin films, performed using 405 nm laser excitation (≈10 um beam spot) in reflection mode, have shown strong exciton emission peaks at 513 nm (λ 1 , n = 1), 571 nm (λ 2 , n = 2), 607 nm (λ 2 , n = 3), and 644 nm (λ 3 , n = 4) (Figure 3d). The occurrence of additional emission peaks for n = 2-4 is attributed to the presence of mixed phase of RP perovskite. [61] The presence of MoS 2 in the perovskite film has shown PL quenching effect, which is observed by recording the PL images of pristine and MoS 2 -doped perovskite thin films (prepared on glass substrate) under UV illumination (LED, λ ex ≈ 300 nm, without any long pass filter) ( Figure 3e). The quenched PL from the perovskite-MoS 2 film is due to electron transfer behavior [38] to MoS 2 (conduction band minima, CBM ≈ 4 eV and valence band maxima, VBM ≈ 5.58) in n = 1-4 which has reduced the radiative combinations in the perovskite due to efficient extraction of photogenerated electrons from perovskite. However, further investigations, such as micro-confocal PL and impedance spectroscopy, are suggested to understand the role of MoS 2 in improving the charge transport properties of RP perovskite films.
To further understand the exciton recombination dynamics in the pristine and MoS 2 -doped perovskite thin films (deposited on glass substrate) the time-resolved PL measurements are performed using time correlated single photon counting (TCSPC) setup with 405 nm pulsed diode laser (see Figure 3f). The PL decay curves corresponding to characteristic emission wavelengths mentioned in Table 1 are probed and lifetime of photogenerated charge carriers is calculated by fitting the decay curves using double exponential decay function [10] (see Figure S6a,b, Supporting Information). The charge carrier lifetimes are found to increase linearly from 276 ps, for n = 1, to 888 ps, for n = 4, for pristine perovskite films (see  Table S6c, Supporting Information), whereas the lifetimes are found to decrease from 240 ps, for n = 2, to 126 ps, for n = 4, for MoS 2 -doped perovskite films (see Figure 3f). The relative decay in the average lifetime for hybrid films reflects the fact that as the perovskite bandgap is decreasing with n-value, their VBM and CBM alignment with MoS 2 improves, which cause efficient dissociation of e-h pairs at the perovskite-MoS 2 heterojunction (see Figure S7, Supporting Information). Such e-h pair dissociation results in forming more decay channels due to efficient transfer of electrons to MoS 2 from perovskite (n = 1-4), while blocking the holes. [38] Thus, these hybrids can be explored for fast response optoelectronics such as photodetectors.

Effect of MoS 2 Doping on Optoelectronic Properties of Hybrids
To investigate the effect of MoS 2 on the optoelectronic properties of RP perovskites, the photocurrent studies are performed on the RP perovskite and MoS 2 -doped RP perovskite thin films by fabricating the vertical configuration photodetector on patterned FTO substrates (FTO/(BA) 2 (MA) n−1 Pb n I 3n+1 -MoS 2 /Al) (see the Experimental Section, Figure 4a). As reported elsewhere, the VBM and CBM of (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) increases to higher energy with the increase in n-value. [67,68] Therefore, MoS 2 functions as ETM [38] with CBM at 4.0 eV [69] that transport photoexcited electrons generated from (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4), whereas inhibited unpaired photoinduced holes with VBM at 5.58 eV (see Figure 4b and Figure S7, Supporting Information). Under illumination, the photogenerated e-h pairs got instantaneously separated and collected by the current collector electrode (Al and FTO) as shown in Figure 4b. The J-V measurements, performed from −1.0 to +1.0 V on (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) and respective hybrids with MoS 2 (12.8 × 10 −3 m) thin films have shown diode like behavior which confirms their semiconducting behavior (Figure 4c,d and Figure S8, Supporting Information) and upon illumination with standard 1 sun light (with AM1.5 filter) an increase in photocurrent is observed. However, for perovskite (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4)+MoS 2 (12.8 × 10 −3 m) hybrid thin films much higher photocurrent is observed for similar measurements, which confirmed that the MoS 2 are actively participating as a only ETM for the (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) without affecting the intrinsic semiconducting properties of perovskite (Figure 4c,d).
Next, to confirm the role of MoS 2 on the transport properties of photogenerated charge carriers in perovskite (n = 2 and 4), we performed the transient photocurrent studies with different molar concentration of MoS 2 . These transient photocurrent responses are recorded by using a potentiostat work station (see the Experimental Section) where a manual shutter is used to switch ON and OFF the illumination every 15 s, providing a square wave-like pulsed excitation. At 0 V bias, upon illumination with 1 sun light the pristine perovskite (n = 2) film has shown fast rise (800 ms) in photocurrent to ≈40 nA cm −2 and when the light is turned off after 15 s the photocurrent decayed rapidly (800 ms) (see Figure 5a and Figure S9a, Supporting Information). However, doping of MoS 2 (4.2 × 10 −3 m) has resulted in higher photocurrent of ≈63 nA cm −2 . Further for 8.5 × 10 −3 m MoS 2 -doped perovskite film (n = 2) similar fast rise (900 ms) in photocurrent to ≈110 nA cm −2 is observed, however when the light is turned off the photocurrent decayed slowly (15 s) (see Figure S9b, Supporting Information). Compared to the pristine perovskite where the recombination rate of photogenerated e-h pairs is relatively high, the slow decay in the photocurrent of hybrids is attributed to the absence of HTM which causes imbalance between the extracted e-h pairs, as MoS 2 extract the electrons very efficiently which leaves large number of holes in the VB of perovskite. [38] Investigations with suitable HTM are required to understand the photogenerated carrier dynamics better. Since n = 4 has lowest bandgap (1.94 eV) among (BA) 2 (MA) n−1 Pb n I 3n+1 , n = 1-4 RP perovskites, therefore we have extensively investigated MoS 2 -doped n = 4 perovskite films. As expected, under similar illumination conditions, at 0 V bias higher photocurrent density of ≈602 nA cm −2 is observed for pristine n = 4 perovskite thin film (see Figure 6a), indicate generation of large number of e-h pairs due to improved absorption in the visible region owing to lower energy bandgap. Similar to n = 2, the pristine n = 4 based photodetector has also shown fast photocurrent rise (≈200 ms) and decay (≈200 ms) in transient photocurrent measurements (see Figure S9c, Supporting Information). Photocurrent response for the (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 4)+MoS 2 hybrids based photodetectors increased to ≈1 µA cm −2 (4.2 × 10 −3 m MoS 2 ), ≈3.8 µA cm −2 (8.5 × 10 −3 m MoS 2 ), and ≈9.8 µA cm −2 (12.8 × 10 −3 m MoS 2 ) (see Figure 5b) on the adding MoS 2 compared to the Table 1. Room-temperature exciton absorption peak wavelength λ abs (nm) and corresponding energy E abs (eV), exciton PL peak wavelength λ PL (nm) and corresponding energy E PL (eV), Stokes shift (meV), energy bandgap (eV), exciton binding energy E B.E. (meV), and lifetime (ps) for (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) thin films.
(BA) 2  www.advmatinterfaces.de reference device (≈0.6 µA cm −2 ). In addition to the improved photocurrent densities compared to (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 2)+MoS 2 relatively shorter photocurrent rise and decays times are observed for (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 4)+MoS 2 hybrids (see Figure S9d, Supporting Information). To further confirm that the faster photocurrents is entirely due to the perovskite-MoS 2 heterojunction, we performed the photocurrent measurements on the pristine MoS 2 drop-casted filmbased photodetector, which showed much slower photocurrent rise (≈6.2 s) and decay (≈15 s) times. The transient photocurrent measurements are performed over 10 cycles under 1 Sun illumination and showed highly stable and strong photocurrent for (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 4)+MoS 2 hybrids. To demonstrate the stable performance of hybrids we have recorded the transient photocurrent in ambient air condition for long cycles (50 cycles of 15 s ON and 15 s OFF) at 0 V bias using a DPSS laser (λ ex ≈ 405 nm, ≈630 mW cm −2 ) without encapsulating the devices. The hybrid (n = 4+MoS 2 ) based photodetector has shown fairly stable performance with slight reduction of ≈7% in the photocurrent density (t = 0 min, 12.89 µA cm −2 and t = 25 min, 11.97 µA cm −2 ), whereas the pristine (n = 4) perovskite-based detector has shown significant reduction of ≈39% (t = 0 min, 1.1 µA cm −2 and t = 25 min, 0.67 µA cm −2 ) over 50 cycles (Figure 6a). Furthermore, the transient photocurrent performed for 200 cycles for the fresh hybrid (n = 4+MoS 2 ) based photodetector under similar conditions has shown fairly stable performance with slight reduction of ≈9.6% in the peak photocurrent density (t = 0.0 h, 12.95 µA cm −2 and t = 1.67 h, 11.70 µA cm −2 ) (see Figure 6b and Figure S11, Supporting Information). We want to emphasize again that here the un-encapsulated device is exposed to the high power laser (λ ex ≈ 405 nm, ≈630 mW cm −2 ) for the stability test which has ≈6.3 times higher intensity compared to the standard light source (1 sun, Xe lamp) and it is noteworthy that photo response profile after 200 cycles is similar to initial cycle. It demonstrated that addition of MoS 2 not only increased the photocurrent response (≈11.8 times) but also enhanced the stability of the RP perovskites. In addition to stable photocurrents, the photodetectors also showed very low dark currents. Further increase in the doping concentration of MoS 2 (17-25.6 × 10 −3 m) in (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 4) reduces the photocurrent density significantly, though these photodetectors have still shown higher photocurrent compared to pristine (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 4) film (Figure 5c). Such decrease in photocurrent of the photodetector for heavily doped MoS 2 (17 to 25.6 × 10 −3 m) in the RP perovskites film is attributed to the fact that excessive presence of MoS 2 nanoflakes lower the intensity of light reaching the perovskite as these excessive MoS 2 flakes settles on the FTO substrate therefore this reduces the transmission of light. Moreover, the excessive presence of MoS 2 also increases the film resistance due to increase in number of junctions. [39] Photoresponsivity (R) is another important parameter for photodetectors which indicate how efficiently the photodetector responds to an optical signal. Responsivity is defined as The difference between the J light and J dark for pristine (n = 2 and 4) and MoS 2 -doped RP perovskite at 0 V applied bias is obtained from the transient photocurrent measurements as shown in Figure 5a-c. For pristine (n = 2) photodetector, the R is ≈0.4 µA W −1 which increased to ≈0.63 and ≈1.1 µA W −1 on doping MoS 2 with molar 4.2 and 8.5 × 10 −3 m concentrations, respectively (see Figure S12, Supporting Information). Similar to n = 2, pristine (n = 4) showed R ≈ 6 µA W −1 , which increased to ≈10, ≈38.8, and ≈98 µA W −1 on doping MoS 2 with 4.2, 8.53, and 12.8 × 10 −3 m molar concentrations, respectively. Whereas for heavily MoS 2 -doped perovskites R decreased from ≈18.8 to ≈13.5 µA W −1 with increase in MoS 2 doping from 17 to 25.6 × 10 −3 m, respectively (Figure 5d), which is still higher than R (≈6 µA W −1 ) obtained for pristine RP perovskite. Thus, an optimum performance of thin film photodetector is obtained for the 12.8 × 10 −3 m MoS 2 -doped RP perovskite hybrid.
The detectivity (D) of a photodetector is used to determine how well a photodetector can detect a weak signal, whereas noise equivalent power (NEP) is used to determine power generated by noise source and these parameters are defined as [70] where R = photoresponsivity, q = charge of electron, J dark = dark current density, and A = active area of the device. It is found that D for pristine device (n = 4) is only 0.71 × 10 8 cm Hz 1/2 W −1 and as we dope MoS 2 of concentration 4.2 and 8.5 × 10 −3 m, it increases slowly to 1.26 × 10 8 cm Hz 1/2 W −1 and for further doping (12.8 × 10 −3 m) it sharply increased to 6.93 × 10 8 cm Hz 1/2 W −1 which is ≈9.76 times higher as compare to its pristine sample. Similarly, NEP is reduced ≈9.76 times for 12.8 × 10 −3 m doped MoS 2 compared to its pristine device (see Table 2 and Figure 5d). Furthermore, as expected from the transient photocurrent measurements for the heavily doped photodetectors (17-25.6 × 10 −3 m), the D decreased and NEP also increased, which is attributed to the increase in the film resistance and poor interaction of light with the perovskite.

Conclusion
The hybrids of RP perovskites (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) and exfoliated few layer MoS 2 nanoflakes are fabricated by solution processing route and their improved structural and optoelectronic properties are investigated extensively. The SEM imaging of RP perovskite-MoS 2 hybrids demonstrated improvement in film packaging and reduction in pinholes compared to pristine RP perovskite thin films. XRD studies showed huge improvement in the crystallinity for the hybrid thin films. Such improved surface morphology and crystallinity are attributed to the heterogeneous nucleation of perovskite crystals

www.advmatinterfaces.de
triggered by the addition of MoS 2 in the RP perovskite precursors. The room temperature optical absorbance measurements demonstrated narrower FWHM for the perovskite-MoS 2 (n = 1-4) hybrids which confirmed that the optical confinement effects are more pronounced due to improved stacking of organic and inorganic layers compare to pristine RP perovskites. PL quenching effect, studied by performing the PL imaging, and the time-resolved PL measurements performed on the hybrid thin films confirmed that the addition of MoS 2 has improved the extraction of photogenerated charge carriers from RP perovskite (n = 2-4). Moreover, the nanoflake structure of MoS 2 helped in the formation of more number of decay channels which resulted in efficient extraction of photoelectrons. The transient photocurrent measurements, performed on vertical photodetectors with RP perovskite-MoS 2 hybrids as active light absorbing material, showed strong, sharp, and stable photocur-rents. The addition of MoS 2 in RP perovskites formed bulk heterojunction where MoS 2 worked as electron transport materials for (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) perovskites which resulted in the improved light absorption and enhancement in photocurrent densities. At 0.0 V bias these hybrid-based photodetectors have shown strong photocurrent densities of ≈9.8 µA cm −2 (n = 4+MoS 2 ) and ≈110 nA cm −2 (n = 2+MoS 2 ) on addition of 12.8 and 8.5 × 10 −3 m MoS 2 , respectively, which are ≈16.28 (n = 4) and ≈2.75 (n = 2) times higher compared to the photocurrent densities obtained for the corresponding pristine RP perovskites under standard 1 sun illumination. Such enhancement in the photocurrent response has resulted in the photoresponsivities of ≈98 µA W −1 (n = 4+MoS 2 ) and ≈1.1 µA W −1 (n = 2+MoS 2 ). In addition to photoresponsivities, n = 4+MoS 2 hybrid photodetector have shown detectivity of 6.93 × 10 8 cm Hz 1/2 W −1 for 12.8 × 10 −3 m doped MoS 2 which is ≈7 times higher compared to its pristine sample (0.71 × 10 8 cm Hz 1/2 W −1 ). Next, we also studied the effect of MoS 2 doping on the optoelectronic properties of RP perovskite (n = 4) by varying the MoS 2 concentrations (4.2 to 25.6 × 10 −3 m) in order to get the optimum doping level for MoS 2 nanoflakes (12.8 × 10 −3 m). Furthermore, the transient photocurrent measurements confirmed the stability of the proposed hybrids by performing 50 cycles on non-encapsulated devices under laser illumination (λ ex ≈ 405 nm CW-DPSS, ≈630 mW cm −2 ) at 0.0 V bias under ambient air conditions. n = 4+MoS 2 hybrid photodetector has shown fairly stable performance with slight reduction of ≈7% in the peak photocurrent density over 50 cycles and 9.6% over 200 cycles for the fresh device, whereas the pristine RP perovskite (n = 4) photodetector has shown significant reduction of ≈39%. With these findings we conclude that the addition of MoS 2 nanoflakes not only increases the photocurrent response  www.advmatinterfaces.de of RP perovskites but also enhances their structural stability, which is very important in order to demonstrate their applications in real-life optoelectronic devices.

Experimental Section
Materials: All the chemicals were purchased from Sigma-Aldrich and used without any further purification. Lead  /MACl, which resulted in the formation of black precipitate. These precipitates again got redissolved when the combined solution was heated to boiling. The final solution was left to cool at room temperature without stirring, which resulted in the formation of brown color rectangular-shaped RP perovskite crystals.
(BA) 2 (MA)Pb 2 I 7 (n = 2): MACl (169 mg, 2.5 mmol) powder was added to the PbI 2 solution which formed black powder precipitate in the solution, which quickly dissolved to give clear bright yellow solution under continuous stirring. In a separate beaker, BA (347 µL, 3.5 mmol) was neutralized in an ice bath in 57% w/w aqueous HI solution (2.5 mL, 19 mmol) to get BAI solution. Next, BAI solution was added to the clear bright yellow solution of PbI 2 /MACl, which resulted in the formation of black precipitate. These precipitates again got redissolved when the combined solution is heated to boiling. The final solution was left to cool at room temperature without stirring, which resulted in the formation of cherry red color rectangular-shaped RP perovskite crystals.
(BA) 2 PbI 4 (n = 1): BA (462 µL, 5 mmol) was neutralized in an ice bath in 57% w/w aqueous HI solution (2.5 mL, 19 mmol) to get BAI solution. Next, BAI solution was added to the clear bright yellow solution of PbI 2 which resulted in the formation of black precipitate. These precipitates again got redissolved when the combined solution is heated to boiling. The final solution was left to cool at room temperature without stirring, which resulted in the formation of yellow color rectangular-shaped RP perovskite crystals.
Separate precursor solution of 0.5 m was prepared for all (BA) 2 (MA) n−1 Pb n I 3n+1 (n = 1-4) RP perovskites by dissolving perovskites in 60 µL of NMP, which correspond to 25.9 mg for n = 1, 44.4 mg for n = 2, 63.0 mg for n = 3, and 81.6 mg for n = 4 RP perovskite. These solutions were used to get the thin films of pristine RP perovskites.
Preparation of MoS 2 Nanoflakes and RP Perovskite-MoS 2 Hybrids: Exfoliation of MoS 2 was carried out using LPET. 50 mg bulk MoS 2 powder were added in 10 mL NMP solvent and ultrasonicated for 25 h. After sonication, the bigger MoS 2 chunks were removed by performing centrifugation at 2000 rpm for 10 min, and the tube was left in the stand for 30 min to further settle the bigger MoS 2 particles. Next, the supernatant suspension was carefully extracted using a micropipette. The final concentration of MoS 2 nanoflakes was estimated using microbalance (Sartorius, Model Number: SE2) and found to be ≈4.12 mg mL −1 or 25.6 × 10 −3 m. Six different millimolar concentration MoS 2 solutions (4.2, 8.5, 12.8, 17.0, 21.3, and 25.6 × 10 −3 m) were prepared by adding NMP solvent accordingly in such a way that total volume of each MoS 2 solution became 60 µL. The RP perovskite-MoS 2 hybrids for a given n-value of perovskite was obtained by adding the RP perovskite crystals (0.5 m) in the MoS 2 solution (60 µL NMP) of fixed molarity (4.2-25.6 × 10 −3 m). The hybrid solution was sonicated for 15 min followed by stirring for 20 min at room temperature. Next, the hybrid solution was immediately used to fabricate thin/thick films.
Photodetector Fabrication: The vertical configuration photodetectors were prepared on fluorine-doped tin oxide (FTO) glass substrate (25 × 10 mm 2 , thickness ≈ 2.3 mm, ≈13 Ω sq −1 ). The middle portion of FTO was covered by the Kapton tape strip (3 mm width) and the remaining FTO was etched by using HCl acid and Zn dust. After etching, the FTO substrates were washed by sonication in acetone followed by IPA, and DI water for ≈30 min each. The required perovskite or hybrid solution was spin-coated over cleaned FTO substrate at ambient air conditions by running a two-step program: 1) 600 rpm for 15 s and 2) 3000 rpm for 30 s. The films were annealed at 50 °C for 30 min in ambient air and at 100 °C for 30 min inside an inert atmosphere (Ar) glove box to remove residual NMP. The Al metal electrodes of 100 nm thickness were deposited on the perovskite films by thermal vapor deposition technique using a shadow mask (slit width 3 mm) to form the vertical configuration photodetector (Al/RP perovskite-MoS 2 /FTO). All the photodetectors were made with a 9 mm 2 (3 × 3 mm 2 ) active area.
Characterization Techniques: SEM and EDAX were carried out using Zeiss EVO 18 Special Edition model. AFM was performed by using Park Systems XE-70 model. Thin film X-ray diffraction measurements were performed on Rigaku-Smart Lab XRD using Cu Kα (λ = 0.1542 nm, 25 kV, 40 mA) X-rays and diffraction angle step size of 0.2° in glancing angle mode. UV-vis absorption spectrums were obtained using Perkin Elmer Lambda 2S UV-vis spectrometer. PL measurements were performed on a confocal microscope (10×) coupled with DPSS 405 nm laser and spectrophotometer. TRPL measurements were performed on Horiba Delta Flex model using time-correlated single-photon counting method. The photocurrent measurements were carried out by using two-electrode potentiostat (BioLogic VMP3) system and for illumination DPSS laser 405 nm (≈630 mW cm −2 ), xenon lamp (Holmarc, Model Number: HO-ARC-XE150/300, 100 mW cm −2 ) with AM1.5 optical interference filter (Sciencetech, AM1.5G-FT-3) were used.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.