High-Performance Photodetectors Based on Lead-Free 2D

Subscriber access provided by UNIV OF OREGON

Functional Nanostructured Materials (including low-D carbon)

High-Performance Photodetectors Based on Lead-Free 2D Ruddlesden-Popper Perovskite/MoS2 Heterostructures Chen Fang, Haizhen Wang, Zixi Shen, Hongzhi Shen, Shuai Wang, Jiaqi Ma, Jun Wang, Hongmei Luo, and Dehui Li ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b20538 • Publication Date (Web): 31 Jan 2019 Downloaded from http://pubs.acs.org on February 5, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

High-Performance Photodetectors Based on Lead-Free 2D Ruddlesden−Popper Perovskite/MoS2 Heterostructures Chen Fang,1 Haizhen Wang,2 Zixi Shen,1 Hongzhi Shen,1 Shuai Wang,1 Jiaqi Ma,1 Jun Wang,1 Hongmei Luo2 and Dehui Li1* 1School

of Optical and Electronic Information and Wuhan National Laboratory for

Optoelectronics, Huazhong University of Science and Technology, Wuhan, 430074, China; 2Department

of Chemical and Materials Engineering, New Mexico State University, Las Cruces, NM 88003, United States

*Correspondence to: Email: [email protected].

Abstract Two-dimensional (2D) Ruddlesden−Popper Perovskites have attracted great interest for their promising applications in high-performance optoelectronic devices owing to their greatly tunable bandgap, layered characteristics and better environmental stability over three-dimensional (3D) perovskites. Here, we for the first time report on photodetectors based on few-layer MoS2 (n-type) and lead-free 2D perovskite (PEA)2SnI4 (p-type) heterostructures. The heterojunction device is capable of sensing light over the entire visible and near-infrared wavelength range with a tunable photoresponse peak. By using few-layer graphene flakes as electrical contact, the performance of the heterostructures can be improved with a responsivity of 1100 A/W at 3 V bias, fast response speed of ~40 ms under zero bias and excellent rectification ratio of 500. Importantly, the quantum efficiency can achieve 38.2% at zero bias, which are comparable or even higher than that of 3D perovskite/2D material photodetectors. Importantly, the spectral response peak of heterojunctions gradually shifts in a wide spectral range from band edge of MoS2 toward that of (PEA)2SnI4 with the external 1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 30

bias. We believe those 2D perovskite/2D material heterostructures with a great diversity represent an interesting system for investigating the fundamental optoelectronic properties and open up a new pathway towards 2D perovskite-based optoelectronic devices. Keywords: 2D perovskite, lead-free, transition-metal-dichalcogenide, graphene, heterostructure, photodetector Introduction Organometal halide perovskites such as CH3NH3PbX3 (X=Cl, Br, I) have been extensively studied in past few years mainly due to the rapid advancement of the power conversion efficiency of perovskite-based solar cells.1-4 Those perovskite materials within 3D frameworks have many advantages such as modest carrier mobility, long carrier diffusion length and strong light absorption in the visible range,5-7 which enable the certified power conversion efficiency of perovskite-based solar cells to more than 23% within a few years.8 In addition to the photovoltaic devices,9-11 the unique and excellent optical properties allow perovskites to be utilized in a wider range of optoelectronic devices such as photodetectors,12-19 lasers and light-emitting devices with decent performance.20,21 However, the lack of long-term stability and toxicity of lead in 3D perovskites are big challenges that severely hinder the further development of 3D perovskite-based optoelectronic devices.22-24 To solve the stability issue of 3D perovskites, 2D Ruddlesden-Popper perovskites with great environmental stability are emerging as an alternative solution and have attracted more and more attention recently.25-34 These 2D perovskites have a general 2


Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


chemical formula R2(A)n−1MX3n+1, where R is a long chain cation spacer, M is a metal cation, A is an organic cation, X is a halogen, and n represents the inorganic layer number of [MX6]2− octahedra sandwiched between the long chain cation spacer R.27,28,35 With such a unique structure, multi-quantum well structures are naturally formed, resulting in a larger exciton binding energy and the layer number n-dependent band gap due to the quantum confinement effect.26,28 Importantly, the presence of the hydrophobicity of the organic chain R would prevent [MX6]4- octahedral sheets from being directly contacted with moisture in ambient conditions, thus leading to a great environmental stability compared to their 3D counterparts.25,28 In addition, by changing either the layer number n and/or chemical composites, their band gap can be readily tuned from the near infrared to violet wavelength range, which would provide great flexibility for optoelectronic applications.26,36,37 In particular, the layered nature of 2D perovskites allows us to exfoliate thin flakes from their bulk counterparts to further integrate with other 2D layered materials to extend their functions. Up to date, 2D perovskite-based hysteresis-free solar cells have been demonstrated with greatly improved environmental stability compared to their 3D counterparts.25,31,35,38 Besides, lead-free perovskites such as (PEA)2SnI4 has also attracted great interests owing to their non-toxicity and superior carrier mobility.39-42 Due to their high carrier mobility and small exciton binding energy, 3D perovskites such as CH3NH3PbX3 (X = Cl, Br and I) have been considered as promising light‐sensitive materials.12 3D perovskite-based photodetectors have been demonstrated with excellent performance and might find important applications in the field of 3



Page 4 of 30

imaging, communication and biomedical sensing systems.43-45 Among various types of perovskite-based photodetectors, perovskite-based heterostructures can exhibit advantages in terms of lower noise, faster response speed, and larger specific detectivity over perovskite-based photoconductors and phototransistors, and can meet the demands of next‐generation photodetectors without an external power source.46 3D perovskitebased heterostructures have been fabricated with fairly well performance by integrating with 2D layered materials such as graphene (Gr) and/or transitional metal dichalcogenides (TMDs).46-52 Nevertheless, 3D perovskites are non-layered materials and the presence of the dangling bond would be detrimental to the efficient carrier transfer and thus the photodetection performance.27,36 To this end, 2D perovskites with layered nature are ideal candidates to form heterostructures with other 2D materials for high performance optoelectronic applications. Those heterostructures with the out-ofplane van der Waals bonding have no surface dangling bonds and thus atomically sharp interfaces, which are particularly important for high performance photodetectors with low noise.26,28 Furthermore, the van der Waals bonding in 2D perovskite/2D material heterostructures can eliminate the influence from lattice matching and deposition temperature, providing a significant advantage over 3D material-based heterostructures. Nevertheless, studies on the heterostructures based on 2D perovskites and other layered 2D materials are still in its infancy. Here, we for the first time report on photodetectors based on few-layer MoS2 (ntype) and lead-free 2D perovskite (PEA)2SnI4 (p-type) heterostructures. The lead-free 2D perovskite (PEA)2SnI4 flakes are exfoliated from the as-synthesized bulk single 4


Page 5 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


crystals and exhibit p-type behavior. The heterojunction device is capable of sensing light over the visible and near-infrared wavelength range with a tunable response peak. By using few-layer graphene flakes as electrical contact, the performance of the photodetectors can be further improved with a high photoresponsivity of 1100 A/W at 3 V bias and 121 mA/W at zero bias, corresponding to a high quantum efficiency of 38.2% at zero bias. In addition, the devices with graphene contact exhibit excellent rectification effect with a rectification ratio of ~500 and fast response speed with rise/fall times of 34/38 ms, which are comparable or even higher than that of 3D perovskite/2D material photodetectors. Results and Discussion The schematic of the crystal structure of (PEA)2SnI4 (PEA=C6H5C2H4NH3) is given in Figure 1a, which presents the [SnI6]4- octahedra layers are stacked between the long chain organic molecules via van der Waals interaction.39 Figure 1b shows the photograph of the as-synthesized (PEA)2SnI4 crystals. The (PEA)2SnI4 crystals with the lateral size ranging from hundreds of μm to few millimeters were synthesized by a solution method (see Experimental section). The black color of the crystals indicates that the bandgap of the 2D perovskites is close to near infrared range. After being exfoliated by mechanical method, the optical microscope (OM) image (Figure 1b) indicates the exfoliated flake has rather smooth surface, which suggests the high crystallinity of the exfoliated flakes and would be beneficial for stacking the vertical heterostructures. Powder X-ray diffraction (XRD) pattern in Figure 1c shows that all diffraction peaks of as-synthesized 2D perovskite crystals can be indexed to the 5



Page 6 of 30

orthorhombic phase of (PEA)2SnI4.39 The narrow full width at half maximum (FWHM) and high signal-to-noise ratio of XRD pattern suggest the excellent crystalline quality of the as-grown crystals. To explore the optical properties of the as-synthesized (PEA)2SnI4 crystals, photoluminescence (PL) and reflection measurement were carried out. As shown in Figure 1d, the pristine (PEA)2SnI4 sample exhibits one emission peak locating at 626 nm, which corresponds to the excitonic emission of (PEA)2SnI4.53 In terms of the reflection spectrum, the peak at 617 nm corresponds to the free exciton while the peak at 511 nm originates from the band-to-band transition. The peak below the excitonic emission might be due to the self-trapped excitons.54 Compared with the reflection spectrum, the exciton peak of PL spectrum shows a Stokes shift of ~9 nm, which is commonly observed in semiconductors. Figure 1e, f present the output and transfer characteristic curves of the asfabricated (PEA)2SnI4 flake device. The inset of Figure 1e displays the device configuration, which has a channel length of 20 μm fabricated on a 300 nm SiO2/Si substrate with 300 nm SiO2 as the dielectric layer. All the electrical measurements were carried out at room temperature in dark. It is evident that the current across the perovskite increases with increasing the negative gate voltage, suggesting a representative p-type semiconductor behavior. The transfer characteristic curves further confirm the p-type behavior of the as-synthesized samples. The huge hysteresis is present in the transfer curves (Figure S1), which might be due to the trap states since the ion migration has been suppressed in 2D perovskites, unlike 3D cases.55,56 This ptype behavior allows us to properly select other n-type 2D materials to construct vertical 6


Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


heterostructures based on (PEA)2SnI4 with a desired rectification ratio for optoelectronic applications. A few-layer MoS2 flake was used as n-type layer to fabricate (PEA)2SnI4/MoS2 heterostructures to investigate their optoelectronic properties. The detailed fabrication process and schematic illustration of the heterostructures are shown in Figure S2 and Figure 2a. (PEA)2SnI4 layer with a thickness of ~100 nm severs as the main light absorbent layer stacking on the top of a few-layer MoS2 flake while 5 nm Cr/50 nm Au works as electrodes. The formation of heterostructures is confirmed by the PL spectra (Figure 2b). The PL intensity is greatly quenched in the area of (PEA)2SnI4/MoS2 heterojunction in comparison to that in area of the perovskite alone, which can be attributed to the efficient separation of photogenerated electrons and holes due to the built-in electric field at the junction area.46 The I−V curves of a typical as-fabricated device both in dark and under a 451-nm illumination with various powers exhibit a clear current rectification behavior, suggesting the excellent diode characteristics in (PEA)2SnI4/MoS2 heterojunction with a rectification ratio of ~15. It is obvious that both forward current and reverse current increase gradually with increasing the light power due to the increased amount of photogenerated charge carriers. Based on previous UPS (Ultraviolet Photoelectron Spectroscopy) studies of (PEA)2SnI4 and MoS2,39,40,57 the energy band alignment of the heterostructure is shown in Figure 2d-f. Under zero bias, the diffusion current and drift current in the junction will be balanced each other in dark while a photocurrent was observed under illumination due to the presence of the built-in field (Figure 2d). Under forward bias, 7



Page 8 of 30

the applied bias would prompt the majority carrier to efficiently transport leading to a large diffusion current (Figure 2f). On the contrary, the reverse bias would block the majority carriers to transport to electrodes, and thus a weak drift current was observed by the minority carriers. As a result, the rectification behavior was observed in our (PEA)2SnI4/MoS2 heterostructures. Under illumination, the photogenerated carriers would contribute to the current either via diffusion current under forward bias or via drift current under reverse bias, thus resulting in the increased current. The as-fabricated device exhibits excellent photoresponse under a 451-nm light illumination (Figure 2c and g). We intentionally selected the 451-nm light to study photoresponse under different external bias and incident light power to avoid the complexity due to the shift of the spectral response peak under different external bias (see below and Figure 2i). As the power of the incident light increases, the photocurrent shows a gradual increase as well under a bias of 3 V. The responsivity, defined as the ratio of the collected current to the incident light power, excludes the influence of the incident light power and thus represents the true sensitivity of a photodetector. The estimated responsivity gradually decreases with the increase of the light power, which might be due to the increased carrier-carrier scattering or reduction of the carrier extraction efficiency from the heating effect.58,59 The maximum responsivity can reach 20 A/W under a bias of 3 V, which is smaller than that in 3D perovskite-based heterostructures (Table 1). The optical switch characteristics under the bias of 0 V, 0.25 V and 3 V reveal the excellent stability and reversibility of our (PEA)2SnI4/MoS2 devices (Figure 2h). The presence of reverse photocurrent at zero bias can verify the 8


Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


presence of the built-in field and an external quantum efficiency (EQE) is evaluated to be ~ 5%. The increase of the photocurrent with increasing bias can be ascribed to the bias induced change of the field strength within the depletion region. Figure 2i presents the spectral response of a typical as-fabricated device under the bias of -1 V, 1 V and 3 V. Similar to the reflection spectrum of (PEA)2SnI4 in Figure 1d, the responsivity spectra consist of a continuously increasing region above 600 nm corresponding to the band to band transition of (PEA)2SnI4 and two narrow peaks locating at 630 nm and 670 nm originating from the exciton peak of (PEA)2SnI4 and MoS2, respectively. Interestingly, the relative intensity of the two narrow peaks strongly depends on the applied external bias. With the increase of the bias, the intensity of the response peak at 630 nm increases much faster than that of 670 nm, which is probably due to the bias dependent carrier extraction efficiency. The majority of the incident light is absorbed by the top layer of (PEA)2SnI4 and the photogenerated carriers within (PEA)2SnI4 at the region out of the junction area can also contribute to the photocurrent. Under a small bias, the photogenerated electrons and holes at the region out of the junction area cannot be separated and quickly recombine, leading to the negligible contribution to the photocurrent. While under a larger bias, the photogenerated electrons and holes at the region out of the junction area can be separated and extracted to electrodes and thus contribute to the current. As a result, the response peak at 630 nm (exciton peak of (PEA)2SnI4) increases rapidly with the applied bias. To further improve the performance of the (PEA)2SnI4/MoS2 heterostructures, we constructed graphene/(PEA)2SnI4/MoS2/graphene devices by using few layer graphene 9



Page 10 of 30

flakes as electrical contact. As a zero-bandgap semimetal, few-layer graphene can easily make ohmic contact with Au electrode.60 Meanwhile, the layered nature of fewlayer graphene, (PEA)2SnI4 and MoS2 flakes allows them to form physically good contact among them, and thus would be beneficial for the photogenerated carriers to be extracted. Furthermore, the smaller work function of few-layer graphene compared to that of Au can also reduce the Schottky barrier between (PEA)2SnI4 or MoS2 and Au electrodes.

The

detailed

fabrication

processes

of

the

graphene/(PEA)2SnI4/MoS2/graphene device is shown in Figure S3. The few-layer graphene flakes were first dry transferred to the prefabricated Au electrodes followed by successively stacking the (PEA)2SnI4 and MoS2 flakes. Figure 3a presents the schematic illustration, optical image and scanning electron microscopy (SEM) image of the as-fabricated graphene/(PEA)2SnI4/MoS2/graphene device, where few-layer graphene, (PEA)2SnI4 and MoS2 flakes are highlighted by the dashed lines with different color. SEM image reveals that the (PEA)2SnI4 flake has very smooth surface after being transferred. The thickness of MoS2 and (PEA)2SnI4 are around 20 nm and 100 nm. Because the presence of the organic chain leads to the extreme large resistance along the out-of-plane direction compared with that of the in-plane direction,28,35 we intentionally contact the (PEA)2SnI4 flake by graphene from bottom side so that the photogenerated carriers can avoid crossing the entire (PEA)2SnI4 flake in the out-ofplane direction. The I-V curves of the graphene/(PEA)2SnI4/MoS2/graphene device in dark and under a 451-nm light illumination exhibit an excellent rectification behavior (Figure 3b 10


Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and c), similar to that of (PEA)2SnI4/MoS2 devices. Nevertheless, both dark current and light current were greatly enhanced after inserting the few-layer graphene, which would prompt the charge extraction. Remarkably, the rectification ratio of the graphene/(PEA)2SnI4/MoS2/graphene device can reach as large as 500, two orders of magnitude larger than that of 3D perovskite/2D material junctions (Table 1) and one order of magnitude larger than (PEA)2SnI4/MoS2 devices, because of the greatly increased forward current although the reverse current is also simultaneously enhanced. To

further

explore

the

optoelectronic

properties

of

the

as-fabricated

graphene/(PEA)2SnI4/MoS2/graphene device, we have carried out the optical switch characteristics

with

different

incident

power

at

zero

bias.

The

graphene/(PEA)2SnI4/MoS2/graphene device exhibits better photoresponse to the 451nm light illumination comparable to the (PEA)2SnI4/MoS2 device (Figure S6). Similarly, the photocurrent increases continuously with the incident light power at zero bias

(Figure

3d).

In

particularly,

the

on-off

ratio

of

the

graphene/(PEA)2SnI4/MoS2/graphene device can reach 102 with the incident power of 5.46 nW at 0 V bias, comparable to 3D perovskite/2D material junctions (Table 1). Similar to that of (PEA)2SnI4/MoS2, the responsivity and EQE of the graphene/(PEA)2SnI4/MoS2/graphene device continuously decreases with the increase of the incident power of the 451-nm light (Figure 3e). The maximum responsivity of 121 mA/W and EQE of 38.2% have been achieved under the 451-nm light illumination with a power of 36 pW, which is the highest value among the perovskite/2D material junction up to date (Table 1). By applying a bias, the similar optical switch 11



Page 12 of 30

characteristics and power dependent responsivity have been observed (Figure S4, S5). In addition, the response speed is a critical parameter for photodetectors, which is characterized by rise and fall time (deﬁned as the time taken for the photocurrent increasing from 10 % to 90 % of the peak value, and vice versa).46 The rise/fall time of the as-fabricated graphene/(PEA)2SnI4/MoS2/graphene device is estimated to be 34/38 ms (Figure 3f), which is similar to that of 3D perovskite/2D material junctions (Table 1). The comparison of the rise/fall time of the graphene/(PEA)2SnI4/MoS2/graphene device and the (PEA)2SnI4/MoS2 device is shown in Figure S7. After inserting the graphene layer, the response speed remains the same as the device without graphene layers, which is expected since the response speed should be limited by the (PEA)2SnI4/MoS2 junction. Similarly, the spectral response of our graphene/(PEA)2SnI4/MoS2/graphene device shows a continuous response band above 600 nm and two response peaks locating at 630 nm and 670 nm (Figure 4a). With the increasing the applied bias, the response peak at 630 nm increases much faster than that at 670 nm, which further confirms that this is due to the bias dependent carrier extraction efficiency, as discussed above. Remarkably, the maximum responsivity can be as large as 1100 A/W at a bias of 3 V, which is two orders of magnitude larger than that of (PEA)2SnI4/MoS2 devices and also larger than that of 3D perovskite/2D material junctions reported (Table 1). This great enhanced responsivity would be ascribed to the efficient carrier extraction and the formation of high quality interfaces among graphene, MoS2 and (PEA)2SnI4. Detectivity is a figure of merit used to characterize the detecting ability of a 12


Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


photodetector and thus allows to be compared among different devices.46,49 In order to extract the detectivity (D*) of the graphene/(PEA)2SnI4/MoS2/graphene device, we first determined the bandwidth of the device to be about 4 Hz under the bias of 0.5 V (Figure S8). The noise power density was then acquired by using a lock-in amplifier according to a previous reported method (Figure 4b).49,62 It is obvious that the 1/f noise, which might originate from the trapping and detrapping process of the large number of charges,63 dominates the noise of our device at the low frequency. At 4 Hz, the noise spectral density (Sn) of our graphene/(PEA)2SnI4/MoS2/graphene device is estimated to be 2.84×10-22 A2/Hz. The noise equivalent power (NEP) was calculated by NEP = (Sn)1/2/(RB1/2),41 where R is the responsivity and B is the bandwidth of the photodetector. The NEP spectrum of the graphene/(PEA)2SnI4/MoS2/graphene device at 0.5 V bias is given in Figure S9. Based on the measured NEP and bandwidth, the detectivity was calculated by D* = (AB)1/2/NEP.49 Figure 4c exhibits the D* spectrum of the as-fabricated graphene/(PEA)2SnI4/MoS2/graphene device under a bias of 0.5 V. The specific detectivity is on the order of 109 Jones from 450 to 800 nm, with the peak value of 8.09×109 Jones around 660 nm, which is slightly lower than that in 3D perovskite/MoS2 devices. We attribute the low detectivity to the high noise and the low device bandwidth caused by the defects at the surface and interfaces. Finally, we tested the repeatability of our graphene/(PEA)2SnI4/MoS2/graphene device. After continuous measurement for 1000 s, the device maintains the high on-off ratio and exhibits negligible current degradation, suggesting the excellent repeatability and good stability (Figure 4d). It is expected that with further optimizing the devices structures and 13



Page 14 of 30

lowering the noise level, the performance of our devices can be further improved. It should be noted that 2D organic-inorganic perovskites usually exhibit a great environmental stability compared to their 3D counterparts owing to the presence of the hydrophobicity of the organic chain spacers. However, organic-inorganic Sn-based perovskites have been proved to be not as stable as that of their Pb counterparts since Sn2+ can be easily oxidized to Sn4+ when exposed to the air .64 Our devices were based on exfoliated (PEA)2SnI4 microplates without any encapsulation layer. Thus the devices were not very stable under ambient conditions and the performance drops a lot after being stored in ambient for one day (Figure S10). Nevertheless, the Sn-based 2D perovskite devices are still more stable compared with Sn-based 3D perovskite devices. Both the synthesis of our Sn-based 2D perovskite crystals and fabrication of the Snbased 2D perovskite microplate heterostructures were carried out in ambient condition, which is impossible for Sn-based 3D perovskites due to the extremely fast degradation in air. We expect that our devices can survive for a longer time after being properly encapsulated. Conclusion In summary, we have successfully demonstrated high-performance photodetectors based on the lead-free 2D perovskite (PEA)2SnI4/MoS2 heterostructures. By using fewlayer graphene as electrodes, the graphene/(PEA)2SnI4/MoS2/graphene device exhibits a high responsivity of 1100 A/W, fast response speed of ~40 ms and a rectification ratio of 500 in the visible wavelength range. All performance parameters of our graphene/(PEA)2SnI4/MoS2/graphene device are comparable to or better than those of 14


Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3D perovskite/2D materials heterostructures. Our findings open up a way towards 2D lead-free perovskite-based heterostructures for the high-performance optoelectronic devices. Acknowledgements D. L. acknowledges the support from NSFC (61674060) and the Fundamental Research

Funds

for

the

Central

Universities,

HUST

(2017KFYXJJ030,

2017KFXKJC002, 2017KFXKJC003 and 2018KFYXKJC016); H. L. thanks the support from New Mexico EPSCoR with NSF-1301346. We thank Hong Cheng engineer in the Analytical and Testing Center of Huazhong University of Science and Technology for the support in PL measurement and thank the Center of MicroFabrication of WNLO for the support in device fabrication. Methods Material Synthesis. The PEAI (C6H5C2H4NH3I) powder was synthesized using a previously reported method.9 The (PEA)2SnI4 crystals for mechanical exfoliation were synthesized by a reported solution method.28 0.0688 g SnO powder, 0.75 ml hypophosphorous acid (H3PO2, 50 wt.% in water) solution and 2.7 ml hydriodic acid (HI, 57 wt.% in water) solution were mixed in a flask and heated to 150 °C under a constant magnetic stirring. Then 0.249 g PEAI powder was injected into the mixture to obtain (PEA)2SnI4 single crystals. After 10 minutes, the stirring was stopped and the solution was left to naturally cool to room temperature. Finally, the as-prepared perovskite plates were dried at 60 °C for 3 hours. Device Fabrication. For (PEA)2SnI4 device, the 5 nm Cr/50 nm Au electrodes on a 300 15



Page 16 of 30

nm SiO2/Si substrate were fabricated by photolithography and followed by thermal evaporation and lift-off process. The substrates were then treated by an oxygen plasma cleaner for 180 s. To make the (PEA)2SnI4 transistors, the (PEA)2SnI4 flake typically with a thickness of 100−120 nm was exfoliated onto a PDMS (polydimethylsiloxane) stamp. Then PDMS stamp with the exfoliated (PEA)2SnI4 flakes was brought to contact with Au electrodes under the aid of optical microscope and micromanipulators. The (PEA)2SnI4 flake was then released from PDMS stamp after waiting for 3 mins. For (PEA)2SnI4/MoS2 device, the few-layer MoS2 flake was first transferred onto one Au stripe of the two-probe electrodes by applying the same procedure to fabricate the (PEA)2SnI4 transistors. Then the (PEA)2SnI4 flake was transferred onto the other strip of the two-probe electrodes and simultaneously contacts with pre-transferred MoS2 flake to form the heterojunction. For graphene/(PEA)2SnI4/MoS2/graphene device, two few-layer graphene flakes were separately transferred onto each of the Au stripe of the two-probe electrode and then (PEA)2SnI4/MoS2 was transferred onto the graphene flakes by using same procedure as those described above. Material characterizations. XRD measurements were performed using a Bruker D2 PHPSER (Cu Kα λ=0.1542 nm, Nickel filter, 25 kV, 40 mA). Optical microscopy images were collected by an Olympus BX53M microscope. Scanning electron microscopy (SEM) images were acquired on the JEOL 7001F field emission scanning electron microscope. PL measurements were carried out on a micro-Raman spectrometer system (Horiba HR550) equipped with a 600 g/mm grating in a backscattering configuration excited by a 473 nm solid-state laser with a power of 1 16


Page 17 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


μW. The reflection measurement was performed in the same system illuminated by a halogen tungsten lamp with a power of 150 W. Photoconductivity measurements. The responsivity and photocurrent versus incident power were measured by coupling a 451-nm LED with a neutral density filter. By selecting filters with different OD values, we can tune the incident light power and record the photocurrent. For the spectral response measurement, a quartz tungsten halogen lamp (250 W) was dispersed by a monochromator (Horiba JY HR320) and the output monochromatic light beam was collimated onto the devices as the light source. The incident light power was recorded by a pyroelectric detector (Gentec, model APM (D)). The photocurrent was amplified by a low-noise amplifier (Stanford SR570) and subsequently read by a lock-in amplifier (Stanford SR830) equipped with a mechanical chopper (Stanford SR540). The noise current was measured with a lock-in amplifier (Standford SR830) in dark at room temperature. The response speed and bandwidth were measured by a digital oscilloscope (Tektronix MDO3032) coupled with a computer-controlled analogue-to-digital converter (National Instruments model 6030E). All measurements were performed under ambient conditions. Electrical measurements. The output characteristic curves and transfer characteristic curves of (PEA)2SnI4 transistors were measured in a commercial probe station (Lakeshore, PS100) coupled with a precision source/measurement unit (Agilent B2902A).

17



Page 18 of 30

Tables Table 1. Performance of photodetectors based on 3D perovskite/2D material heterostructures. Rise

EQE at

Detectivity

Rectification

time

0 V [%]

[Jones]

ratio

950 at 1 V

22 ms

-

-

High-Performance Photodetectors Based on Lead-Free 2D

Recommend Documents