† Corresponding author. E-mail:
In recent years, the rapid progress of metal halide perovskite solar cells has been witnessed by the rocketing power conversion efficiency. In addition, perovskites have opened up a great opportunity for high performance photodetectors (PDs), due to their attractive optical and electrical properties. This review summarizes the latest progress of perovskite-based PDs, aiming to give a comprehensive understanding of the material design and device engineering in perovskite PDs. To begin with, the performance parameters and device configurations of perovskite PDs are introduced, which are the basis for the next discussion. Next, various PDs based on perovskites in different morphologies are discussed from two aspects: the preparation method, and device performance. Then, several device engineering strategies to enhance the performance of perovskite-based PDs are highlighted, followed by the introduction of flexible and narrow-band perovskite PDs. Finally, key issues and major challenges of perovskite PDs that need to be addressed in the future are outlined.
Photodetectors (PDs), which operate by converting incident light into electrical signals, are the mainstay of a variety of industrial and scientific applications, such as imaging systems, optical communications, environmental surveillance, and biological sensing.[1–5] The semiconductor, as an active layer, is the essential part of the PD, and can absorb the incident photons and generate electron and hole pairs. Besides, a built-in or applied electric field is necessary to separate the electron and hole pairs, which are collected by electrons to produce an electric current. Up to now, a variety of semiconductors has been applied for conventional PDs, including Si, Ge, group II–VI (PbS and PbSe), group IIB–VI (Hg1−xCdxTe), and group III–V (GaN and InxGa1−xAsyP1−y). These PDs have been fully investigated and there has been impressive progress on the improvement of photodetection and device configuration design. However, there are several drawbacks, such as the complex manufacturing, high cost, and mechanical inflexibility, which limit broad applications of conventional PDs.
Low-temperature solution-processed semiconductors, such as organic semiconductors, nanomaterials, and quantum dots, are an emerging class of photoactive materials because they are compatible with the simple, low-cost, large area, and flexible manufacturing process.[5] Among them, metal halide perovskites have attracted wide research interests due to their direct bandgaps, long carrier diffusion lengths, high absorption coefficients, and high defect tolerance.[6–10] These outstanding features make metal halide perovskites widely applied in solar cells.[11–14] As a result, the power conversion efficiency of perovskite solar cells has risen from 3.8% to 23.3% in the past few years.[15,16] Based on these significant improvements, the perovskite film formation, chemical composition, and hole/electron transport materials have been amply developed.[17,18] In addition, metal halide perovskites also have great potential in many other optoelectronics, such as light-emitting diodes (LEDs), lasers, and PDs.[19–25] Particularly, there already have been plenty of research studies about perovskite PDs in material compositions and morphologies, as well as device architectures and engineering. Various strategies have been developed to synthesize perovskites in different morphologies (polycrystalline films, single crystals, and low-dimensional nanostructures) for PDs. Meanwhile, device-engineering strategies are applied to improve the performance of the perovskite PDs. Therefore, it is necessary to give a comprehensive review of the perovskite PDs in material compositions and device engineering.
In this review, we first introduce the performance parameters that influence the performance of perovskite PDs, the configurations of which can be divided into three types: photodiodes, photoconductors, and phototransistors. Next, we mainly discuss PDs based on perovskites in different morphologies, including polycrystalline films, single crystals, nanowires/nanorods, quantum dots/nanocrystals, and two-dimensional (2D)/quasi-2D perovskites. Then, several strategies to fabricate high-performance or special-purpose perovskite PDs are illustrated and analyzed. Finally, we provide a brief conclusion and outlook about both the achievement and challenges of perovskite PDs.
There are several parameters used to evaluate the performance of perovskite PDs, including responsivity R, detectivity D *, linear dynamic range (LDR), and response time (rise/decay time). A brief overview of these parameters is presented as follows.
The responsivity is defined as the ratio of photocurrent to incident-light intensity, which measures the electrical output per optical input. It is given by
![]() |
![]() |
Another parameter closely related to responsivity is photoconductive gain G, which is defined as the average number of circuit electrons generated per photocarrier pair.[26] The device responsivity can be calculated from photoconductive gain
![]() |
![](cpb_28_1_018502/cpb_28_1_018502_ieqn1.gif)
![]() |
The specific detectivity D * characterizes the weakest level light that can be detected by PDs, which is determined by the responsivity and noise of PDs. The D* is given by
![]() |
![]() |
Another parameter characterizes the noise level of the detector, and the detectorʼs ability to detect weak light signals is known as the noise equivalent power (NEP), which is defined as the signal incident light power on the PD when the signal-to-noise ratio S/N=1 is detected (the signal-to-noise ratio is the ratio of the effective value of the signal to the effective value of the noise). Since the noise level is proportional to the root of the measurement bandwidth Δ f, the NEP specifies measurement results at a 1 Hz bandwidth. Consequently, the NEP is equal to the reciprocal of D *:
![]() |
One important characteristic for PDs is the capacity to perform an identical responsivity over a wide range of light intensity, which is known as the LDR:
![]() |
![](cpb_28_1_018502/cpb_28_1_018502_ieqn2.gif)
![](cpb_28_1_018502/cpb_28_1_018502_ieqn3.gif)
PDs are not only evaluated by sensitivity, but also by speed. The response time is characterized by rise time τr and fall time τf of its response to the optical signal. Generally, the rise time and fall time are the times taken for the photocurrent to rise to 63.2% and drop to 36.8% of the steady-state values, respectively.[25,28] However, when PDs are modulated in a pulsed case, the rise time is defined as the time to transit between 10% to 90% of the peak output value, and vice versa for the fall time.[29–32]
Generally, there are three main types of configurations for perovskite PDs: photodiodes, photoconductors, and phototransistors.[25,33–37] According to the spatial layout of the photoactive medium and electrodes, perovskite PDs can be divided into two major categories: the lateral type, and the vertical type (Fig.
![]() | Fig. 1. (a) Categories of perovskite PD configurations. (b) Schematic diagram of photoconductor. (c) Schematic diagram of p–i–n photoconductor. |
A photodiode is a p–n junction that has a depleted semiconductor region with a high electric field that serves to separate photogenerated electron–hole pairs. Photodiodes derived from a p–n photodiode include, but are not limited to, the following: p–i–n photodiodes (Fig.
![]() | Table 1.
The advantages and disadvantages of different device configurations. . |
A photoconductor is based on the photoconductive effect, in which an electric field applied to the material by an external voltage source causes the electrons and holes to be transported.[26] In comparison to the photodiode, the perovskite photoconductor has a simple configuration and an easy process (Fig.
![]() | Fig. 2. Different configurations of perovskite PDs, including (a) a conventional photodiode,[34] (b) an inverted photodiode,[38] (c) a lateral photoconductor,[47] (d) a vertical photoconductor,[48] (e) a phototransistor,[25] and (f) a phototransistor with a hybrid functional layer.[45] |
As mentioned above, the lateral structure photoconductor is relatively facile to fabricate. However, it usually has a large dark current and a relatively slow response, and thus the phototransistor is designed with a gate electrode to modulate the charge transport. Phototransistors usually have a three-terminal configuration, which is similar to field-effect transistors. The gate electrode is isolated from the perovskite layer by a thin dielectric film, and thus the charge transport can be modulated by applying a gate voltage (Fig.
Generally, halide perovskites have the formula ABX3, in which A and B are cations, and X represents halogen anion (Cl−, Br−, and I−) (Fig.
![]() | Fig. 3. (a) The structure of perovskites with formula ABX3, in which A represents cations (MA, FA, and Cs+), B represents metal cations (Pb2+ and Sn2+), and X represents halogen anions (Cl−, Br−, and I−).[75] (b) Schematic structures of 2D (RNH3)2[ABX3]nBX4 perovskites (n = 0, 1, and 2) and 3D ABX3 perovskite (![]() ![]() ![]() |
The thermal and structural instability of organic cation lead perovskite is an urgent problem that needs to be solved. Three-dimensional (3D) MA-based perovskites suffer from intrinsic instability due to the relatively volatile MA cations that are released from the film during heating, and this process is accelerated in the presence of moisture. In addition, MAPbI3 was found to be prone to degradation following exposure to moisture, heat, oxygen, and light.[55–58] It was previously reported that FA-based perovskites exhibited much greater thermal stability than MA-based perovskites.[59] Nevertheless, the black perovskite phase of pure FAPbI3 is thermodynamically stable only above 160 °C.[60] Although all-inorganic perovskite (CsPbI3) has better thermal stability, the most stable phase at room temperature is not a black phase with high photoelectric conversion efficiency. Notably, recent studies have revealed that mixed perovskites showed improved stability and device performance. The MA/FA mixture demonstrates that a small amount of MA induces a preferable crystallization of FA perovskite into its photoactive black phase, resulting in a more thermally and structurally stable composition than pure MA or FA compounds.[60,61] Furthermore, by substituting FA with Cs cations, FAPbI3 can be stabilized in the desired black α-phase to form a thermally and structurally stable perovskite.[62,63] Besides, the band gap and the optical absorption band border can be modulated by the substitution of A cations since the band gap changes from 1.48 eV for FAPbI3 to 1.73 eV for CsPbI3.[64] Although the band gap can also be modulated by the ratio of halogens, the range of the adjustment is limited considering the structural stability of perovskite. By substituting FA with Cs cations, the region of structural instability in the Br–I phase space could be pushed to higher energies, and thus potentially achieve a structurally stable mixed-halide perovskite with a band gap of 1.75 eV.[59] For B cations, the bivalent lead ion is usually the best choice to attain high photoelectric conversion efficiency despite the lead being toxic. Sn, Ge, Bi, Cu, and many other elements have been used for replacing Pb in order to avoid the toxicity to the human body and environment.[65–70]
Different from the 3D perovskites mentioned above, layered perovskites have a general formula of (RNH3)2[ABX3]nBX4, where R denotes a long chain alkyl or aromatic group (such as aliphatic or aromatic alkyl ammonium), and n is relevant to the number of repeated structural units that contain the A cation[71] (Fig.
As mentioned in Section
![]() | Table 2.
Materials and key parameters of perovskite PDs. . |
Polycrystalline film is one of the most common types of perovskite materials. There are many universal strategies to form polycrystalline perovskite films, such as thermal annealing, solvent annealing, atmospheric control, solvent engineering, chemical vapor deposition, vacuum flash-assisted solution processing, and the use of additives.[76–84] These strategies have been applied successfully in fabricating not only high-efficiency perovskite solar cells, but also perovskite film PDs. The preparation processes of perovskite polycrystalline films are simple and compatible with different configurations of perovskite PDs, including photoconductors, photodiodes, and phototransistors. Xie and co-workers proposed the first organic–inorganic hybrid perovskite photoconductor, the configuration of which is shown in Fig.
![]() | Fig. 4. (a) Fabrication procedure for perovskite thin film using the vapor–solution fabrication method on a patterned ITO/PET substrate. (b) Photograph of a flexible perovskite PD in the bending situation (with interdigital electrodes on the PET substrate). (c) Wavelength-dependent EQE of the flexible PD at different voltages. (d) Photocurrent and EQE of the flexible PD as a function of repetitive bending cycles at 1 V.[84] (e) Energy diagram of the perovskite PD under a slight reverse bias. (f) Current density–voltage curves of PDs with and without the hole-blocking layer. PD1: without hole-blocking layers; PD2 and PD3: with BCP and PFN as the hole-blocking layers, respectively. (g) Transient photocurrent response at a pulse frequency of 100 kHz with a device area of 0.1 cm2 (blue line) and 0.01 cm2 (red line). Transient photocurrent response of a silicon diode is shown for comparison.[38] (h) Transfer characteristics of a perovskite-based phototransistor in the dark (black and red symbols) and under light illumination (blue and magenta symbols). (i) EQE and R of the hybrid perovskite PD at different wavelengths and a VDS value of −30 V. (j) Temporal photocurrent responses highlighting a rise time of ![]() ![]() |
As for photodiodes, Yang and co-workers presented a solution-processed high-performance PD based on organic–inorganic hybrid CH3NH3PbI3−xClx film using an “inverted” device configuration,[38] the energy diagram of which is shown in Fig.
Although the configuration of the phototransistor is different from that of the photoconductor, the polycrystalline perovskites applied are similar. Wu and co-workers fabricated a phototransistor based on polycrystalline perovskite (Fig.
Compared with polycrystalline perovskites, single-crystal perovskites possess many unique properties: high purity, few grain boundaries, and enhanced thermal and moisture stabilities. There are several reported methods for synthesizing perovskite single crystals: inverse temperature crystallization (ITC),[48,92–95] antisolvent vapor-assisted crystallization,[96] top-seed solution growth,[97] bottom-seeded solution growth,[98] and solvent acidolysis crystallization.[99] Among of them, ITC is the simplest and most commonly used approach. It is suitable to use the ITC method when the perovskite exhibits inverse temperature solubility behavior in certain solvents. MAPbX3 perovskites exhibit fascinating inverse solubility trends in solvents like γ-butyrolactone (GBL), N,N-dimethylformamide (DMF), dimethylsulphoxide (DMSO), or their mixtures.[92] As shown in Fig.
![]() | Fig. 5. (a) Schematic of the ITC apparatus and MAPbI3 crystal growth at different time intervals.[92] (b) Device structure of a single-crystal perovskite PD. (c) EQE spectra of a 0.3 mm thick MAPbBr3 single crystal under different reverse biases. (d) EQE spectra of MAPbBr3 single crystals with different thicknesses cleaved from the same large single crystal, measured under a −4 V bias.[36] (e) Photograph of a 27 × 27 photoconductor array fabricated on a 34 × 38 mm2-sized MAPbBr3 single crystal. (f) EQE and detectivity of the MAPbBr3 single crystal photoconductor illuminated by monochromatic light with its wavelength changing from 325 nm to 650 nm at a 4 V bias.[102] |
There have been extensive researches about one-dimensional (1D) nanowire materials, such as carbon nanotubes, metal nanowires, metal oxide nanowires, and polymer nanowires, due to their high aspect ratio in conjunction with superior mechanical, electronic, and optical properties.[109,110] Hence, it is desirable to fabricate perovskite nanowires based optoelectronics. Generally, perovskite nanowires are prepared by the solution-phase synthesis method.[111–114] Xue and co-workers demonstrated a simple and effective method of hot-injection for synthesis of CsPb(Br/I)3 nanorods with a uniform rod-shaped structure,[112] the synthesis process of which is shown in Fig.
![]() | Fig. 6. (a) Schematic diagram of the synthesis process for CsPb(Br/I)3 nanorods. (b) Schematic diagram of the PD based on CsPb(Br/I)3 nanorods.[112] (c) Schematic description of the aligned nanowire growth. (d) Optical images of nanowires grown by a modified EISA method.[119] (e) Schematic illustration of the fluid-guided antisolvent vapor-assisted crystallization (FGAVC) method for the fabrication of the MAPbI3 nanowire array.[123] |
Based on the simple solution-phase method, there are several strategies to boost the performance of perovskite nanowires, including the self-assembly method and template-assisted method. Song and co-workers used the one-step self-assembly method to grow CH3NH3PbI3 nanowires,[118] and further applied an evaporation-induced self-assembly (EISA) method to prepare high-ordered perovskite nanowires.[119] In order to obtain a large-area aligned nanowire distribution, the substrates were tilted for a small angle of less than 15° to guide the nanowire alignment (Figs.
Besides the commonly used self-assembly method, the template-assisted method opened up another avenue for high-performance perovskite nanowire PDs. Zhang and co-workers demonstrated a facile method for preparing porous MAPbBr3 nanowires by a self-template-directed reaction of Pb-containing nanowires with HBr and MABr at room temperature in solution.[122] Jie and co-workers reported a facile FGAVC method for large-scale fabrication of high-quality single-crystalline perovskite nanowire arrays.[123] As shown in Fig.
When the size of 3D materials is decreased to the nanometric scale, the tiny spherical particles will exhibit completely different optical and electronic properties compared with their bulk counterparts. Their size-dependent optical band gaps, size tunability, large extinction coefficient, high photoluminescence quantum yield (PLQY), and multiple exciton generation characteristics of quantum dots make them especially suitable for use in optoelectronic devices.[124–128] Perovskite quantum dots/nanocrystals also exhibit quantum-size effects similar to conventional quantum dots, and thus they can be applied in optoelectronic devices.
Generally, two methods have been widely used for the preparation of perovskite quantum dots/nanocrystals: (i) the ligand-assisted re-precipitation (LARP) method[129,130] and (ii) the hot-injection method.[99,131,132] Pérez-Prieto and co-workers first demonstrated solution-phase synthesis of perovskite colloidal quantum dots, in which the MAPbBr3 quantum dots with a size of 6 nm were prepared via re-precipitation using medium-length alkyl chain organic ammonium cations as ligands.[133] Dong and co-workers also reported a simple and versatile LARP technique for the fabrication of MAPbX3 QDs with absolute PLQYs of 50%–70%.[130] The process for the LARP technique and reaction system is illustrated in Fig.
![]() | Fig. 7. (a) Schematic illustration of the reaction system and process for the LARP technique. The starting materials in the precursor solution (top right) and the typical optical image of the colloidal MAPbBr3 solution (bottom right) are illustrated.[130] (b) The reversible anion exchange reaction of MAPbX3 nanocrystals. (c) I–V curves of MAPbBr3 film under 365 nm and 505 nm irradiation (![]() ![]() |
According to the results of early research work, perovskite quantum dots/nanocrystals were most applied in LEDs.[130,131,134] However, more and more studies suggested that PDs based on perovskite quantum dots/nanocrystals also have good performance.[135–139] Early in 2013, Wang and co-workers reported a CsPbBr3 quantum dot PD, the performance of which was enhanced by employing a bilayer composite film of mesoporous TiO2 (mp-TiO2) CsPbBr3 quantum dots as photosensitizers.[135] As a result, this PD demonstrated a significantly improved on/off ratio by about three orders of magnitude. Song and co-workers synthesized MAPbBr3 plate-type nanocrystals by a novel solution reaction using octylamine as the capping ligand.[136] The synthesis of MAPbBr3−xClx and MAPbBr3−xIx nanocrystals was achieved by the halide exchange reaction of MAPbBr3 with MACl and MAI in an isopropyl alcohol solution, respectively, and the nanocrystals with different chemical compositions demonstrated full-range band gap tuning over a wide range (1.6–3 eV) (Fig.
Since the long-term stability became a major concern for perovskite nanocrystalline applications, all-inorganic perovskite nanocrystals gradually showed advantages owing to their superior stability against oxygen and moisture. Lee and co-workers presented PDs based on all-inorganic CsPbX3 nanocrystals that were synthesized by a hot-injection method.[140] Besides, the CsPbCl3 and CsPbI3 were acquired by halide exchange reactions using lithium salts (LiX, X=Cl, Br, and I). PDs made of CsPbI3 nanocrystals exhibited a high on/off photocurrent ratio of 105, as well as rise and decay times of 24 ms and 29 ms, respectively. The same group also demonstrated a high-sensitivity hybrid PD based on graphene–CsPbBr3−xIx perovskite nanocrystals, which exhibits a high photo-responsivity of
In recent years, 2D materials have been extensively investigated to be applied in PDs due to their excellent optical, electrical, and mechanical properties. Hence, it is desirable to reduce the dimensionality of bulk perovskite to 2D and investigate the optical–electrical characteristic of 2D perovskite-based devices. Currently, 2D perovskite materials can be divided into two types: the non-van der Waals type, and the van der Waals type. The former is derived from a 3D perovskite framework by cutting down the thickness to a single or few unit cells, such as nanoplates,[141] nanosheets,[142–145] and nanoflakes.[146] The latter is built by inserting long-chain organic cations in the “A” position to block the interaction of inorganic [BX6]4− bilayers, mainly including those quasi-2D perovskites with a formula of (RNH3)2[ABX3]nBX4 (
For non-van der Waals-type 2D perovskites, there are several ways to realize the synthesis process, including the simple solution-phase method, the solution process combined with the vapor-phase conversion method, and the seed-mediated solution growth method. Bao and co-workers reported a combined solution process and vapor-phase conversion method to synthesize 2D hybrid organic–inorganic perovskite (Fig.
![]() | Fig. 8. (a) Schematic illustrations of solution process to fabricate 2D PbI2 nanosheets, vapor-phase conversion process to transfer PbI2 to 2D MAPbI3 perovskite nanosheets, and the crystal structure of single unit cell thick 2D MAPbI3 perovskite. (b) Atomic force microscope (AFM) topography images of 2D MAPbI3 nanosheets. (c) Schematic image of a transistor device based on 2D perovskite. (d) I–V curves of the 2D perovskite-based device under different light intensities.[145] (e) High-magnification SEM images of 2D ultrathin CsPbBr3 nanosheets. (f) AFM image and height profile of monolayer and bilayer CsPbBr3 nanosheets with a thickness of 1.6 nm and 3.3 ± 0.2 nm, respectively. (g) I–V curves of the PD in the dark and under 442 nm light (![]() |
Song and co-workers reported high-performance flexible visible light PDs based on 2D CsPbBr3 nanosheets.[143] The high-quality and easily dispersed CsPbBr3 nanosheets (Fig.
Van der Waals-type 2D perovskites, namely, the quasi-2D perovskite, can be synthesized by non-colloidal and colloidal methods.[71] Most of the non-colloidal methods, such as the liquid-phase crystallization method[152] and direct solution-phase growth,[153], produce small crystals in solution, which are not suitable for large-area optoelectronics requiring thickness control and complete surface coverage. On the contrary, the colloidal method is easy and has a low cost associated with preparing large-area thin films. Huang and co-workers reported PDs based on 2D layer-structured perovskite with a formula of (C4H9NH3)2(CH3NH3)n−1PbnI3n+1 (n = 1, 2, and 3).[154] The perovskite film of three different 2D layer-structured perovskite materials with different layer numbers (n = 1, 2, and 3) were fabricated by spin-coating followed by thermal annealing. The structures of the layered perovskite and PDs are illustrated in Fig.
Although there are various perovskite materials for PDs, it is not enough to meet the requirements of high-performance, diverse applications, and high stability, since the different types of perovskites mentioned above have their respective shortcomings. Hence, besides the perovskite materials, device engineering of the PDs is crucial. There are highly developed strategies for perovskite PDs, such as hybridizing with other materials and interface/band gap engineering. In addition, devices with special functions, such as flexible and narrow-band perovskite PDs also enlarge the application field of perovskites.
It is a common approach to improve the performance of 2D-material PDs using hybrid systems.[155,156] In addition, the hybrid structure of perovskites has been widely applied in solar cells. Therefore, it is rational to employ the hybrid structure to boost the performance of perovskite PDs. There are various materials for hybrid perovskite PDs, such as 2D materials, polymers, metal oxides, and quantum dots.
In recent years, conventional 2D materials, which have excellent optical, electric, thermal, and mechanical properties, have been widely explored as functional layers for PDs.[155] Accordingly, many 2D materials have been studied as functional layers in PDs hybridized with perovskites. These 2D materials include, but are not limited to, graphene and its derivatives,[37,45,138,157,158] transition-metal dichalcogenides (TMDs) (such as MoS2 and WS2),[159–161] and h-BN.[162]
Cho and co-workers first fabricated a novel phototransistor consisting of graphene/MAPbI3 hybrid layers.[45] The schematic diagram of the phototransistor is illustrated in Fig.
For TMDs, Wu and co-workers fabricated high-performance PDs based on perovskite MAPbI3/WS2 hybrid layers for the first time.[159] The schematic device structure of the PDs is shown in Fig.
![]() | Fig. 9. (a) Schematic device structure of the hybrid WS2/perovskite photoconductor. (b) Bias dependence of the on/off ratios measured on the two samples with and without the WS2 layer.[159] (c) Molecular structures and schematic diagram of MAPbI3/C8BTBT hybrid PDs. (d) Responsivity of MAPbI3 and MAPbI3/C8BTBT hybrid PDs as a function of the light wavelength.[167] |
Organic semiconductors have been widely used in perovskite photodiodes as the hole/electron transport layer. Generally, the charge transport layers in perovskite photodiodes are used to modify the energy levels and reduce the dark currents.[24,90,168] Burn and co-workers demonstrated a low noise, IR-blind visible perovskite photodiode with performance metrics equivalent to commercially available silicon photodiodes.[24] The [6,6]-phenyl-C61-butyric acid methyl ester (PC60BM)/C60 cathode interlayer has three functions: (i) improving the diode temporal response; (ii) suppressing the dark current; and (iii) controlling the spectral response. Hence, the performance of the photodiode was greatly enhanced without compromising the LDR. Similarly, hole transport layers can also be employed to reduce the dark current and control the spectral response.[90]
Besides being charge transport layers in photodiodes, organic semiconductors are widely applied in photoconductors. PCBM was applied to perovskite hybrids in order to fabricate performance-enhanced photoconductors or phototransistors via enhancing the charge transfer.[169,170] Yang and co-workers fabricated a flexible and air-stable perovskite photoconductor based on the MAPbI3/dioctylbenzothieno [2,3-b] benzothiophene (C8BTBT) bulk heterojunction;[167] the schematic device structure is exhibited in Fig.
Metal oxides have been widely applied in perovskite optoelectronic devices as charge transport layers due to their suitable energy levels and high mobility.[21,173] Metal oxides can also be applied in perovskite PDs for the following reasons: (i) providing a mesoporous scaffold for perovskites; (ii) providing a high efficient charge transfer layer; (iii) band gap engineering; (iv) providing surface passivation; and (v) providing a blocking layer to reduce the dark current. Mesoporous TiO2 was not only applied in perovskite to provide a mesoporous scaffold for perovskites, but also in perovskite photodiodes. Tress and co-workers fabricated high-gain, low-voltage MAPbI3 PDs based on FTO/TiO2 MP scaffold/MAPbI3/Spiro-MeOTAD HTM/Au (FTO: fluorine-doped tin oxide; MP: mesoporous; HTM: hole transporting material), which showed the highest responsivity of
Recently, amorphous indium gallium zinc oxide (IGZO) has become an attractive channel material for organic–inorganic hybrid phototransistors.[180] The advantageous features of IGZO, such as high field-effect carrier mobility and low-temperature large-area processing capabilities, make it possible to apply it in perovskite transistors. With an optical bandgap of about 3 eV, amorphous IGZO is ready to sense ultraviolet light with wavelengths shorter than 420 nm. Zhou and co-workers developed a hybrid phototransistor where IGZO was capped with a solution-processed perovskite layer.[181] The device structure of the hybrid phototransistor is depicted in Fig.
![]() | Fig. 10. (a) Schematic of MAPbI3/IGZO hybrid phototransistors. (b) Responsivities of IGZO transistors and MAPbI3/IGZO phototransistors at VGS=−10 V.[181] (c) Schematic of perovskite/PbSe hybrid PD. (d) Responsivity and detectivity of the device as a function of the wavelength ranging from 300 nm to 1500 nm at bias voltage VDS=2 V and drain source voltage VGS=0 V.[46] |
The perovskite quantum dots have been discussed above; however, there are various materials that can be processed into CQDs, such as graphene, PbS, PbSe, HgTe, and CdSe/ZnS. They have been widely applied in PDs due to their ability to limit spectral absorption and offering a wide range of bandgaps.[32,127,128,182–184] It is advisable to combine the perovskite with CQDs of other materials to enhance the performance of PDs. Gong and co-workers fabricated MAPbI3/PbS QDs hybrid perovskite photodiodes, in which PbS QDs induced an additional trap-state onto the surface of the MAPbI3 film, thus forming ohmic contact at the MAPbI3/aluminum (Al) interface.[185] As a result, the photodiodes based on hybrid perovskite achieved an EQE of ∼4500%, a responsivity of
Layers are stacked together in a vertical perovskite device, producing several interfaces and interfacial regions. Hence, the interfaces and band alignment have significant influences on the charge processes in addition to the qualities of the layers themselves. As a result, interface engineering and band alignment engineering have become effective ways to promote the performance of perovskite PDs, especially the photodiodes.
Sargent and co-workers produced a thin, pinhole-free compact layer of Al2O3 on the nanostructured TiO2 to passivate the surface trap states, and further added a thin layer of PCBM to minimize the dark current.[33] Zhang and co-workers introduced interface engineering to improve the responsivity and response speed of the perovskite photodiode by replacing the TiO2 transport layer with a SnO2 layer, and reduced the dark current by adding an inter-layer of poly(vinylpyrrolidone) between the electron transport layer and perovskite layer. The SnO2 layer has a lower conduction band and lower trap density when compared with TiO2 layer, thus leading to a better band alignment and a more efficient carrier transport.[34] Besides, organic semiconductors and 2D materials were also involved in the interface/band gap engineering of perovskite PDs. Huang and co-workers fabricated high-performance perovskite PDs with ingenious device design, in which the
![]() | Fig. 11. (a) Device structure of the MAPbI3 PD. (b) Photocurrent and dark current of the MAPbI3, and the photocurrent of the reference photovoltaic device, which has a structure of ITO/PEDOT:PSS/MAPbI3/PCBM/C60/BCP/Al. (c) Energy diagram of the MAPbI3 PD at reverse bias under illumination.[31] (d) Device structure of perovskite–graphene hybrid PDs. (e) The responsivity and EQE versus light intensity relationships for the graphene–perovskite and graphene–perovskite–gold nanoparticle (Au-NP) devices. (f) Schematic of generation, diffusion, and transfer of photo-induced carriers in the perovskite layer with and without the influence of Au-NPs.[37] |
Sum and co-workers introduced a vacancy-engineered MoS2 monolayer to achieve ultrafast hole transfer at the interface of MoS2 and perovskite.[187] It was found that creating S vacancies in MoS2 strongly improves the hole transfer rate between MAPbI3 and MoS2 layers. Importantly, this work highlights the feasibility of applying defect engineering in tuning the band alignment of 2D transition metal dichalcogenides (TMDCs)–perovskite hybrid PDs. Sun and co-workers inserted Au-NPs into MAPbI3/graphene hybrid PDs to achieve a significant responsivity enhancement.[37] The schematic of the device structure is illustrated in Fig.
Flexible PDs have been widely applied in wearable devices, stretchable displays, and electronic skins due to their reduced cost and light weight.[188–190] Fortunately, perovskites can be prepared by solution processing, which renders the possibility for them to be assembled into flexible PDs. In addition to the perovskite itself being compatible with flexible PDs, we have discussed several strategies to fabricate a flexible PD based on perovskites, including reducing the dimensions of bulk perovskite and integrating them with 2D materials and organic semiconductors. As we mentioned in Section
![]() | Fig. 12. (a) Morphologies of perovskite crystals networks. (b) Schematic of a PD (top), and photograph of network PD arrays (bottom). (c) I–t curve evolution at different bending cycles (0–10000 times) at a fixed angle of 40°.[118] (d) Schematic illustration of fabricating the flexible MAPbI3/PDPP3T PD. (e) UV–vis–IR absorption spectra of MAPbI3 and MAPbI3/PDPP3T composite films. (f) Responsivity versus bending cycle number at a radius of 7 mm under 650 nm and (g) 835 nm.[171] |
For hybrid perovskite flexible PDs, the perovskite was usually hybridized with organic semiconductors, such as C8BTBT, PDPP3T, PCBM, single-walled carbon nanotubes (SCNTs), and rhodamine B.[167,171,172,191,192] Shi and co-workers reported a flexible broad-band PD based on a bilayer composite film of MAPbI3/PDPP3T.[171] The fabrication process is illustrated in Fig.
Narrowband PDs used for spectrally selective photodetection are widely required in imaging. Spectral discrimination in state-of-the-art narrowband PDs is usually realized by one of several approaches: (i) by combining broadband PDs with bandpass filters,[193] (ii) by using photoactive materials with narrowband absorption,[194] (iii) by intentionally enhancing absorption in a particular wavelength range via the plasmonic effect,[195] and (iv) by splitting the light into its component colors.[196] Perovskites have good prospects for application in narrowband PDs due to the controllable band gap, large carrier diffusion length, and large absorption coefficient. These unique properties also extend the strategies to carry out the device engineering for narrowband perovskite PDs.
The vertical PDs based on single-crystal perovskite are well suited for narrowband PDs, which has been discussed in Section
![]() | Fig. 13. (a) Schematic of the working principles of CCN photodiodes. (b) Working mechanism and performance of the red narrowband CCN photodiodes. (c) EQEs at −0.5 V of the red narrowband photodiodes fabricated with different ratios of PbI2 and PbBr2.[101] (d) Schematic showing a narrowband PD with a thick active layer and integrated perovskite layer. (e) Device structures of narrowband PDs with an integrated perovskite layer (left) and thick active layer (right). (f) EQE spectra of devices with a thick active layer and devices with an integrated perovskite-1 filter layer with different mixed-halides ratios.[199] |
Based on the common practice of combining broadband PDs with bandpass filters to achieve narrowband detection, Huang and co-workers demonstrated a new method to tune the spectral response band of perovskite PDs with perovskite filters.[199] They prepared two types of devices, the working principle schematic diagram of which is exhibited in Fig.
In summary, we introduce and discuss perovskite PDs in terms of chemical compositions, material morphologies, device architectures, and engineering strategies. The use of perovskites in PDs is compelling due to their unique electric and optical properties, simple and low-cost processing, and mechanical flexibility. There have been various strategies for synthesizing perovskites, such as solution processing, the EISA method, and chemical vapor deposition. Several types of perovskite materials for PDs have been demonstrated, including polycrystalline films, single crystals, nanowires/nanorods, quantum dots/nanocrystals, and 2D/quasi-2D perovskites. PDs based on various perovskite materials possess very different merits regarding photoelectric performance. The key parameters of perovskite PDs are listed and compared in Table
In terms of the device engineering of perovskite PDs, the core issues are different for the three types of PDs (photodiodes, photoconductors, and phototransistors). For perovskite photodiodes, the universal strategies to enhance the performance were introducing an interlayer to regulate the energy band and passivate the traps on the interface, which could suppress the dark current and boost the photocurrent. For perovskite photoconductors, it is pivotal to increase the gain of the device. According to Eq. (
[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] | |
[61] | |
[62] | |
[63] | |
[64] | |
[65] | |
[66] | |
[67] | |
[68] | |
[69] | |
[70] | |
[71] | |
[72] | |
[73] | |
[74] | |
[75] | |
[76] | |
[77] | |
[78] | |
[79] | |
[80] | |
[81] | |
[82] | |
[83] | |
[84] | |
[85] | |
[86] | |
[87] | |
[88] | |
[89] | |
[90] | |
[91] | |
[92] | |
[93] | |
[94] | |
[95] | |
[96] | |
[97] | |
[98] | |
[99] | |
[100] | |
[101] | |
[102] | |
[103] | |
[104] | |
[105] | |
[106] | |
[107] | |
[108] | |
[109] | |
[110] | |
[111] | |
[112] | |
[113] | |
[114] | |
[115] | |
[116] | |
[117] | |
[118] | |
[119] | |
[120] | |
[121] | |
[122] | |
[123] | |
[124] | |
[125] | |
[126] | |
[127] | |
[128] | |
[129] | |
[130] | |
[131] | |
[132] | |
[133] | |
[134] | |
[135] | |
[136] | |
[137] | |
[138] | |
[139] | |
[140] | |
[141] | |
[142] | |
[143] | |
[144] | |
[145] | |
[146] | |
[147] | |
[148] | |
[149] | |
[150] | |
[151] | |
[152] | |
[153] | |
[154] | |
[155] | |
[156] | |
[157] | |
[158] | |
[159] | |
[160] | |
[161] | |
[162] | |
[163] | |
[164] | |
[165] | |
[166] | |
[167] | |
[168] | |
[169] | |
[170] | |
[171] | |
[172] | |
[173] | |
[174] | |
[175] | |
[176] | |
[177] | |
[178] | |
[179] | |
[180] | |
[181] | |
[182] | |
[183] | |
[184] | |
[185] | |
[186] | |
[187] | |
[188] | |
[189] | |
[190] | |
[191] | |
[192] | |
[193] | |
[194] | |
[195] | |
[196] | |
[197] | |
[198] | |
[199] | |
[200] | |
[201] | |
[202] | |
[203] | |
[204] | |
[205] | |
[206] | |
[207] | |
[208] | |
[209] | |
[210] | |
[211] | |
[212] | |
[213] | |
[214] | |
[215] | |
[216] | |
[217] | |
[218] | |
[219] | |
[220] | |
[221] | |
[222] |