Self-assembled heterogeneous chiral perovskite thin film, preparation method thereof and perovskite thin film transistor

Abstract

The invention belongs to the technical field of perovskite materials, and particularly relates to a self-assembled heterogeneous chiral perovskite thin film, a preparation method thereof and a perovskite thin film transistor. Chiral organic amine salt, achiral organic amine salt and stannous halide are dissolved in an organic solvent to obtain a perovskite precursor solution, the perovskite precursor solution is deposited on a to-be-deposited layer, annealing treatment is performed to obtain the self-assembled heterogeneous chiral perovskite thin film, and the thin film has the advantages of high crystal quality, high resolution, high resolution and the like. The material has high-efficiency spin selectivity and excellent carrier transmission performance, and still has high-efficiency spin selectivity even under the condition that the doping amount of chiral molecules is extremely low, so that highly-asymmetric circularly polarized light resolution capability is realized, carrier loss at a grain boundary is effectively inhibited, carrier transmission across grains is promoted, and the resolution efficiency of circularly polarized light is improved. And the photoelectric property of the device can be improved.

Description

Technical Field

[0001] This invention belongs to the field of perovskite material technology, specifically relating to a self-assembled heterogeneous chiral perovskite thin film and its preparation method, and a perovskite thin film transistor. Background Technology

[0002] As an emerging material in chiral electronics, chiral perovskites have shown great potential in the field of circularly polarized light-resolved biomimetic retinal sensors. Their inherent chiral-induced spin selectivity, excellent carrier transport characteristics, and structural tunability are conducive to the design of devices with multiple biomimetic functions.

[0003] However, preparing chiral perovskites that combine high spin selectivity with excellent carrier transport performance still faces severe challenges: on the one hand, although high concentrations of chiral molecules can endow materials with high spin selectivity, large steric hindrance will hinder effective binding with the inorganic framework during crystallization, resulting in poor crystal quality and weak carrier transport performance; on the other hand, although reducing the chiral molecule content helps to improve crystal quality and carrier transport efficiency, it will inevitably weaken its spin selectivity.

[0004] In recent years, researchers have found that partially replacing chiral molecules with achiral molecules can improve the film quality of perovskite films. For example, Chinese Patent 202210708526.X uses traditional stoichiometry for doping. Although this can improve the surface crystallinity of pure chiral perovskite nanosheets to some extent, its improvement on film quality and response to circularly polarized light is still limited. Summary of the Invention

[0005] In view of the problems existing in the prior art, the purpose of the present invention is to provide a self-assembled heterogeneous chiral perovskite thin film and its preparation method, and a perovskite thin film transistor.

[0006] To achieve the above objectives, the first aspect of the present invention provides a method for preparing a self-assembled heterogeneous chiral perovskite thin film, comprising the following steps: A chiral organic amine salt, an achiral organic amine salt, and tin halide are dissolved in an organic solvent to obtain a perovskite precursor solution. This perovskite precursor solution is then deposited onto the layer to be deposited, and after annealing, the self-assembled heterogeneous chiral perovskite film is obtained. The molar amounts of the chiral organic amine salt, the achiral organic amine salt, and the tin halide are a, b, and c, respectively, and a, b, and c satisfy the following condition: 2:1 < (a+b):c ≤ 2.5:1, and a≠0, b≠0, which means that the ratio of the sum of the moles of chiral organic amine salts and non-chiral organic amine salts to stannous halide is greater than 2 and less than or equal to 2.5.

[0007] This invention is based on a strategy of partially replacing chiral molecules with achiral molecules. By doping chiral and achiral molecules in a non-stoichiometric manner and forming a perovskite precursor solution with metal halides, a self-assembled heterostructure chiral perovskite thin film with high crystallinity is prepared. Its heterostructure consists of chiral-deficient grains and chiral-rich grain boundaries. The chiral-rich grain boundaries act as in-plane spin valves, enabling the material to maintain high spin selectivity even with extremely low chiral molecule doping levels, thereby achieving highly asymmetric circularly polarized light resolution. Simultaneously, this structure effectively suppresses carrier loss at grain boundaries and promotes carrier transport across grains, contributing to improved optoelectronic performance of the device.

[0008] The inventors further revealed the mechanism by which the self-assembled heterogeneous chiral perovskite film still achieves high spin selectivity and excellent carrier transport performance under low chiral molecule incorporation: by utilizing the difference in binding energy between chiral and achiral organic cations, their crystallization rate and spatial distribution can be controlled, and the formation of a self-assembled heterogeneous microstructure composed of chiral-rich grain boundaries and chiral-deficient grains can be successfully guided. This structure has not been reported in tin / lead-based chiral perovskites, providing a new design idea for developing high-performance transistor-type biomimetic retinal sensors.

[0009] Preferably, a, b, and c satisfy the following condition: (a+b):c = 2.2:1.

[0010] Preferably, the molar ratio of the chiral organic amine salt, the non-chiral organic amine salt, and the tin halide is a:b:c = 0.4:1.8:1.

[0011] In some embodiments, the non-chiral organic amine salt includes phenylethyl ammonium iodide (PEAI).

[0012] In some embodiments, the chiral organic amine salt includes at least one of R-methylbenzylammonium iodide (R-MBAI) and S-methylbenzylammonium iodide (S-MBAI).

[0013] In some embodiments, the organic solvent includes at least one of dimethyl sulfoxide, N,N-dimethylformamide, N,N-dimethylacetamide, and N-methylpyrrolidone.

[0014] In some embodiments, the concentration of the perovskite precursor solution is 0.05-0.25 mol / L.

[0015] In some embodiments, the annealing treatment is performed at a temperature of 90-110°C for a time of 10-20 minutes.

[0016] In a second aspect, the present invention provides a self-assembled heterogeneous chiral perovskite thin film, which is prepared by the above method.

[0017] In a third aspect, the present invention provides a perovskite thin-film transistor, the perovskite thin-film transistor comprising a gate electrode, a gate insulating layer, a source electrode, a drain electrode, and a light-absorbing layer, wherein the light-absorbing layer is the aforementioned self-assembled heterogeneous chiral perovskite thin film.

[0018] In some embodiments, the gate electrode is silicon, the gate insulating layer is silicon dioxide, and the thickness of the gate insulating layer is 50-150 nm.

[0019] In some embodiments, the thickness of the silicon dioxide is 50-150 nm.

[0020] In some embodiments, the source electrode and drain electrode are made of at least one of Au, Ni, Ag, and Al.

[0021] In some implementations, the thickness of the light-absorbing layer is 50-100 nm.

[0022] In some implementations, the thickness of the source electrode and the drain electrode is 30-50 nm.

[0023] In some embodiments, a conductive channel is provided between the source electrode and the drain electrode, the conductive channel having a length of 50-200 μm and a width of 500-1500 μm.

[0024] In some implementations, the perovskite thin-film transistor has a structure that is either a bottom-gate bottom-contact structure or a bottom-gate top-contact structure.

[0025] In a fourth aspect, the present invention provides a method for fabricating a perovskite thin-film transistor, comprising the following steps: depositing a source electrode and a drain electrode on a gate electrode covered with a gate insulating layer by vacuum evaporation to obtain a layer to be deposited; and then fabricating a self-assembled heterogeneous chiral perovskite thin film by the method described above to obtain the perovskite thin-film transistor.

[0026] Compared with the prior art, the present invention has the following beneficial effects: (1) The self-assembled heterochiral perovskite thin film prepared by the present invention has high crystal quality and has both high spin selectivity and excellent carrier transport performance: even when the material has extremely low chiral molecule doping, it still has high spin selectivity, thereby achieving highly asymmetric circular polarization light resolution, effectively suppressing carrier loss at grain boundaries, and promoting carrier transport across grains, which helps to improve the optoelectronic performance of the device.

[0027] (2) The chiral transistors prepared based on this thin film exhibit excellent chiral optoelectronic performance, with an asymmetry factor as high as 1.98, close to the theoretical limit of 2.0. They have both high spin selectivity and excellent carrier transport performance, and can be used to design a variety of biomimetic chiral perovskite sensors. The prepared chiral transistor perovskite sensors can simultaneously realize a variety of biomimetic functions such as synaptic behavior, light adaptation, color recognition, information encryption and binocular visual information reconstruction, effectively simulating part of the visual mechanism of the human eye. Attached Figure Description

[0028] Figure 1 (a) is a schematic diagram of a gold electrode, and (b) is a schematic diagram of a chiral perovskite transistor, where V D V is the drain, Vs is the source, and gate is the gate electrode.

[0029] Figure 2 (a) is a SEM image of the pure chiral (R-MBA)₂SnI₄ perovskite film of Comparative Example 1, and (b) is a SEM image of the (R-MBA)₂SnI₄ perovskite film of Comparative Example 2. 0.7 PEA 0.3 SEM images of homogeneous chiral perovskite films of 2SnI4, (c) is a comparative example 3 (R-MBA). 0.4 PEA 0.6 SEM images of homogeneous chiral perovskite films of 2SnI4, (d) is the (R-MBA) film of Comparative Example 4. 0.1 PEA 0.9 SEM images of homogeneous chiral perovskite films of 2SnI4, (e) showing the (R-MBA) film of Example 1. 0.2 PEA 0.9 SEM images of 2SnI4 heterochiral perovskite films, (f) showing the (S-MBA) film of Example 2. 0.2 PEA 0.9 SEM image of a heterogeneous chiral perovskite film of 2SnI4.

[0030] Figure 3 (a) is a schematic diagram of a homogeneous chiral perovskite, (b) is a schematic diagram of a heterogeneous chiral perovskite, (c) is an AFM image of the homogeneous chiral perovskite film of Comparative Example 4, (d) is an IR-AFM image of the homogeneous chiral perovskite film of Comparative Example 4, (e) is the AFM height profile and IR-AFM intensity profile at the grain boundaries of the homogeneous chiral perovskite film of Comparative Example 4, (f) is an AFM image of the heterogeneous chiral perovskite film of Example 1, (g) is an IR-AFM image of the heterogeneous chiral perovskite film of Example 1, and (h) is the AFM height profile and IR-AFM intensity profile at the grain boundaries of the heterogeneous chiral perovskite film of Example 1.

[0031] Figure 4(a) shows the X-ray diffraction spectra of different types of perovskites, where Heterogeneous is the heterochiral perovskite of Example 1, Homogeneous is the homochiral perovskite of Comparative Example 4, MBA2SnI4 is the pure chiral perovskite, and PEA2SnI4 is the achiral perovskite. (b) shows the formation energies of the achiral perovskite PEA2SnI4 and the chiral perovskite MBA2SnI4. (c) shows the transfer curves of the heterochiral perovskite transistor device of Example 1 under dark conditions and circularly polarized light irradiation, along with the corresponding gph values. DS The voltage was -30 V, the irradiation wavelength was 607 nm, and the irradiation power density was 1 μW cm⁻¹. -2 (d) shows the transfer curves and corresponding gph values ​​of the homogeneous chiral perovskite transistor device of Comparative Example 4 under dark conditions and circularly polarized light illumination, where V DS The voltage was -30 V, the irradiation wavelength was 607 nm, and the irradiation power density was 1 μW cm⁻¹. -2 .

[0032] Figure 5 (a) is a schematic diagram of mCP-AFM measurement; (b) is the average IV curve of the homogeneous chiral perovskite sample of Comparative Example 4 under different implantation spin polarizations, where P is the spin polarization efficiency, and the average value of 20 IV curves measured from different positions of the sample is taken; (c) is the average IV curve of the heterogeneous chiral perovskite sample of Example 1 under different implantation spin polarizations, where P is the spin polarization efficiency, and the average value of 20 IV curves measured from different positions of the sample is taken.

[0033] Figure 6 (a) shows the PPF behavior under the action of RCP and LCP pulse pairs, where the frequency is 1 Hz, the wavelength is 607 nm, the pulse width is 200 ms, and the intensity is 12 μW cm⁻¹. -2 V GS 30 V, V DS (a) shows the EPSC behavior under 20 consecutive RCP and LCP pulses, with a frequency of 1 Hz, a wavelength of 607 nm, a pulse width of 200 ms, and an intensity of 12 μW / cm². -2 V GS 30 V, V DS (c) is -0.1 V, and (c) is the change of EPSC increment with the number of RCP and LCP pulses.

[0034] Figure 7 (a) I changing over time under different light intensities DS Among them, the wavelength is 450nm, (b) AI 100sThe value is a function of light intensity and incident light wavelength. (c) is a schematic diagram of light-adapted imaging, (d) is a schematic diagram of dark-adapted imaging, and (e) is the evolution of photocurrent in light-adapted and dark-adapted imaging.

[0035] Figure 8 (a) Schematic diagram of an artificial vision system using a retinal sensor array based on CPL decomposition and convolutional neural networks (CNNs) for encrypted information preprocessing and recognition in the CIFAR-10 dataset; (b) Schematic diagram of images encrypted with different levels of LCP noise and decrypted by a retinal sensor array based on CPL decomposition; (c) Recognition accuracy of cat patterns under different levels of LCP noise after 100 training cycles; (d) Confusion matrix of training results based on a homogeneous chiral perovskite retinal sensor array after 100 cycles, where the LCP noise level is 100%; (e) Confusion matrix of training results based on a heterogeneous chiral perovskite retinal sensor array after 100 cycles, where the LCP noise level is 100%; (f) Recognition accuracy of cat patterns with an EPSC decay time span after 100 training cycles, where the LCP noise level is 100%.

[0036] Figure 9 (a) is a schematic diagram of virtual stereo reconstruction using binocular vision based on two CPL-resolved retinal morphology sensor arrays with opposite chirality; (b) is a photocurrent map from the CPL multiplexing screen of the R array and S array; (c) is a superimposed photocurrent map of the R array and S array used for virtual stereo reconstruction. The plus sign and circle mark represent the photocurrent signal points of the left and right images, respectively, and the green line indicates their point-to-point correspondence; (d) is the virtual stereo depth reconstruction result of the "HIT" pattern by the CPL-resolved retinal morphology sensor array based on heterogeneous chiral perovskite; (e) is a comparison of the reconstruction depth with the designed depth of the three letters ("H", "I" and "T"); and (f) is the virtual stereo depth reconstruction result of the "HIT" pattern by the sensor array based on homogeneous chiral perovskite. Detailed Implementation

[0037] The following detailed embodiments further illustrate the content of the present invention. These embodiments do not constitute a limitation on the scope of protection of the present invention. Non-essential modifications and adjustments made by others based on the concept of the present invention still fall within the scope of protection of the present invention. The raw materials, reagents, or devices used in the embodiments are all available from conventional commercial sources or can be obtained through existing technical methods. Unless otherwise specified, the experimental or testing methods are conventional methods in the art.

[0038] Example 1 (R-MBA) 0.2 PEA 0.9)2SnI4 heterochiral perovskite thin films and transistors A method for fabricating a self-assembled heterogeneous chiral perovskite thin-film transistor includes the following steps: S1. Substrate Pretreatment: A transistor-type sensor is fabricated on a cleaned p+-Si / SiO2 substrate using a bottom-gate bottom-contact structure. The SiO2 insulating layer is 100 nm thick and has a capacitance per unit area of ​​34 nF cm⁻¹. -2 The substrate cleaning process involved sequentially immersing the substrate in deionized water, acetone, and isopropanol for 15 minutes of ultrasonic cleaning to remove surface contaminants and improve interface quality. After cleaning, a 40 nm thick gold layer was deposited in a vacuum environment as the source and drain electrodes. A channel with a length (L) of 100 μm and a width (W) of 1000 μm was formed between the source and drain electrodes. A schematic diagram of the gold electrodes is shown below. Figure 1 As shown in figure a. Subsequently, the substrate with electrodes was subjected to ultraviolet ozone treatment for 30 min to further improve surface properties. After treatment, the substrate was transferred to a glove box for spin-coating deposition of perovskite thin films.

[0039] S2. Preparation of perovskite precursor solution: Dissolve R-MBAI, PEAI, and SnI2 in a mixed solvent of DMF and DMSO at a molar ratio of 0.4:1.8:1 (DMF to DMSO volume ratio of 9:1). Place the solution in a glove box and stir for 12 hours under nitrogen atmosphere, water oxygen concentration <1ppm, and 60℃ to ensure complete dissolution, obtaining a 0.1M (R-MBAI) precursor solution. 0.2 PEA 0.9 )2SnI4 perovskite precursor solution S3. Spin-coating deposition of perovskite thin films: The (R-MBA) film obtained in step S2 is then applied... 0.2 PEA 0.9 The 2SnI4 perovskite precursor solution was spin-coated onto the substrate at a rotation speed of 4000 rpm for 50 s, resulting in a deposition thickness of 50 nm.

[0040] S4. Annealing: The spin-coated substrate is annealed at 100°C for 10 minutes to obtain a transistor-type sensor. A schematic diagram of the transistor-type sensor structure is shown below. Figure 1 As shown in b.

[0041] Example 2 (S-MBA) 0.2 PEA 0.9 )2SnI4 heterochiral perovskite thin films and transistors The only difference from Example 1 is that in step S2, R-MBAI is replaced with S-MBAI in equimolar amounts; the remaining components and preparation steps are exactly the same as in Example 1.

[0042] Comparative Examples 1-4: Pure chiral perovskite thin film R-MBA2SnI4 and transistors, homogeneous chiral perovskite thin film (R-MBA2SnI4 ... 1-x PEAx)2SnI4 and transistors The only difference between Comparative Examples 1-4 and Example 1 is that the molar ratio of R-MBAI, PEAI and SnI2 is different in step S2, as shown in Table 1 below. The amounts of the remaining components and the preparation methods are exactly the same as in Example 1. Test Example 1: Scanning Electron Microscopy (SEM) The R-MBA prepared in Examples 1-2 and Comparative Examples 1-4 were observed by SEM. x PEA y The surface morphology of the 2SnI4 perovskite film is shown in the following results. Figure 2 As shown. When y=0, the chiral organic molecules in pure R-MBA2SnI4 chiral perovskite have large steric hindrance, hindering the crystallization of chiral perovskite. Therefore, chiral perovskite exhibits poor film quality, with uneven surface morphology and high density of trenches and pinholes. Figure 2 a) This would impair its carrier dynamics, leading to the loss of fundamental transistor behavior in the device. Partially replacing the chiral molecule R-MBA with the achiral molecule PEA can improve the thin film quality of chiral perovskites by reducing the octahedral tilt in the lattice. For films conforming to conventional stoichiometry (R-MBA)... 1-x PEA x For 2SnI4 perovskite thin films (where the traditional stoichiometric ratio satisfies x + y = 1), the film quality improves as y increases from 0.3 to 0.9. Figure 2 However, in terms of crystal quality, film uniformity, grain size, and obvious grain boundaries, it is still significantly inferior to the heterogeneous chiral perovskite films of Examples 1-2. Figure 2 ef).

[0043] Test Example 2: Atomic Force Microscopy (AFM) and Infrared Atomic Force Microscopy (IR-AFM) The heterogeneous chiral perovskite film prepared in Example 1 and the homogeneous chiral perovskite film prepared in Comparative Example 4 were observed by AFM and IR-AFM. The results are as follows: Figure 3 As shown. Figure 3 a-3b are schematic diagrams of the structures of homogeneous and heterogeneous chiral perovskite films. Atomic force microscopy (AFM) reveals that the homogeneous chiral perovskite exhibits distinct grains and concave grain boundaries. Figure 3 c). R-MBA molecules were observed to be uniformly distributed within the grains using infrared atomic force microscopy (IR-AFM). Figure 3 d and 3e). However, for heterochiral perovskites, AFM images show that the grain boundaries change from concave to convex (d and 3e). Figure 3 f). Accordingly, IR-AFM images revealed the non-uniform distribution of R-MBA molecules at the grain boundaries (f). Figure 3 (g and 3h) indicates the formation of a self-assembled heterostructure microstructure consisting of chiral grain boundaries and chiral grains.

[0044] Test Example 3: X-ray Diffraction (XRD) The heterogeneous chiral perovskite film of Example 1 and the homogeneous chiral perovskite film of Comparative Example 4 were tested by XRD, and the results are as follows: Figure 4 As shown in Figure a. The X-ray diffraction (XRD) spectrum of the heterochiral perovskite of Example 1 shows a weak shoulder peak at 6.26°, corresponding to the (020) diffraction peak of R-MBA2SnI4. Figure 4 a), which is attributed to the filling of grain boundaries by R-MBA2SnI4. In addition, the diffraction peaks at approximately 5.6° of the homogeneous chiral perovskite of Comparative Example 4 and the heterogeneous chiral perovskite of Example 1 are slightly shifted compared to the (002) diffraction peak of PEA2SnI4, indicating that a small amount of R-MBA molecules are incorporated into the grains.

[0045] Test Example 4: Density Functional Theory (DFT) Calculation The formation of self-assembled heterochiral perovskites was understood through density functional theory (DFT) calculations, and the results are as follows: Figure 4 As shown in b, at 100℃ (the annealing temperature used in the crystallization process), the formation energy of PEA2SnI4 is -1.238 eV per atom, while the formation energy of chiral perovskite R-MBA2SnI4 is -1.142 eV per atom. Figure 4 (b) The difference of 96 meV is significantly higher than the thermal activation energy at 100 °C (31.8 meV), which is believed to affect the crystallization preference of different components in heterochiral perovskites, and the crystallization of R-MBA2SnI4 with higher formation energy will be inhibited. Considering the radial growth direction of the grains, it is expected that excess chiral R-MBA molecules will accumulate at the grain edges, and the small portion of R-MBA2SnI4 with a slower crystallization rate will be confined to the grain boundaries.

[0046] Test Example 5: Carrier Transport Characteristics and Photoelectric Performance of Transistor Sensors Unique self-assembled heterogeneous microstructures facilitate cross-grain carrier transport and spin selectivity, fundamental to the chiral electronic performance of transistor-type circularly polarized retinal sensors. Perovskite grain boundaries are typically defective and geometrically discontinuous between grains, hindering cross-grain carrier transport, crucial for transistor-type devices. Chiral components filling grain boundaries connect adjacent grains and passivate grain boundary defects, thereby suppressing carrier loss and promoting cross-grain carrier transport.

[0047] This invention fabricates a heterogeneous chiral perovskite transistor sensor with a non-stoichiometric ratio (referred to as the heterogeneous sensor, corresponding to Example 1) and a homogeneous chiral perovskite transistor sensor with a conventional stoichiometric ratio (referred to as the homogeneous sensor, corresponding to Comparative Example 4) based on R-MBA, and characterizes the carrier transport properties of these two sensors for the first time. Specifically, the transfer curves of the two sensors under dark conditions are compared, and the on-state current of the heterogeneous sensor (3.3 × 10⁻⁶) is... -5 A) Significantly higher than that of similar sensors (2.4 × 10⁻⁶). -6 A). Meanwhile, the carrier mobilities (μ) extracted from the transfer curves of the heterogeneous and homogeneous sensors were 0.36 and 0.12 cm⁻¹, respectively. 2 V -1 s -1 This confirms the suppression of carrier loss and the promotion of transgrain carrier transport in heterochiral perovskites.

[0048] This invention also characterizes the photoelectric performance of two types of sensors under the same illumination intensity of 607 nm right-handed circularly polarized light (RCP) and left-handed circularly polarized light (LCP). In both types of sensors, a significant improvement in the transfer curves under RCP and LCP illumination is observed, attributed to the optical gating effect caused by the accumulation of photogenerated carriers. However, in the heterogeneous sensor, the optical gating effect under RCP illumination is more significant than that under LCP illumination, and a threshold voltage difference (ΔV) between the two transfer curves can be obtained. TH The voltage is 3.3V. This ΔV... TH This results in a high photocurrent asymmetry factor (gph) and a gate voltage (V) GS The window is widened, with a linewidth of 10 V, which facilitates obtaining a high gph during operation. The peak gph of the heterogeneous sensor can reach as high as 1.98, approaching the theoretical limit of 2.0 for complete differentiation between RCP and LCP. In contrast, the transfer curves of the homogeneous sensor under RCP and LCP irradiation almost overlap, and the ΔV between the two transfer curves is very small. TH The value is 0.1 V, reflecting the subtle difference in optical gating effects under RCP and LCP illumination. Although a peak gph as high as 1.17 can be obtained, homogeneous sensors are almost indistinguishable during operation from the narrow V values ​​of circularly polarized light.GS Window. ΔV between the two types of sensors. TH The differences are significant, indicating that heterogeneous sensors have a clear CPL discrimination capability.

[0049] Test Example 6: Magnetic Conductive Probe Atomic Force Microscopy (mCP-AFM) To further demonstrate the difference between the heterogeneous sensor (Example 1) and the homogeneous sensor (Comparative Example 4), this invention uses magnetic conductive probe atomic force microscopy (mCP-AFM) equipped with a Co / Cr coated probe to characterize the spin selectivity of the two chiral perovskites. This probe is capable of injecting carriers with a single spin polarization state (represented as spin-up or spin-down). This invention prepared a sample with a single bottom electrode and positioned the probe at a lateral distance of approximately 100 μm from the bottom electrode to simulate in-plane carrier transport in a transistor-type sensor. A schematic diagram of the measurement is shown below. Figure 5 As shown in a. For homogeneous chiral perovskites, the spin polarization efficiency (P) quantitative carrier spin selection effect is 13% at 2 V ( Figure 5 b). The inventors anticipate that the weak spin selectivity may be due to the limited R-MBA incorporated into the material. In contrast, the p-value of heterochiral perovskites is significantly increased, reaching 68% at 2 V ( Figure 5 c). The inventors anticipate that chiral grain boundaries act as spin valves in the plane, thereby achieving cross-grain boundary spin selectivity, such as... Figure 3 As shown in b. Correspondingly, in heterogeneous devices, the spin polarization asymmetry induced by circularly polarized light absorption is amplified during transgrain boundary transport, thereby suppressing spin relaxation. In this case, the distribution of photogenerated carriers accumulated in heterogeneous devices under RCP and LCP irradiation differs significantly, which explains the difference in ΔV between the two types of sensors. TH The huge difference between light-gated effects and light-gated effects.

[0050] Test Example 7: Investigating the Synaptic Plasticity of a Retinal Bionic Sensor Thanks to the quantum well structure of two-dimensional perovskites, which enables intrinsic storage of charge carriers, circularly polarized-resolved heterogeneous sensors can also achieve a variety of biomimetic functions, thus being used in circularly polarized-resolved biomimetic retinal sensors.

[0051] This invention uses a 4200 semiconductor analyzer, a waveform generator, and a laser to perform IT curve testing on heterogeneous samples (Example 1), and adjusts the number of laser pulses to change it from STP to LTP.

[0052] This invention investigates synaptic behavior that simulates the memory and learning functions of biological neurons. The sensor of this invention exhibits memory of incident signals, displaying a highly asymmetric circularly polarized photoluminescence (CPL) response. Paired pulse facilitation (PPF), a key feature of short-term plasticity (STP), refers to the phenomenon where successive stimulation of the preneuron leads to signal enhancement in the postneuron; this is also demonstrated by the sensor of this invention through a highly asymmetric CPL response. Figure 6 a). This process is related to the temporal and spatial encoding / decoding of visual information and can be quantified by the PPF index.

[0053] Besides STP, long-term plasticity (LTP) with a highly asymmetric CPL response can also be achieved through a repetitive learning process. It was found that EPSC gradually saturates after 20 consecutive LCP pulses, while EPSC continuously increases after 20 consecutive RCP pulses. Figure 6 b), this is also evident in the summary of EPSC changes (∆EPSC) under different CPL pulses ( Figure 6 c). The inventors anticipate that the sensor of the present invention exhibits a more significant EPSC dependence on incident CPL signals with the same spin orientation as the chiral perovskite used in the device. This is likely due to the effective spin selectivity in the heterogeneous chiral perovskite suppressing the accumulation of carriers with opposite polarization states. This characteristic ensures effective memorization and learning of incident CPL signals with the same spin orientation as the chiral perovskite used in the device, while restricting opposite signals, which is beneficial for applications such as encrypted visual information perception.

[0054] Test Example 8: Investigating the Light Adaptability of a Retinal Bionic Sensor The human retina possesses visual adaptation capabilities, namely light adaptation and dark adaptation, enabling it to self-adjust to ambient light intensity and achieve effective visual perception in both bright and dim environments. Our retinal biomimetic sensor demonstrates its ability to adapt to variations in ambient light intensity across the visible spectrum. Taking variable-intensity 450 nm illumination as an example, by adjusting the gate voltage (V0) during a 100-second dynamic adaptation process... GS The dynamic drain-source current (Io) of the retinal bionic sensor of the present invention (i.e., the transistor obtained in Example 1) was investigated. DS )characteristic( Figure 7 a). Under strong light irradiation (40 mW cm⁻¹) -2 Up to 40 µW cm -2 ), by applying V GS Light adaptation can be achieved at ≤0 V. Conversely, under weak light irradiation (4 µW cm⁻¹), -2 Up to 40 nW cm -2 ), by applying V GSA value greater than 0 V enables dark adaptation. During this process, V can be adjusted appropriately. GS This can weaken the adaptability. The fitness index (AI) is the final state photocurrent (If). f ) and initial state photocurrent (I i The ratio of AI to 100 s can be used to quantify adaptability. 100s ),like Figure 7 As shown in b. Specifically, AI 100s <1 indicates that the photocurrent is suppressed, corresponding to light adaptation, at 40 mW / cm². -2 AI scores as low as 0.72 can be obtained at this time. 100s Value. Furthermore, AI 100s >1 indicates enhanced photocurrent, corresponding to dark adaptation, at 40 nW / cm². -2 Up to 4×10 can be obtained at this time. 5 AI 100s value.

[0055] To clearly demonstrate the adaptability of the retinal bionic sensor of this invention, a 7×7 sensor array was fabricated. The pattern of the letter "H" was sensed by measuring the time-dependent drain current under a fixed gate voltage under continuous strong and weak light illumination conditions. In the light adaptation test, 13 sensors sensing the "H" pattern were exposed to 40 mW cm⁻¹ light. -2 Under illumination, the remaining 36 sensors for background sensing were exposed to 4 mW cm⁻¹ light, while the other 36 sensors for background sensing were exposed to 4 mW cm⁻¹ light. -2 Under the light ( Figure 7 c). In the dark adaptation test, 13 sensors sensing the "H" pattern were exposed to 0.4 µW cm⁻¹. -2 Under illumination, the remaining 36 sensors for background sensing were exposed to 40 nW cm⁻¹ light, while the other 36 sensors were exposed to light at the same level. -2 Under the light ( Figure 7 d). In light adaptation and dark adaptation tests, when illumination begins at 0s, the retinal bionic sensor array can barely distinguish the "H" pattern from the background ( Figure 7 e). Over time, the visual adaptation effect gradually emerges, resulting in a significant increase in the contrast of the "H" pattern, which is remarkably similar to the adaptation process of the retina. This biomimetic in-sensor visual adaptation function can be effectively applied to artificial vision systems, thereby simplifying external circuit requirements and reducing reliance on complex processing algorithms.

[0056] Test Example 9: Exploring the potential applications of retinal bionic sensors in artificial vision systems. Based on the characteristics of the CP-resolved retinal sensor described above, we will now focus on its potential applications in artificial vision systems that require highly asymmetric circularly polarized light response.

[0057] Visual information can be encoded using circularly polarized light in orthogonal modes (such as intensity and color) to achieve high-capacity, multi-channel encryption against forgery and eavesdropping. Circularly polarized light-resolved retinal sensors can directly decrypt the circularly polarized light-encrypted visual information during the sensing process and perform data processing within the device, which is considered a key component of visual information encryption.

[0058] As a demonstration, the inventors generated an image of a cat using RCP pulses and then encrypted the RCP image with background noise using LCP pulses, resulting in an encrypted hybrid image. Figure 8 a). The encrypted mixed image is decomposed into red, green, and blue (RGB) channels, which are then collected sequentially by a CP-resolved biomimetic retinal sensor array based on heterogeneous chiral perovskite. This array preprocesses the encrypted mixed image by selectively responding to the RCP components in the image. The photocurrent collected by the sensor is then fed into a convolutional neural network (CNN) for convolution and pooling operations, followed by unrolling and input into a fully connected network with 576 input neurons. Finally, the network performs a high-level recognition task through 10 output neurons. The CNN model is then trained using a subset of 5000 images from the Canadian Institute for Advanced Research (CIFAR-10) dataset, using the "adam" optimization algorithm, trained for 100 epochs, and applied to 10 image recognition tasks. Specifically, different levels of noise-level LCP are used to simulate different degrees of visual information encryption (…). Figure 8 b). As the LCP noise level increases, the cat image generated by RCP gradually becomes indistinguishable in the encrypted mixed image. Regardless of the LCP noise level in the encrypted image, our CPL decomposition retinal sensor (i.e., the transistor obtained in Example 1) is able to effectively decrypt the image and clearly identify the cat pattern in the decrypted image. Under 100% LCP noise conditions, after 100 training cycles, the recognition accuracy of the target image decrypted by the CP decomposition retinal sensor of this invention reaches 95% ( Figure 8 (CP-resolved in c). For direct comparison, a sensor based on homogeneous chiral perovskite with limited CPL discrimination capability (i.e., the transistor obtained in Comparative Example 4) was also used to perceive encrypted images. The results showed that with increasing LCP noise levels, the recognition accuracy dropped significantly to 54.6% after 100 training cycles, which can be considered as an inability to recognize encrypted images. Figure 8 The inventors further identified patterns from the CIFAR-10 dataset, and obtained the confusion matrix of the two types of sensors under 100% LCP noise (in c). Figure 8(d and 8e) demonstrate that the CPL-decomposed retinal sensor of the present invention can achieve accurate recognition, which is beneficial for constructing artificial vision systems for visual information encryption. After sensing visual information, the EPSC (Earnings Per Crush Rate) of the sensor at different decay times (10 s, 30 s, and 60 s) was also measured to evaluate the contrast enhancement function within the sensor, which simulates the human retina. As the decay time increases, the difference in recognition accuracy between the two types of sensors also increases ( Figure 8 f). The CPL-resolution bionic retinal sensor of the present invention has a preferential CPL response and can obtain long-term EPSC from the RCP signal, which facilitates preprocessing within the sensor, thereby enabling optical information sensing and long-term memory retention capabilities.

[0059] Test Example 10: Exploring the potential applications of sensors in building artificial vision systems for immersive human-computer interaction. Having demonstrated the potential applications of a circularly polarized-resolved retinal sensor based on a heterogeneous chiral perovskite with a single chirality, the inventors further explored the potential applications using two circularly polarized-resolved retinal sensor arrays with opposite chirality, namely the transistors obtained in Example 1 and Example 2, respectively. By separating an image into two images with opposite polarization states, a display can generate three-dimensional spatial information of an object, as is common in 3D movies. The left and right eyes then perceive one of the polarized images via polarizers, and parallax is generated through the binocular parallax between the two images, which is then used to reconstruct virtual three-dimensional spatial information. Utilizing its highly asymmetric circular polarization response, our retinal sensor can be used in artificial vision systems that achieve binocular stereoscopic vision directly from a 3D display, which is promising for immersive human-computer interaction.

[0060] Figure 9 This diagram schematically illustrates 3D depth reconstruction using two CPL-resolved retinal arrays with opposite rotational orientations (i.e., the transistors obtained in Example 1 and Example 2, denoted as the R array and S array, respectively). Two sets of "HIT" patterns are generated using a CPL multiplexing screen, one using RCP and the other using LCP. The spatial distribution of the six letters in the two sets of "HIT" patterns on the screen is specifically designed to encode the depth contours of the letters during binocular stereoscopic vision reconstruction. These patterns are then focused by convex lenses and projected onto the R and S arrays. Figure 9 b shows the corresponding photocurrent mappings from the two arrays, indicating that the R and S arrays capture patterns with the same spin, respectively, due to their highly asymmetric CPL responses. This is further demonstrated by superimposing the photocurrent mappings of the two arrays and extracting the matched feature points. Figure 9c) can reconstruct the 3D spatial information of the "HIT" pattern. Figure 9 d). The inventors further compared the depth of the reconstructed three letters with the depth designed based on the spatial distribution of the two patterns on the screen. The results showed that the spatial positions of the reconstructed patterns and the designed patterns had a very high degree of consistency, with an average error of approximately 3%. Figure 9 e), which demonstrates the potential of the circularly polarized-resolved retinal sensor of the present invention in constructing artificial vision systems for immersive human-computer interaction. In contrast, due to the limited circularly polarized resolution, sensor arrays based on homogeneous chiral perovskites cannot distinguish the six letters with different polarizations, nor can they reconstruct the 3D spatial information of the corresponding patterns. Figure 9 f).

[0061] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A method for preparing a self-assembled heterogeneous chiral perovskite thin film, characterized in that, Includes the following steps: Chiral organic amine salts, non-chiral organic amine salts, and stannous halides are dissolved in an organic solvent to obtain a perovskite precursor solution. The obtained perovskite precursor solution is deposited onto the layer to be deposited, and after annealing, the self-assembled heterochiral perovskite film is obtained. The molar numbers of the chiral organic amine salt, the non-chiral organic amine salt, and the stannous halide are a, b, and c, respectively, and a, b, and c satisfy the following condition: 2:1 < (a+b):c ≤ 2.5:1, and a≠0, b≠0.

2. The preparation method according to claim 1, characterized in that, The molar ratio of the chiral organic amine salt, the non-chiral organic amine salt, and the stannous halide is a:b:c = 0.4:1.8:

1.

3. The preparation method according to claim 1, characterized in that, The non-chiral organic amine salts include phenylethyl ammonium iodide; The chiral organic amine salt includes at least one of R-methylbenzylammonium iodide and S-methylbenzylammonium iodide; The organic solvent includes at least one of dimethyl sulfoxide, N,N-dimethylformamide, N,N-dimethylacetamide, and N-methylpyrrolidone.

4. The preparation method according to claim 1, characterized in that, The concentration of the perovskite precursor solution is 0.05-0.25 mol / L.

5. The preparation method according to claim 1, characterized in that, The annealing process is carried out at a temperature of 90-110℃ for 10-20 minutes.

6. A self-assembled heterogeneous chiral perovskite thin film, characterized in that, It is prepared by the method described in any one of claims 1-5.

7. A perovskite thin-film transistor, characterized in that, The perovskite thin-film transistor includes a gate electrode, a gate insulating layer, a source electrode, a drain electrode, and a light-absorbing layer; the light-absorbing layer is a self-assembled heterogeneous chiral perovskite thin film as described in claim 6.

8. The perovskite thin-film transistor as claimed in claim 7, characterized in that, Includes at least one of the following features: (1) The gate electrode is silicon; (2) The gate insulating layer is silicon dioxide; (3) The thickness of the gate insulating layer is 50-150 nm; (4) The source electrode and drain electrode are made of at least one of Au, Ni, Ag and Al; (5) The thickness of the light-absorbing layer is 50-100 nm; (6) The thickness of the source electrode and the drain electrode is 30-50 nm; (7) A conductive channel is provided between the source electrode and the drain electrode, the length of the conductive channel is 50-200μm and the width is 500-1500μm.

9. The perovskite thin-film transistor as claimed in claim 7, characterized in that, The perovskite thin-film transistor has either a bottom-gate bottom-contact structure or a bottom-gate top-contact structure.

10. A method for fabricating a perovskite thin-film transistor, characterized in that, Includes the following steps: A source electrode and a drain electrode are deposited on a gate electrode covered with a gate insulating layer by vacuum evaporation to obtain a layer to be deposited. Then, a self-assembled heterogeneous chiral perovskite thin film is prepared by the method described in any one of claims 1-5 to obtain the perovskite thin film transistor.

Images