A black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector, its fabrication method, and its application.

By using a black arsenic phosphorus molybdenum disulfide heterojunction structure and gate voltage modulation, a wide-spectrum detection from visible light to mid-infrared light is achieved, solving the problems of insufficient spectral response range and multi-level optical response in existing technologies. It can be applied to gas molecule recognition, carbon emission monitoring and thermal scene sensing, improving the system's integration and intelligence level.

CN123124468APending Publication Date: 2026-11-17SHANGHAI INSTITUTE OF TECHNICAL PHYSICS CHINESE ACADEMY OF SCIENCES
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611281568.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-24
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing two-dimensional mid-infrared photoelectronic retina devices have shortcomings in terms of wide spectral response range, multi-level gate voltage reconfigurable optical response, and detection of molecular fingerprints exceeding 4μm.

Method used

By employing a black arsenic phosphorus molybdenum disulfide heterojunction structure and utilizing the broad-spectrum absorption characteristics of narrow-bandgap black arsenic phosphorus material, and by controlling the heterointerface barrier and carrier separation and transport processes through gate voltage, a broad-spectrum detection from visible light, near-infrared to mid-infrared is achieved, forming multiple stable and distinguishable photoresponse states.

Benefits of technology

It achieves wide-spectrum detection from visible light to mid-infrared, covering the mid-infrared molecular fingerprint band of more than 4μm, and can be used for gas molecule identification, carbon emission monitoring and thermal scene perception, reducing the data transmission volume and processing latency of infrared sensing systems, and improving system integration and intelligence.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123124468A_ABST
    Figure CN123124468A_ABST
Patent Text Reader

Abstract

This invention relates to the field of low-dimensional semiconductor optoelectronic devices and infrared photodetector technology, specifically a black arsenic phosphorus molybdenum disulfide heterojunction reconfigurable broadband photodetector, comprising a substrate, a gate dielectric layer, a molybdenum disulfide layer, a black arsenic phosphorus layer, and two metal electrodes. The substrate is a silicon substrate, which also serves as the back gate electrode. The gate dielectric layer is disposed above the silicon substrate. The molybdenum disulfide layer is disposed above the gate dielectric layer. The black arsenic phosphorus layer is disposed above the molybdenum disulfide layer and partially overlaps with it. The overlapping region forms a black arsenic phosphorus molybdenum disulfide van der Waals heterojunction, serving as the main photosensitive region and photogenerated carrier separation region of the device. The two metal electrodes form electrical contacts with the molybdenum disulfide layer and the black arsenic phosphorus layer, respectively. This invention also includes a method for fabricating the photodetector and its applications. This invention enables a single device to achieve broadband detection and obtain multiple stable and distinguishable photoresponse states, thereby realizing the encoding, weighting, and preprocessing of optical information.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of low-dimensional semiconductor optoelectronic devices and infrared photoelectric detection technology, and in particular to a black arsenic phosphine molybdenum disulfide heterojunction reconfigurable broadband photodetector, its preparation method, and its application. Background Technology

[0002] Existing literature discloses a two-dimensional mid-infrared optoelectronic retina device that enables simultaneous perception and encoding of mid-infrared information. The literature information is: Wang FK et al., A two-dimensional mid-infrared optoelectronic retina enabling simultaneous perception and encoding, Nature Communications, 2023, 14, 1938, DOI: 10.1038 / s41467-023-37623-5. This device is based on a two-dimensional b-AsP / MoTe2 van der Waals heterostructure and can provide a reference for mid-infrared target perception and information encoding. However, there is still room for further improvement in terms of wide spectral response range, multi-level gate voltage reconfigurable optical response, and detection of molecular fingerprints exceeding 4 μm. Summary of the Invention

[0003] The purpose of this invention is to provide a black arsenic phosphorus molybdenum disulfide heterojunction reconfigurable broadband photodetector, its fabrication method, and its application. It mainly solves the problems existing in the prior art. By utilizing the broadband absorption characteristics of narrow bandgap black arsenic phosphorus material and using gate voltage to control the heterojunction barrier and carrier separation and transport process, a single device can achieve broadband detection from visible light, near infrared to mid-infrared, and obtain multiple stable and distinguishable light response states. This enables the encoding, weighting, and front-end preprocessing of optical information, reducing the data transmission volume, power consumption, and processing delay of traditional infrared sensing systems.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is to provide a black arsenic phosphorus molybdenum disulfide heterojunction reconfigurable broadband photodetector, characterized in that it includes a substrate, a gate dielectric layer, a molybdenum disulfide layer, a black arsenic phosphorus layer, and two metal electrodes; the substrate is a silicon substrate, which also serves as the back gate electrode; the gate dielectric layer is disposed above the silicon substrate to achieve electrical insulation and electric field coupling between the back gate electrode and the photoelectric transport channel of the device; the molybdenum disulfide layer is disposed above the gate dielectric layer; the black arsenic phosphorus layer is disposed above the molybdenum disulfide layer and partially overlaps with the molybdenum disulfide layer; a black arsenic phosphorus molybdenum disulfide van der Waals heterojunction is formed in the overlapping region of the black arsenic phosphorus layer and the molybdenum disulfide layer, serving as the main photosensitive region and photogenerated carrier separation region of the device; the two metal electrodes respectively form electrical contacts with the molybdenum disulfide layer and the black arsenic phosphorus layer, for applying source-drain bias voltage, forming carrier transport channels, and collecting the photocurrent generated by the device.

[0005] Furthermore, the black arsenic phosphorus layer is made of b-As 0.3 P 0. 7. The material has a band gap of 0.22 eV and is particularly responsive to mid-infrared light with wavelengths exceeding 4 μm; the molybdenum disulfide layer has a band gap of 1.2 eV and is used to form the van der Waals heterointerface with band misalignment with the black arsenic phosphorus layer, and to regulate the interface barrier, built-in electric field, and carrier separation and transport processes under the action of an applied gate voltage.

[0006] Furthermore, the metal electrode is a chromium-gold composite electrode, comprising a chromium layer and a gold layer; wherein the chromium layer is disposed on the surface of the molybdenum disulfide layer and the black arsenic phosphorus layer to improve the adhesion between the electrode and the black arsenic phosphorus or molybdenum disulfide; the gold layer is disposed above the chromium layer to reduce the electrode resistance and improve the electrical signal transmission capability; the two metal electrodes are respectively located at both ends of the heterostructure, so that the black arsenic phosphorus layer, the molybdenum disulfide layer and the van der Waals heterojunction formed by their overlap together constitute a photoelectric transport channel.

[0007] This invention also provides a method for fabricating the above-mentioned black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector, characterized by comprising the following steps:

[0008] Step S101: Clean the substrate and the gate dielectric layer to remove organic contaminants, particles and other impurities from the material surface;

[0009] Step S102: Molybdenum disulfide flakes and black arsenic phosphorus flakes are obtained by mechanical peeling.

[0010] Step S103: Transfer the molybdenum disulfide sheet onto the silicon dioxide surface;

[0011] Step S104: Optically position the black arsenic phosphorus sheet and the molybdenum disulfide sheet, transfer the black arsenic phosphorus sheet above the molybdenum disulfide sheet, so that the two partially overlap and form the van der Waals heterojunction;

[0012] Step S105: Electrode patterns are formed by photolithography or electron beam exposure, and a chromium-gold metal layer is deposited by metal evaporation or sputtering.

[0013] Step S106: The metal electrodes connected to the black arsenic phosphorus layer and the molybdenum disulfide layer are formed by a stripping process, thereby obtaining the reconfigurable broadband photodetector.

[0014] Furthermore, in step S102, the molybdenum disulfide sheet is directly peeled off to the silicon dioxide surface, and step S103 is skipped, proceeding directly to step S104.

[0015] Furthermore, during the fabrication and use of the device, an inert gas environment is used for material stripping, transfer, and device processing, and an encapsulation protective layer is set on the device surface; the encapsulation protective layer uses hexagonal boron nitride or oxide insulating material to improve the environmental stability and service life of the device.

[0016] This invention also provides an application of the above-mentioned black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector, characterized in that it includes:

[0017] Step S201: Project the light signal to be detected onto the photodetector, so that the photodetector generates a corresponding photocurrent output according to the incident light intensity, wavelength and polarization state;

[0018] Step S202: Based on the pre-established correspondence between gate voltage and photocurrent state, select the appropriate gate voltage so that the photodetector outputs the required photoresponse state.

[0019] Step S203: The multi-level photoresponse states corresponding to different gate voltages are used as optical information encoding values ​​or calculation weights. The input optical information is then processed in the backend, and the classification results are output.

[0020] Further, in step S203, the input image is resized and normalized, and a feature extraction network is used to convert the input image into a feature vector; the feature vector is processed and calculated by a fully connected neural network to output the recognition results of different target categories.

[0021] Furthermore, the multi-level photoresponse states are mapped to the connection weights between nodes in a fully connected neural network; by changing the gate voltage, the photoresponse states and connection weights corresponding to the photodetector are adjusted, so that the input optical information undergoes partial weighting and preprocessing before being input into the back-end processor.

[0022] In view of the above technical features, the present invention provides a black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector, its fabrication method, and its application, which have the following significant advantages compared with the prior art:

[0023] 1. This invention utilizes the narrow bandgap and wide spectral absorption range of black arsenic phosphorus, enabling the photodetector to produce significant responses to incident light at wavelengths of 520 nm, 638 nm, 1310 nm, 1550 nm, and 4050 nm, thereby achieving broad-spectrum detection from visible light and near-infrared to mid-infrared bands. The response to 4050 nm incident light allows the device to cover the mid-infrared molecular fingerprint band with wavelengths exceeding 4 μm, which can be used for gas molecule recognition, carbon emission monitoring, industrial leak early warning, and thermal scene sensing.

[0024] 2. This invention enables continuous control of carrier separation and transport states at heterojunctions by adjusting the gate voltage. Within a gate voltage range of −20 V to 40 V, a single device can achieve 15 stable and distinguishable photocurrent states, thereby improving the programmability of the photoresponse and providing a hardware foundation for multi-level optical information encoding, computational weight configuration, and front-end information preprocessing.

[0025] 3. By mapping multi-level photoresponse states to neural network connection weights, this invention enables front-end encoding, weighting, and classification processing of infrared images. In the infrared target recognition embodiment, the recognition accuracy for pedestrians, bicycles, cars, and other vehicles exceeds 99%. Therefore, this technical solution integrates broadband detection, multi-state control, and information processing functions in a single device, reducing the amount of raw data transmission and back-end computational burden. This helps reduce the power consumption and processing latency of the infrared sensing system, and improves system integration and intelligence. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of a preferred embodiment of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention;

[0027] Figure 2 This is a preferred embodiment of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention, showing the dark-state drain-source current-drain-source voltage characteristic curves under different gate voltage conditions;

[0028] Figure 3This is a preferred embodiment of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention, showing the light-drain-source current-drain-source voltage characteristic curves under different gate voltage conditions;

[0029] Figure 4 This is a preferred embodiment of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention, showing the output current curves of the device when the device is periodically irradiated with incident light of different wavelengths.

[0030] Figure 5 This is a preferred embodiment of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention, showing the short-circuit current or photocurrent curves output by the device when the gate voltage is changed.

[0031] Figure 6 This is a flowchart of a preferred embodiment of the method for fabricating a black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention.

[0032] Figure 7 This is a process demonstration diagram of a preferred embodiment of the fabrication method of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention.

[0033] Figure 8 This is a flowchart of a preferred embodiment of the application of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention.

[0034] Figure 9 This is an image processing flowchart of a preferred embodiment of the application of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention.

[0035] Figure 10 This is a system block diagram of another preferred embodiment of the application of the black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector of the present invention.

[0036] In the diagram: 10 - Black arsenic phosphorus molybdenum disulfide photodetector; 20 - Back-end processing unit;

[0037] 11-Silicon substrate, 12-Gate dielectric layer, 13-Molybdenum disulfide layer, 14-Black arsenic phosphorus layer, 15-First metal electrode, 16-Second metal electrode;

[0038] 21-Feature extraction network, 22-Classifier. Detailed Implementation

[0039] The present invention will be further described below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0040] Example 1: A black arsenic phosphorus disulfide molybdenum photodetector.

[0041] Please see Figure 1 This embodiment provides a black arsenic phosphorus molybdenum disulfide photodetector. In this embodiment, the black arsenic phosphorus molybdenum disulfide photodetector 10 includes, from bottom to top, a silicon substrate 11, a gate dielectric layer 12, a molybdenum disulfide layer 13, a black arsenic phosphorus layer 14, a first metal electrode 15, and a second metal electrode 16.

[0042] A silicon / silicon dioxide substrate is provided. The silicon substrate 11 is located at the bottom of the device and also serves as the back gate electrode. The silicon dioxide layer, as the gate dielectric layer 12, is disposed above the silicon substrate 11 to achieve electrical insulation and electric field coupling between the back gate electrode and the optoelectronic transport channel of the device.

[0043] Molybdenum disulfide layer 13 is disposed above gate dielectric layer 12, serving to form a van der Waals heterojunction with black arsenic phosphorus layer and participate in carrier transport. Molybdenum disulfide layer 13 and black arsenic phosphorus layer 14 form a van der Waals heterojunction with bandgap, and under the action of an applied gate voltage, the interface barrier, built-in electric field, and carrier separation and transport processes are adjusted. In this embodiment, the thickness of molybdenum disulfide layer is approximately 58 nm, and the bandgap is approximately 1.2 eV.

[0044] A black arsenic phosphorus layer 14 is disposed above and partially overlaps with a molybdenum disulfide layer 13. In the overlapping region, the black arsenic phosphorus and molybdenum disulfide form a van der Waals heterojunction. The black arsenic phosphorus layer uses b-As... 0.3 P 0.7 The material has a band gap of approximately 0.22 eV and is capable of absorbing incident light in the visible to mid-infrared (including wavelengths above 4 μm) range. In this embodiment, the thickness of the black arsenic phosphorus layer is approximately 79 nm.

[0045] The first metal electrode 15 and the second metal electrode 16 form electrical contacts with the molybdenum disulfide layer 13 and the black arsenic phosphorus layer 14, respectively, leading the molybdenum disulfide layer 13 and the black arsenic phosphorus layer 14 to the peripheral circuit for electrical signal testing and photocurrent collection. The first metal electrode 15 forms an electrical contact with the molybdenum disulfide layer 13, and the second metal electrode 16 forms an electrical contact with the black arsenic phosphorus layer 14. Both the first metal electrode 15 and the second metal electrode 16 are chromium-gold composite electrodes, consisting of a chromium layer disposed on the surface of the molybdenum disulfide layer 13 or the black arsenic phosphorus layer 14, and a gold layer disposed above the chromium layer.

[0046] When incident light irradiates the overlapping region of the black arsenic phosphorus / molybdenum disulfide heterojunction, the black arsenic phosphorus layer 14 absorbs the incident photons and generates photogenerated electron-hole pairs. The built-in electric field formed at the heterojunction interface causes the photogenerated electrons and photogenerated holes to undergo spatial separation—the separated charge carriers are transported along the black arsenic phosphorus layer 14 and the molybdenum disulfide layer 13 to the corresponding first metal electrode 15 and second metal electrode 16 (depending on the band alignment direction), thereby forming a measurable photocurrent signal.

[0047] When the gate voltage applied to the back gate electrode of the silicon substrate 11 is changed, the electron concentration, Fermi level position, and band arrangement of the black arsenic phosphorus / molybdenum disulfide heterointerface in the molybdenum disulfide layer 13 can be adjusted, thereby altering the photoresponse characteristics of the device under different gate voltages. Specifically, when a positive gate voltage is applied, electrons accumulate in the molybdenum disulfide layer 13, enhancing the potential difference and built-in electric field across the heterointerface, thus promoting the separation and collection of photogenerated electron-hole pairs and increasing the photocurrent output of the device. When a negative gate voltage is applied, electrons in the molybdenum disulfide layer 13 are gradually depleted, reducing the potential difference and built-in electric field at the heterointerface, suppressing the separation and transport of photogenerated carriers, and correspondingly reducing the photocurrent output of the device. Therefore, by adjusting the gate voltage stepwise, the carrier transport state at the heterointerface can be continuously changed, allowing the same photodetector to generate multiple stable and distinguishable photocurrent output states under the same or different incident light conditions.

[0048] Please see Figure 2 The fabricated photodetector was placed in a dark-state testing environment, with the two metal electrodes (Cr / Au) connected to the source and drain test terminals, respectively, and the silicon substrate connected to the gate test terminal. The drain-source voltage Vds was scanned under gate voltages Vg of −40 V, −30 V, −20 V, −10 V, 0 V, 10 V, 20 V, 30 V, and 40 V to obtain the dark-state drain-source current-drain-source voltage characteristic curves under different gate voltage conditions.

[0049] Test results show that the IV curves of the device exhibit significant nonlinear characteristics under different gate voltage conditions. Furthermore, the device displays different current variation patterns in the forward and reverse drain-source bias regions, indicating that the black arsenic phosphorus / molybdenum disulfide heterojunction interface forms an effective carrier transport barrier, and the device possesses significant rectification characteristics. Under a specific test condition, when Vg = 0 V and Vds = 1.5 V, the rectification ratio of the device is approximately 50, further demonstrating the formation of an effective heterojunction transport barrier within the device.

[0050] The device also exhibits asymmetric gate voltage regulation characteristics. Compared to when Vg = 0 V, the device current changes only to a relatively limited extent after a positive gate voltage is applied; however, as the gate voltage is adjusted in the negative direction, the device current is significantly suppressed, and as the absolute value of the negative gate voltage increases, the device current at the same drain-source voltage gradually decreases.

[0051] The reason for the above phenomenon is that MoS2 is an electronic semiconductor material. When a positive gate voltage is applied, electrons accumulate further in the MoS2 layer, the Fermi level shifts upward, and the carrier supply and built-in electric field at the heterojunction are enhanced. However, since MoS2 is already in a relatively significant electron accumulation state under zero gate voltage and non-negative gate voltage conditions, the regulatory effect of further increasing the positive gate voltage on the dominant carrier transport process gradually weakens. Therefore, compared with Vg = 0 V, the device current only changes to a limited extent in the positive gate voltage region. When a negative gate voltage is applied, electrons in the MoS2 layer are gradually depleted, the Fermi level shifts downward, the carrier supply at the heterojunction is reduced, and the potential difference and built-in electric field at the black arsenic phosphorus / molybdenum disulfide heterojunction are weakened. At the same time, the reduced electron supply also inhibits the filling of interface defect states and the defect-assisted carrier injection process, thereby significantly reducing the device current.

[0052] The above results demonstrate that the gate voltage can regulate the carrier concentration in the MoS2 layer, the barrier morphology of the heterostructure interface, and the main carrier transport paths within the device. Specifically, the negative gate voltage has a significantly stronger suppressive effect on device current than the positive gate voltage has a stronger enhancing effect.

[0053] Please see Figure 3 The overlapping region of the black arsenic phosphorus / molybdenum disulfide heterojunction was irradiated with a laser with a wavelength of 520 nm, and the photo-drain-source current-drain-source voltage characteristics of the device were tested under gate voltages Vg of −40 V, −30 V, −20 V, −10 V, 0 V, 10 V, 20 V, 30 V and 40 V.

[0054] When 520 nm incident light irradiates the heterojunction region, the black arsenic phosphorus layer and the MoS2 layer absorb the incident photons and generate photogenerated electron-hole pairs. Due to the band dislocation between the two materials, the effective heterojunction potential formed at the heterojunction interface can drive the photogenerated electrons and holes to spatially separate. The separated electrons and holes are transported along the MoS2 layer and the black arsenic phosphorus layer to the corresponding metal electrodes, respectively, thereby forming a photocurrent in the external circuit of the device and generating a measurable short-circuit current Isc and open-circuit voltage Voc.

[0055] Test results show that the IV curve of the device under illumination exhibits a gate voltage dependence trend consistent with that under dark conditions. When the gate voltage increases from 0 V to 40 V, the short-circuit current Isc and open-circuit voltage Voc of the device remain basically stable, indicating that within the non-negative gate voltage range, MoS2 is already in a state of significant electron accumulation, and the effective potential of the black arsenic phosphorus / molybdenum disulfide heterojunction and the separation and collection capabilities of photogenerated carriers are close to saturation. In contrast, under negative gate voltage, both Isc and Voc of the device decrease significantly and gradually decrease with the increase of the absolute value of the negative gate voltage. When Vg = −40 V, both Isc and Voc tend to disappear, indicating that at this point, the effective potential of the heterojunction is insufficient to effectively dominate the photoresponse of the device.

[0056] The reason for the above phenomenon is that the negative gate voltage gradually depletes the electrons in the MoS2 layer and shifts the Fermi level of MoS2 downward, thereby changing the band bending and potential distribution of the heterojunction. As the heterojunction potential difference and built-in electric field weaken, the separation efficiency and collection efficiency of photogenerated carriers decrease accordingly. At Vg = −40 V, the electrical and photoelectric responses of the device are no longer mainly controlled by the black arsenic phosphorus / molybdenum disulfide heterojunction, but are gradually limited by the metal-semiconductor contact barrier.

[0057] In summary, the carrier supply and photogenerated carrier separation capabilities at the heterojunction gradually approach saturation within the non-negative gate voltage region. Conversely, the negative gate voltage depletes electrons in the MoS2 layer, weakens the effective potential of the heterojunction, and suppresses carrier transport at the interface, causing the device to gradually transition from a normal photoresponse state to a suppressed photoresponse state. Therefore, by changing the gate voltage, the relative dominance of the heterojunction and metal-semiconductor contact in the device can be adjusted, enabling reconfigurable control of the device's photoelectric response state.

[0058] Please see Figure 4 The overlapping region of the black arsenic phosphorus / molybdenum disulfide heterojunction of the device was illuminated with incident light at wavelengths of 520 nm, 638 nm, 1310 nm, 1550 nm, and 4050 nm, respectively. The incident light was periodically turned on and off, and the change of the device's output current over time was measured. During the test, a predetermined drain-source voltage could be applied between the two metal electrodes, or the short-circuit current Isc of the device could be measured under zero drain-source bias conditions.

[0059] Test results show that under the above wavelength conditions, the device output current changes significantly when the incident light is turned on; when the incident light is turned off, the device output current returns to near its initial state. After repeatedly turning the incident light on and off, the device can still output repetitive and distinguishable photocurrent signals.

[0060] Specifically, 520 nm and 638 nm correspond to the visible light band, 1310 nm and 1550 nm to the near-infrared band, and 4050 nm to the mid-infrared band. The device generates significant photocurrent responses under these wavelength conditions, indicating that the black arsenic phosphorus / molybdenum disulfide heterojunction can achieve broad-spectrum detection from visible and near-infrared to mid-infrared light. In particular, the device can still generate a repeatable and identifiable photocurrent signal under 4050 nm incident light, indicating that the detection wavelength of the device can extend to 4.05 μm and cover the mid-infrared region with wavelengths exceeding 4 μm.

[0061] Since mid-infrared light with wavelengths exceeding 4 μm contains vibrational absorption characteristics of various gas molecules, the device can be applied to gas molecule recognition, carbon emission monitoring, industrial gas leak early warning, thermal scene perception, and mid-infrared optical detection.

[0062] Please see Figure 5 Under 1550 nm incident light irradiation, by changing the gate voltage applied to the silicon substrate, the short-circuit current or photocurrent of the device under different gate voltage conditions is measured, and multiple stable and mutually distinguishable photoresponse states are extracted from a single device.

[0063] In one specific embodiment, 15 gate voltage states are selected within a gate voltage range of −20 V to 40 V, and the photocurrent output under each gate voltage condition is measured. The photocurrent corresponding to each gate voltage is sequentially defined as the first photoresponse state to the fifteenth photoresponse state. Different photoresponse states have different photocurrent output values ​​and can be selected and reconstructed by changing the gate voltage.

[0064] As the gate voltage is adjusted from negative to positive, the electron concentration in the MoS2 layer gradually increases, altering the barrier morphology, built-in electric field, and carrier transport channels at the heterojunction, thus causing a gradual change in the device's output photocurrent. In the negative gate voltage region, electrons in the MoS2 layer are depleted, significantly suppressing photogenerated carrier separation and transport at the heterojunction, resulting in a relatively small output photocurrent. As the gate voltage gradually increases, the effective potential of the heterojunction and its carrier collection capability enhance, leading to a corresponding increase in the device's output photocurrent. In the higher non-negative gate voltage region, due to the accumulation of electrons in the MoS2 layer and the gradual approaching saturation of the heterojunction's carrier separation capability, the current variation between adjacent photoresponse states can gradually decrease, but the different states remain distinguishable.

[0065] The aforementioned 15 photoresponse states can be used as different optical information encoding values, grayscale values, storage states, or calculation weights. When the device is used for optical information processing, the corresponding gate voltage is selected based on the pre-established correspondence between the gate voltage and the photoresponse state.

[0066] Example 2: A method for preparing a black arsenic phosphorus disulfide molybdenum photodetector.

[0067] Please see Figure 6 and Figure 7 This embodiment provides a method for preparing a black arsenic phosphorus disulfide molybdenum photodetector as described in Embodiment 1, comprising the following steps:

[0068] Step S101 involves cleaning the silicon substrate 11 and the gate dielectric layer 12 to remove organic contaminants, particles, and other impurities from the material surface. The silicon / silicon dioxide substrate is cleaned sequentially with acetone, isopropanol, and deionized water, and then dried with nitrogen gas.

[0069] Step S102: Molybdenum disulfide flakes and black arsenic phosphorus flakes are obtained by mechanical peeling.

[0070] MoS2 flakes are obtained from the MoS2 bulk material using a mechanical peeling method, and MoS2 flakes with suitable thickness, smooth surface and no obvious cracks are selected by optical microscopy before proceeding to step S103.

[0071] In other embodiments, the MoS2 sheet can also be directly mechanically peeled off to the silica surface. When direct mechanical peeling off to the silica surface is selected, proceed to step S104.

[0072] Step S103: Transfer the molybdenum disulfide sheet to the silicon dioxide surface.

[0073] MoS2 wafers are picked up using PDMS or other transfer media and transferred to predetermined positions on the surface of a silicon / silicon dioxide substrate under a micro-alignment system.

[0074] Step S104 involves optically positioning the black arsenic phosphorus flakes and molybdenum disulfide flakes, transferring the black arsenic phosphorus flakes over the molybdenum disulfide flakes to partially overlap and form a van der Waals heterojunction. Black arsenic phosphorus flakes are obtained from the bulk material using a mechanical exfoliation method, and flakes with intact surfaces and suitable thickness are selected using an optical microscope. The black arsenic phosphorus flakes are picked up using PDMS or other transfer media and transferred over the MoS2 flakes under a micro-alignment system, resulting in partial overlap and the formation of a van der Waals heterojunction. To reduce the possibility of oxidation or performance degradation of the black arsenic phosphorus material in air, the exfoliation, screening, and transfer processes of the black arsenic phosphorus flakes are preferably performed in a nitrogen or argon-protected environment.

[0075] In step S105, electrode patterns are formed by photolithography or electron beam exposure, and chromium / gold metal layers are deposited by metal evaporation or sputtering. After forming the van der Waals heterostructure, photoresist or electron beam exposure resist is spin-coated onto the device surface, and electrode patterns are formed by photolithography or electron beam exposure. The two electrode patterns are respectively located in the non-overlapping regions of the MoS2 layer and the black arsenic phosphorus layer, so that the two electrodes subsequently formed establish electrical contacts with the MoS2 layer and the black arsenic phosphorus layer, respectively. Cr layer and Au layer are deposited sequentially by evaporation.

[0076] In step S106, metal electrodes are formed by a stripping process, which are respectively connected to the black arsenic phosphorus layer 14 and the molybdenum disulfide layer 13, thereby obtaining a reconfigurable broadband photodetector. After metal deposition is completed, the sample is placed in a stripping solution for metal stripping to remove the metal from the non-electrode regions, thereby forming two Cr / Au composite electrodes connected to the MoS2 layer and the black arsenic phosphorus layer, respectively.

[0077] During device fabrication and use, an inert gas environment can be used for material stripping, transfer, and device processing. A protective encapsulation layer can be applied to the device surface to reduce the likelihood of oxidation and performance degradation of black arsenic phosphorus. The protective encapsulation layer can be made of hexagonal boron nitride, alumina, hafnium oxide, polymer insulating materials, or other materials capable of isolating water and oxygen to improve the device's environmental stability and lifespan.

[0078] Example 3: An application based on a black arsenic phosphorus disulfide molybdenum photodetector.

[0079] Please see Figure 8 and Figure 9 This embodiment provides an application of the black arsenic phosphorus molybdenum disulfide photodetector described in Embodiment 1, specifically a classification method, including the following steps:

[0080] Step S201: The light signal to be detected is projected onto the photodetector, so that the photodetector generates a corresponding photocurrent output according to the incident light intensity, wavelength and polarization state.

[0081] The infrared light signal or infrared image to be detected is projected onto the black arsenic phosphorus molybdenum disulfide photodetector 10. This black arsenic phosphorus molybdenum disulfide photodetector 10 adopts the structure described in Example 1. When incident light irradiates the overlapping region of the black arsenic phosphorus / molybdenum disulfide heterojunction, the black arsenic phosphorus layer 14 absorbs the incident photons and generates photogenerated electron-hole pairs. The built-in electric field formed at the heterojunction interface causes the photogenerated electrons and holes to spatially separate. The separated charge carriers are transported along the black arsenic phosphorus layer 14 and the molybdenum disulfide layer 13 to the corresponding metal electrodes, thereby forming a measurable photocurrent signal. The photodetector generates a corresponding photocurrent output based on the incident light intensity, wavelength, and polarization state.

[0082] Step S202: Based on the pre-established correspondence between gate voltage and photocurrent state, select the appropriate gate voltage so that the photodetector outputs the required photoresponse state.

[0083] By changing the gate voltage applied to the back gate electrode, the electron concentration, Fermi level position, and band arrangement of the black arsenic phosphorus / molybdenum disulfide heterointerface in the molybdenum disulfide layer 13 can be adjusted. When a positive gate voltage is applied, electrons accumulate in the molybdenum disulfide layer 13, enhancing the potential difference and built-in electric field across the heterointerface. This promotes the separation and collection of photogenerated electron-hole pairs, increasing the photocurrent output of the device. When a negative gate voltage is applied, electrons in the molybdenum disulfide layer 13 are gradually depleted, reducing the potential difference and built-in electric field at the heterointerface. This suppresses the separation and transport of photogenerated carriers, correspondingly reducing the photocurrent output of the device. By progressively adjusting the gate voltage, the carrier transport state at the heterointerface can be continuously changed, allowing the same photodetector to generate multiple stable and distinguishable photocurrent output states under the same or different incident light conditions.

[0084] Based on a pre-established correspondence between gate voltage and photocurrent state, a corresponding gate voltage is selected to enable the photodetector to output the desired photoresponse state. In one specific embodiment, selecting 15 gate voltage states within a gate voltage range of −20 V to 40 V yields 15 distinguishable photoresponse states.

[0085] Step S203: The multi-level photoresponse states corresponding to different gate voltages are used as optical information encoding values ​​or calculation weights. The input optical information is then processed in the backend (weighting, filtering and preprocessing) to output the classification results.

[0086] The multi-level photoresponse states corresponding to different gate voltages are used as optical information encoding values ​​or calculation weights to weight, filter, and preprocess the input optical information. The aforementioned 15 photoresponse states can be used as different optical information encoding values, grayscale values, storage states, or calculation weights, respectively. When the device is used for optical information processing, the appropriate gate voltage can be selected according to the characteristics of the optical signal to be processed, causing the device to output a predetermined photoresponse state, thereby realizing multi-level modulation and front-end processing of the optical signal.

[0087] In the infrared image recognition process, the infrared image to be recognized is first resized and normalized. In one specific embodiment, the infrared image is resized to 224×224 pixels. Subsequently, a feature extraction network is used to extract image features. In one specific embodiment, a ResNet18 network is used for feature extraction, converting each infrared image into a 512-dimensional feature vector. The multi-level photoresponse states are mapped to the connection weights Rij between nodes in a fully connected neural network, where i represents the previous layer node and j represents the next layer node. Before weight mapping, the multi-level photoresponse states can be normalized, linearly scaled, piecewise mapped, or calibrated using a lookup table to establish a correspondence between the actual photocurrent output by the device and the weight values ​​required by the neural network. The feature vectors are processed and calculated by the fully connected neural network to output recognition results for different target categories.

[0088] By changing the gate voltage, the photoresponse state and connection weights corresponding to the photodetector are adjusted, allowing the input optical information to undergo partial weighting and preprocessing before being input to the back-end processor. Besides infrared image classification, the photoresponse state can also be used for tasks such as optical information encoding, image preprocessing, edge extraction, target recognition, and multi-dimensional optical information perception.

[0089] Under different gate voltage conditions, multiple photoresponse current values ​​differ, thus the mapped connection weights Rij also differ. This method can be applied to target recognition scenarios in infrared images. In one specific implementation, the fully connected neural network has an input feature dimension of 512, a hidden layer with 64 nodes, and an output layer with 4 nodes, corresponding to target categories such as pedestrians, bicycles, cars, and other vehicles. Under appropriate dataset partitioning and network training conditions, after mapping 15 photoresponse states to fully connected neural network weights, the infrared image recognition accuracy can reach approximately 99%.

[0090] Example 4: Another application based on a black arsenic phosphorus disulfide molybdenum photodetector.

[0091] Please see Figure 10 This embodiment provides another application based on the black arsenic phosphorus disulfide photodetector, specifically an in-sensor computing system for implementing the classification method of Embodiment 3. This computing system includes a black arsenic phosphorus disulfide photodetector 10 and a back-end processing unit 20.

[0092] The black arsenic phosphorus molybdenum disulfide photodetector 10 adopts the black arsenic phosphorus molybdenum disulfide photodetector structure described in Example 1. Its core is a black arsenic phosphorus / molybdenum disulfide van der Waals heterojunction, and it has a back gate electrode structure. This detector generates multiple distinguishable photoresponse states under different gate voltages, which are used to represent optical information encoding values ​​or calculate weights.

[0093] By changing the gate voltage applied to the back gate electrode, the electron concentration, Fermi level position, and band arrangement of the black arsenic phosphorus / molybdenum disulfide heterointerface in the molybdenum disulfide layer 13 can be adjusted. When a positive gate voltage is applied, electrons accumulate in the molybdenum disulfide layer 13, enhancing the potential difference and built-in electric field across the heterointerface, thereby promoting the separation and collection of photogenerated electron-hole pairs and increasing the photocurrent output of the device. When a negative gate voltage is applied, electrons in the molybdenum disulfide layer 13 are gradually depleted, reducing the potential difference and built-in electric field at the heterointerface, suppressing the separation and transport of photogenerated carriers, and correspondingly reducing the photocurrent output of the device. By adjusting the gate voltage stepwise, the carrier transport state of the heterointerface can be continuously changed, allowing the same photodetector to generate multiple stable and distinguishable photocurrent output states under the same or different incident light conditions. In one specific embodiment, selecting 15 gate voltage states within the range of −20 V to 40 V yields 15 distinguishable photoresponse states. The different optical response states mentioned above can be used as optical information encoding values, weight values, or storage states, respectively, to realize multi-level modulation and front-end processing of optical signals.

[0094] The back-end processing unit 20 receives signals from the black arsenic phosphorus molybdenum disulfide photodetector 10. The back-end processing unit 20 includes a feature extraction network 21 and a classifier 22.

[0095] The feature extraction network 21 is used to extract features from the input light signal or image to obtain feature vectors. In this embodiment, the feature extraction network 21 is a ResNet18 network, which is used to extract features from the input signal to obtain feature vectors. The residual connection structure effectively extracts target contours, edges, grayscale distribution, texture, and structural features from the image. In other embodiments, the feature extraction network 21 can also be a convolutional neural network or a spiking neural network. During infrared image recognition, the infrared image to be recognized is first resized and normalized to 224×224 pixels, and then the ResNet18 network is used for feature extraction, converting each infrared image into a 512-dimensional feature vector.

[0096] The classifier 22 is a three-layer fully connected network, including an input layer, a hidden layer, and an output layer. In one specific embodiment, the input feature dimension of the fully connected neural network is 512, the hidden layer includes 64 nodes, and the output layer includes 4 nodes. The connection weights of the classifier 22 are determined by mapping multiple photoresponse states output by the photodetector 10 under different gate voltages. Before weight mapping, the multi-level photoresponse states can be normalized, linearly scaled, piecewise mapped, or calibrated using a lookup table to establish a correspondence between the actual photocurrent output by the device and the weight values ​​required by the neural network. The classifier 22 is used to output classification results based on the feature vector and the connection weights, and can output recognition results for different target categories, including pedestrians, bicycles, cars, and other vehicles, but the target classification category can be adjusted according to actual application requirements.

[0097] During system operation, the black arsenic phosphorus disulfide photodetector 10 outputs multiple distinguishable photoresponse states under different gate voltages, which are then transmitted to the back-end processing unit 20. The feature extraction network 21 extracts features from the input signal to obtain feature vectors. The connection weights of the classifier 22 are determined by these multiple photoresponse states. Subsequently, the classifier 22 outputs the classification result based on the feature vectors and connection weights. By changing the gate voltage, the bandgap arrangement and carrier transport characteristics of the heterojunction can be adjusted, thereby obtaining different photoresponse states under different gate voltages. By changing the gate voltage, the corresponding photoresponse state of the device and the neural network weights can be adjusted, allowing image features to undergo partial weighting and preprocessing before being input to the back-end processor. After calculation by the fully connected neural network, recognition results for different target categories can be output.

[0098] This system can be applied to target recognition scenarios in infrared images, and the output classification results can be target category labels (such as pedestrians, bicycles, cars, and other vehicles). The in-sensor computation method is not limited to using a ResNet18 network; convolutional neural networks, spiking neural networks, or other network structures with feature extraction capabilities can also be used. In addition to infrared image classification, the photoresponse states can also be used for tasks such as optical information encoding, image preprocessing, edge extraction, target recognition, and multi-dimensional optical information perception. Under appropriate dataset partitioning and network training conditions, after mapping the 15 photoresponse states to fully connected neural network weights, the infrared image recognition accuracy can reach approximately 99%.

[0099] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector, characterized in that, The device includes a substrate, a gate dielectric layer, a molybdenum disulfide layer, a black arsenic phosphorus layer, and two metal electrodes. The substrate is a silicon substrate and also serves as the back gate electrode. The gate dielectric layer is disposed above the silicon substrate to achieve electrical insulation and electric field coupling between the back gate electrode and the photoelectric transport channel of the device. The molybdenum disulfide layer is disposed above the gate dielectric layer. The black arsenic phosphorus layer is disposed above the molybdenum disulfide layer and partially overlaps with it. A black arsenic phosphorus molybdenum disulfide van der Waals heterojunction is formed in the overlapping region of the black arsenic phosphorus layer and the molybdenum disulfide layer, serving as the main photosensitive region and photogenerated carrier separation region of the device. The two metal electrodes form electrical contacts with the molybdenum disulfide layer and the black arsenic phosphorus layer, respectively, to apply source-drain bias voltage, form carrier transport channels, and collect the photocurrent generated by the device.

2. The black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector according to claim 1, characterized in that, The black arsenic phosphorus layer is made of b-As0.3P0.7 material with a band gap of 0.22 eV, which is particularly responsive to mid-infrared light with wavelengths exceeding 4 μm. The molybdenum disulfide layer has a band gap of 1.2 eV and is used to form the van der Waals heterointerface with the black arsenic phosphorus layer with band misalignment. Under the action of the applied gate voltage, it regulates the interface barrier, the built-in electric field, and the separation and transport of charge carriers.

3. The black arsenic phosphorus disulfide molybdenum heterojunction reconfigurable broadband photodetector according to claim 1, characterized in that, The metal electrode is a chromium-gold composite electrode, comprising a chromium layer and a gold layer; wherein the chromium layer is disposed on the surface of the molybdenum disulfide layer and the black arsenic phosphorus layer to improve the adhesion between the electrode and the black arsenic phosphorus or molybdenum disulfide; the gold layer is disposed above the chromium layer to reduce the electrode resistance and improve the electrical signal transmission capability; the two metal electrodes are respectively located at both ends of the heterostructure, so that the black arsenic phosphorus layer, the molybdenum disulfide layer and the van der Waals heterojunction formed by their overlap together constitute a photoelectric transport channel.

4. A method for fabricating a black arsenic phosphorus disulfide heterojunction reconfigurable broadband photodetector as described in any one of claims 1 to 3, characterized in that, Includes the following steps: Step S101: Clean the substrate and the gate dielectric layer to remove organic contaminants, particles and other impurities from the material surface; Step S102: Molybdenum disulfide flakes and black arsenic phosphorus flakes are obtained by mechanical peeling. Step S103: Transfer the molybdenum disulfide sheet onto the silicon dioxide surface; Step S104: Optically position the black arsenic phosphorus sheet and the molybdenum disulfide sheet, transfer the black arsenic phosphorus sheet above the molybdenum disulfide sheet, so that the two partially overlap and form the van der Waals heterojunction; Step S105: Electrode patterns are formed by photolithography or electron beam exposure, and a chromium-gold metal layer is deposited by metal evaporation or sputtering. Step S106: The metal electrodes connected to the black arsenic phosphorus layer and the molybdenum disulfide layer are formed by a stripping process, thereby obtaining the reconfigurable broadband photodetector.

5. The preparation method according to claim 4, characterized in that, In step S102, the molybdenum disulfide sheet is directly peeled off to the silicon dioxide surface, and step S103 is skipped, proceeding directly to step S104.

6. The preparation method according to claim 4, characterized in that, During device fabrication and use, an inert gas environment is used for material stripping, transfer, and device processing, and an encapsulation protective layer is applied to the device surface. The encapsulation protective layer uses hexagonal boron nitride or oxide insulating material to improve the environmental stability and service life of the device.

7. An application of a reconfigurable broadband photodetector based on a black arsenic phosphorus disulfide heterojunction as described in any one of claims 1 to 3, characterized in that, include: Step S201: Project the light signal to be detected onto the photodetector, so that the photodetector generates a corresponding photocurrent output according to the incident light intensity, wavelength and polarization state; Step S202: Based on the pre-established correspondence between gate voltage and photocurrent state, select the appropriate gate voltage so that the photodetector outputs the required photoresponse state. Step S203: The multi-level photoresponse states corresponding to different gate voltages are used as optical information encoding values ​​or calculation weights. The input optical information is then processed in the backend, and the classification results are output.

8. The application according to claim 7, characterized in that, In step S203, the input image is resized and normalized, and a feature extraction network is used to convert the input image into a feature vector; the feature vector is processed and calculated by a fully connected neural network to output the recognition results of different target categories.

9. The application according to claim 8, characterized in that, The multi-level photoresponse states are mapped to the connection weights between nodes in a fully connected neural network; by changing the gate voltage, the photoresponse states and connection weights corresponding to the photodetector are adjusted, so that the input optical information is partially weighted and preprocessed before being input to the back-end processor.