Primary motion detector based on PN junction channel, fabrication method and sensor

Abstract

This invention relates to the field of motion detection technology, and discloses a primary motion detector based on a PN junction channel, its fabrication method, and a sensor. The primary motion detector includes a first phototransistor and a second phototransistor. The first phototransistor is configured to operate under a source-drain forward bias, and its photocurrent decay process during turn-off has a first fast decay time constant and a first slow decay time constant. The second phototransistor is configured to operate under a source-drain reverse bias, and its photocurrent during turn-off has a second fast decay time constant and a second slow decay time constant, wherein the first fast decay time constant is greater than the second fast decay time constant. When a light pulse moves along the direction from the first phototransistor to the second phototransistor, a directional current pulse can be detected at the output of the primary motion detector. This invention achieves high-efficiency and low-power discrimination of the direction of light signals.

Description

Technical Field

[0001] This invention relates to the field of motion detection technology, specifically to a primary motion detector based on a PN junction channel, its fabrication method, and a sensor. Background Technology

[0002] Since the last century, people have continuously drawn inspiration from the visual systems of arthropods and insects, inventing specialized visual sensors and small, efficient biomimetic visual sensors. Although the fly's brain and visual system are small and relatively simple, they excel at processing crucial visual information. Their ability to recognize direction comes from H1 neurons, a type of directional movement-sensitive cell known in the animal world; the primary motion detector unit is a subunit of the H1 neuron. For example... Figure 1 As shown, the principle of the primary motion detector unit can be explained by the Hassenstein–Reichardt (HR) model, which consists of three parts: transmission, delay, and product. Inputs A and B represent two adjacent box units that are selectively activated by optical stimulation of two photoreceptor cells R1 and R6 within a single ommatidia. Hassenstein and Reichardt inferred that the nonlinear interaction between the two filtered signals is a multiplier, and this primary motion detector unit is assumed to sense motion in the preferred direction of the H1 neuron and drive the H1 neuron with positive polarity.

[0003] The sequential illumination R1→R6 of the photoreceptor cells drives two box units A→B, transmitting an excitatory signal (+) to the H1 neuron. The left arm describes a second-order low-pass (LP) filter, and the boxes in the right arm describe a first-order high-pass (HP) filter. The left and right arms are connected by a multiplier. Figure 1 In (a), the two horizontal lines at the bottom are the channels of H1 neurons, and "+" represents potential stimulation. Figure 1 The amplitude-frequency response curves of the two filter outputs under different circuit conditions were also shown. Figure 1 (b) is a typical second-order low-pass filter. Figure 1 (c) in the diagram represents the resonant state of the low-pass filter. Figure 1 (d) in the diagram is a typical second-order high-pass filter. Figure 1 (e) in the figure represents the high-pass band-stop filter characteristic.

[0004] Currently, most bionic vision sensors are composed of complex transistor circuits using traditional silicon-based processes, resulting in high manufacturing costs and complex manufacturing processes. These machine vision technologies based on the traditional von Neumann architecture rely on independent computing units, such as optical sensing units, storage units, and processing units. Data transmission leads to energy consumption issues during processing. One optimization solution is to replace traditional silicon-based transistors with transistors that enable in-sensor computing and simplify circuit design to reduce energy consumption. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides an elementary motion detector (EMD) based on a PN junction channel, its fabrication method, and a sensor. Mimicking biological vision, it is applied to basic orientation recognition. Inspired by the visual nerves of flies, this invention utilizes the photoelectric and storage characteristics of specific phototransistors to design a phototransistor based on a vertical PN junction channel. Relying on the electrical characteristics of the PN junction under both forward and reverse biases, the interconnection method of the PN junction channel phototransistors is designed to form the elementary motion detector circuit. This invention offers advantages such as low power consumption and circuit miniaturization. The elementary motion detector based on a PN junction channel phototransistor proposed in this invention provides a new approach for future biomimetic vision sensors.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides a primary motion detector based on a PN junction channel, comprising: a first phototransistor M1 and a second phototransistor M2; the channels of the first and second phototransistors are both based on a vertical PN junction formed by molybdenum disulfide and tungsten selenide, and the first and second phototransistors are connected in series; the first phototransistor is configured to operate under a source-drain forward bias, and the photocurrent attenuation process during turn-off has a first fast decay time constant and a first slow decay time constant; the second phototransistor is configured to operate under a source-drain reverse bias, and the photocurrent during turn-off has a second fast decay time constant and a second slow decay time constant, wherein the first fast decay time constant is greater than the second fast decay time constant; when a light pulse moves along the direction from the first phototransistor to the second phototransistor, a directional current pulse can be detected at the output terminal of the primary motion detector.

[0008] In one embodiment, the source-drain forward bias voltage is greater than 0V, and the source-drain reverse bias voltage is less than 0V.

[0009] In one embodiment, the first fast decay time constant and the second fast decay time constant are in the range of 0.2 seconds to 0.8 seconds, and the first slow decay time constant and the second slow decay time constant are greater than 9.5 seconds.

[0010] In one embodiment, the molybdenum disulfide and tungsten selenide are stacked vertically by van der Waals forces to form a Type-II band-aligned vertical PN junction channel.

[0011] In one embodiment, the first phototransistor and the second phototransistor further include: a substrate, a back gate electrode located on the substrate, a dielectric layer covering the back gate electrode, a source metal electrode electrically connected to a molybdenum disulfide layer, and a drain metal electrode electrically connected to a tungsten selenide layer; the tungsten selenide layer is located on the dielectric layer, and the molybdenum disulfide layer is located on the tungsten selenide layer.

[0012] In one embodiment, the drain of the first phototransistor is electrically connected to the source of the second phototransistor, and the output of the primary motion detector is a measurement node between the drain of the second phototransistor and the ground terminal.

[0013] In one embodiment, the photocurrent attenuation curves of the first and second phototransistors for the light pulse are given by the formula. Fit, where, For fast decay time constant, The slow decay time constant, This represents the reference current in the dark state. Indicates the fast decay coefficient. Indicates the slow decay coefficient. Indicates time, This represents photocurrent.

[0014] Secondly, the present invention provides a method for preparing a primary motion detector as described in any embodiment of the first aspect, comprising:

[0015] A back gate electrode and a dielectric layer covering the back gate electrode are deposited on a substrate;

[0016] Tungsten selenide material is transferred to the surface of the dielectric layer to form a tungsten selenide layer, and molybdenum disulfide material is transferred onto the tungsten selenide layer to form a molybdenum disulfide layer. A vertical PN junction channel is formed by photolithography and development process.

[0017] The source metal electrode and drain metal electrode of the first phototransistor and the second phototransistor are respectively fabricated on the vertical PN junction channel by photolithography and development process.

[0018] The drain of the first phototransistor is electrically interconnected with the source of the second phototransistor.

[0019] In one embodiment, the transfer of tungsten selenide material to the surface of the dielectric layer specifically includes:

[0020] (1) Apply 15% anisole solvent to tungsten selenide on a sapphire substrate and dry at 90°C for two minutes;

[0021] (2) Soak the homogenized tungsten selenide in a 1 mol / L potassium hydroxide solution at 70°C for 8 minutes;

[0022] (3) Place the tungsten selenide in deionized water for peeling. After the tungsten selenide peels off the sapphire substrate, attach the dielectric layer of the target substrate to the tungsten selenide. After successful transfer, dry the target substrate at 50 degrees Celsius for 5 minutes.

[0023] Thirdly, the present invention provides a biomimetic vision sensor comprising an array of a plurality of primary motion detectors as described in any embodiment of the first aspect, arranged in a predetermined space, each of the primary motion detectors being configured to detect light motion information in a specific direction.

[0024] This invention relates to a primary motion detector based on a phototransistor, whose core component employs a tungsten selenide (WSe2)-molybdenum disulfide (MoS2) PN junction channel transistor. This detector achieves high-efficiency, low-power discrimination of optical signal direction by simply cascading two devices and changing the power supply conditions of a single device, demonstrating significant application potential in intelligent optical sensing systems. The core advantages of this detector are reflected in the following aspects:

[0025] (1) High integration and simplification at the device and circuit level: Traditional optical signal direction recognition systems usually require multiple modules such as discrete photodetectors, signal amplification circuits, analog-to-digital converters, and back-end digital processors. This invention utilizes the characteristics of the integrated phototransistor to integrate optical sensing and preliminary logic operation functions at the single device level, simplifying the system architecture, reducing the complexity of peripheral circuits, and lowering the system size and manufacturing cost.

[0026] (2) Low power consumption characteristics: The two-dimensional material used has excellent low static power consumption characteristics, which is suitable for portable and IoT devices that are extremely sensitive to power consumption; the integrated sensing and computing device avoids the large amount of energy consumption caused by long-distance transmission and complex digital processing of the original optical signal. Attached Figure Description

[0027] Figure 1 This is a basic block diagram of a primary motion detector based on the Hassenstein-Reichart model.

[0028] Figure 2 This is a schematic diagram of the primary motion detector circuit of the present invention.

[0029] Figure 3 This is a process flow diagram of the phototransistor of the present invention.

[0030] Figure 4 This is a schematic diagram of the connection method of the present invention.

[0031] Figure 5 The graph shows the test results of the phototransistor of the present invention under different biases.

[0032] Figure 6 This is a schematic diagram of the data fitting process for slow relaxation of photocurrent.

[0033] Figure 7 This is a schematic diagram of the data fitting process for fast relaxation of photocurrent.

[0034] Figure 8 This is a schematic diagram of the preferred direction simulation circuit and its results for this invention.

[0035] Figure 9 This is a schematic diagram of the simulation circuit for the non-preferred direction of the present invention and its results. Detailed Implementation

[0036] A preferred embodiment of the present invention will now be described in detail with reference to the accompanying drawings.

[0037] This invention utilizes the persistent photoconductivity (PPC) effect of two-dimensional material phototransistors and the rapid relaxation of photocurrent in PN junction channel phototransistors. The persistent photoconductivity effect is dominated by a charge transfer and interface trap model. Inherent defects or adsorbates (such as hydroxyl groups, oxygen vacancies, and water-oxygen molecules) exist at the interface between the two-dimensional material and the substrate (e.g., SiO2 / Si) or dielectric layer. Photon energy is absorbed by MoS2, generating electron-hole pairs, which may also directly excite carriers in the interface trap state. Driven by a vertical electric field (derived from the gate voltage or built-in field), photogenerated carriers (especially holes) tunnel or cross the potential barrier and are irreversibly or irreversibly trapped by deep-level traps at the interface; for example, holes are trapped in oxygen-related traps in SiO2. The trapped charges establish a persistent, directional built-in electric field at the interface. This electric field effectively n-type doping of the two-dimensional material layer through capacitive coupling (if holes are trapped, it is equivalent to injecting electrons into the channel), resulting in its threshold voltage V. thA negative drift occurs. Even after illumination stops, these trapped charges, due to their extremely deep energy levels and high potential barriers, have a very low probability of thermally escaping at room temperature, thus maintaining the high conductivity state of the channel. Recovery of the conductivity state depends on the recombination of trapped charges through extremely slow processes such as thermally assisted tunneling, or quenching through external stimuli such as applying a reverse gate voltage or increasing temperature. For two-dimensional PN junctions with van der Waals contacts, such as molybdenum disulfide / tungsten selenide, type-II band alignment is achieved, providing a strong driving force for the effective spatial separation of photogenerated electron-hole pairs. Under illumination, excitons or free carriers can quickly cross the potential barrier at the PN junction interface, achieving spatial separation of charges. Furthermore, the spatially separated electrons and holes are located in the two layers of the heterojunction, greatly suppressing direct radiative recombination and non-radiative Auger recombination, and the built-in electric field accelerates the separation of photogenerated carrier pairs. Therefore, when illumination stops, there are no large numbers of long-lived carrier pairs waiting to recombine, and the photocurrent relaxes rapidly.

[0038] like Figure 2 As shown, this invention utilizes the photoelectric and storage properties of molybdenum disulfide and tungsten selenide to design a primary motion detector circuit composed of two phototransistors. The first phototransistor M1 is forward-biased from source to drain and has a tungsten selenide-molybdenum disulfide PN junction channel, exhibiting slow attenuation of photocurrent during turn-off. The second phototransistor M2 is reverse-biased from source to drain and has a tungsten selenide-molybdenum disulfide PN junction channel transistor, exhibiting rapid attenuation of photocurrent during turn-off. M1 and M2 can generate a current response to light pulses, and the attenuation curve of the photocurrent can be obtained from... The fitting shows that the attenuation process of the photocurrent has two time constants; where I represents the magnitude of the photocurrent (in amperes) and t represents the time (in seconds). For fast decay time constant, The decay time constant is slow. The fast decay time constant of the first phototransistor M1 is greater than that of the second phototransistor M2. To ensure proper power supply to the circuit, when the point light source moves from M1 to M2, a spike pulse current can be obtained at the output terminal due to the relatively slow relaxation of the photocurrent in M1; conversely, no pulse current can be obtained if the light source moves from M1 to M2. Figure 2 In the diagram, VDD represents the positive power supply, VG1 represents the gate voltage of M1, VG2 represents the gate voltage of M2, and Iout represents the source-drain current of M2.

[0039] In photoelectric testing of phototransistors, the depletion region width of a forward-biased tungsten selenide-molybdenum disulfide vertical PN junction phototransistor is smaller, the PN junction depletion layer barrier is lowered, and the photocurrent relaxation process is mainly dominated by persistent photoconductivity, resulting in a relatively large photocurrent decay time constant. Conversely, in a reverse-biased tungsten selenide-molybdenum disulfide vertical PN junction phototransistor, the depletion layer barrier is increased, and the photocurrent relaxation process is dominated by the built-in electric field. The built-in electric field drives carrier drift, accelerating photogenerated electron-hole recombination in the depletion region, resulting in a relatively small photocurrent decay time constant. Based on the differences in phototransistor power supply conditions, this invention connects a first phototransistor and a second phototransistor in series to realize the function of a primary motion detector. The process flow and connection method of the phototransistors are as follows: Figure 3 and Figure 4 .

[0040] The fabrication method of the primary motion detector circuit includes the following steps:

[0041] Step 1: Pattern the substrate, deposit the back gate electrode, and remove the adhesive.

[0042] Step 2: Deposit dielectric material on the back gate electrode.

[0043] Step 3: Pattern the dielectric material, deposit the drain electrode, and remove the adhesive.

[0044] Step 4: Transfer tungsten selenide material to the dielectric surface; transfer molybdenum disulfide material to tungsten selenide, and form the transistor channel through photolithography, development, etching and other processes.

[0045] Step 5: Perform spin coating, photolithography and development, define the source stage, and deposit and strip the source metal electrodes of the device to connect the source and drain metal electrodes of two different phototransistors.

[0046] Step 6: Test the electrical performance of the obtained device.

[0047] In the testing, this method utilized a Keysight B1500 semiconductor analyzer and a high / low temperature vacuum probe station. Light pulses of 520 nm green light were used for illumination. The photocurrent relaxation rates of the two phototransistors showed significant differences under different source-drain biases. Figure 5 The test results of the phototransistor under different biases are shown. Figure 5 (a) in the figure shows the light pulse test curve of the phototransistor with the source and drain biased in the forward direction; Figure 5 (b) Photopulse test curve of the phototransistor with reverse source-drain bias. The photocurrent relaxation curve of the tungsten selenide-molybdenum disulfide PN junction phototransistor of this invention is fitted using a double exponential function:

[0048] .

[0049] When the source and drain are forward biased, the slow relaxation time of the photocurrent is approximately 1.3 s, and when the source and drain are reverse biased, the fast relaxation time of the photocurrent is approximately 20 ms.

[0050] like Figure 6 As shown, the slow relaxation process of photocurrent can be represented by a double exponential function. The fit is very good (red dashed line in the figure), where A represents the reference current in the dark state, B and D represent the fast decay coefficient and the slow decay coefficient, respectively, and C and E represent the fast decay time constant and the slow decay time constant, respectively.

[0051] like Figure 7 As shown, the fast relaxation time of the photocurrent is 20ms.

[0052] In Spice simulation, gate electrical pulses are used to simulate optical pulses, and the timing of the electrical pulses is set to simulate the direction of light spot movement. Figure 8 The simulation circuit diagram and results are shown for the preferred direction. Figure 8 (a) shows the simulation circuit diagram for the preferred direction. The simulation circuit consists of capacitor C1, resistor R2, power supply V1, power supply V2, power supply V3, diode D1, first phototransistor M1, second phototransistor M2, and resistor R1. The preferred direction refers to the direction of the light pulse movement that enables the primary motion detector circuit to generate an effective output signal. Figure 8 (b) in the figure represents the simulation results for the preferred direction. Figure 9 The simulation circuit diagram and results are for the non-preferred direction. Figure 9 (a) is the simulation circuit diagram for the non-preferred direction, which has the same structure as the simulation circuit for the preferred direction, but the timing of the electrical pulses is different. Pluuse1 and Pluuse2 represent different timing sequences. Figure 9 (b) in the diagram represents the simulation results for the non-preferred direction. Where V... pluse1 V represents the pulse voltage of the first phototransistor M1. pluse2 The pulse voltage represents the second phototransistor M2; an RC circuit is used to simulate the relaxation behavior of the phototransistor under forward source-drain bias. A measuring resistor R1 is connected to the ground terminal of the fast relaxation transistor M2, and the current I in R1 is detected. R1 When the timing of the gate pulses follows the preferred direction, R1 detects the output pulse current; otherwise, there is no pulse current.

[0053] In one embodiment, the present invention also provides a biomimetic vision sensor comprising a sensing array consisting of multiple primary motion detector circuits (EMDs) arranged at equal intervals in a plane. Each primary motion detector circuit has two phototransistors positioned along its preferred direction, with the preferred directions of different primary motion detector circuits pointing in different directions. When a small light spot moves along a specific direction, only the primary motion detector circuit whose preferred direction is said specific direction will generate a significant current pulse at its output, while the other primary motion detector circuits will have no pulse output or a weak output. Thus, the biomimetic vision sensor can simultaneously sense and distinguish light motion information from multiple different directions on a plane.

[0054] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0055] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the accompanying drawings may include multiple steps or stages, which are not necessarily completed at the same time, but may be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but may be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0056] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0057] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention, and no reference numerals in the claims should be construed as limiting the scope of the claims.

[0058] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A primary motion detector based on a PN junction channel, characterized in that, include: First phototransistor M1 and second phototransistor M2; The channels of both the first and second phototransistors are based on a vertical PN junction formed by molybdenum disulfide and tungsten selenide, and the first and second phototransistors are connected in series. The first phototransistor is configured to operate under a source-drain forward bias, and the photocurrent decay process during turn-off has a first fast decay time constant and a first slow decay time constant. The second phototransistor is configured to operate under a source-drain reverse bias, and the photocurrent during turn-off has a second fast decay time constant and a second slow decay time constant, wherein the first fast decay time constant is greater than the second fast decay time constant. When the light pulse moves along the direction from the first phototransistor to the second phototransistor, a directional current pulse can be detected at the output of the primary motion detector.

2. The primary motion detector based on a PN junction channel according to claim 1, characterized in that, The source-drain forward bias voltage is greater than 0V, and the source-drain reverse bias voltage is less than 0V.

3. The primary motion detector based on a PN junction channel according to claim 1, characterized in that, The first fast decay time constant and the second fast decay time constant are in the range of 0.2 seconds to 0.8 seconds, and the first slow decay time constant and the second slow decay time constant are greater than 9.5 seconds.

4. The primary motion detector based on a PN junction channel according to claim 1, characterized in that, The molybdenum disulfide and tungsten selenide are stacked vertically by van der Waals forces to form a Type-II band-aligned vertical PN junction channel.

5. A primary motion detector based on a PN junction channel according to claim 1, characterized in that, The first phototransistor and the second phototransistor further include: a substrate, a back gate electrode located on the substrate, a dielectric layer covering the back gate electrode, a source metal electrode electrically connected to the molybdenum disulfide layer, and a drain metal electrode electrically connected to the tungsten selenide layer; the tungsten selenide layer is located on the dielectric layer, and the molybdenum disulfide layer is located on the tungsten selenide layer.

6. The primary motion detector based on a PN junction channel according to claim 1, characterized in that, The drain of the first phototransistor is electrically connected to the source of the second phototransistor, and the output terminal of the primary motion detector is the measurement node between the drain of the second phototransistor and the ground terminal.

7. A primary motion detector based on a PN junction channel according to claim 1, characterized in that, The photocurrent attenuation curves of the first and second phototransistors for the light pulse are given by the formula. Fit, where, For fast decay time constant, The slow decay time constant, This represents the reference current in the dark state. Indicates the fast decay coefficient. Indicates the slow decay coefficient. Indicates time, This represents photocurrent.

8. A method for preparing a primary motion detector as described in any one of claims 1 to 7, characterized in that, include: A back gate electrode and a dielectric layer covering the back gate electrode are deposited on a substrate; Tungsten selenide material is transferred to the surface of the dielectric layer to form a tungsten selenide layer, and molybdenum disulfide material is transferred onto the tungsten selenide layer to form a molybdenum disulfide layer. A vertical PN junction channel is formed by photolithography and development process. The source metal electrode and drain metal electrode of the first phototransistor and the second phototransistor are respectively fabricated on the vertical PN junction channel by photolithography and development process. The drain of the first phototransistor is electrically interconnected with the source of the second phototransistor.

9. The preparation method according to claim 8, characterized in that, The transfer of tungsten selenide material to the surface of the dielectric layer specifically includes: (1) Apply 15% anisole solvent to tungsten selenide on a sapphire substrate and dry at 90°C for two minutes; (2) Soak the homogenized tungsten selenide in a 1 mol / L potassium hydroxide solution at 70°C for 8 minutes; (3) Place the tungsten selenide in deionized water for peeling. After the tungsten selenide peels off the sapphire substrate, attach the dielectric layer of the target substrate to the tungsten selenide. After successful transfer, dry the target substrate at 50 degrees Celsius for 5 minutes.

10. A biomimetic vision sensor, characterized in that, It includes an array of primary motion detectors arranged in a predetermined space as described in any one of claims 1 to 7, each of the primary motion detectors being configured to detect light motion information in a specific direction.

Images