Position-sensitive Te nanowire photoelectric detector and preparation method and application thereof

Abstract

The invention discloses a position-sensitive Te nanowire photoelectric detector and a preparation method and application thereof, and belongs to the technical field of photoelectric detection and neuromorphic calculation. The invention discloses a suspended Te nanowire to minimize parasitic interference of a substrate so as to improve the collection efficiency of photon-generated carriers, and also can realize rapid thermal isolation to accurately control the temperature gradient so as to realize the photoelectric response of a wide spectrum at 1310-1650nm under a zero offset value. The position-adjustable positive and negative photoconductive characteristics generated by the Seebeck effect are utilized to show the calculation application of the device in the sensing under the zero offset value, 85% of accuracy is realized in the image identification application, and the problems that the calculation power consumption in the traditional sensing is large, the device separation degree is poor, the application in the communication wave band is difficult to realize and the like are solved. The device provides a thought for developing a low-power-consumption visual sensing system under wide spectrum response.

Description

Technical Field

[0001] This invention belongs to the field of photoelectric detection and neuromorphic computing technology, specifically relating to a position-sensitive Te nanowire photodetector, its fabrication method, and its application. Background Technology

[0002] A significant portion of the raw data generated in sensing nodes is unstructured and redundant. In typical designs, these sensing systems are physically separated from computing units due to varying functional requirements and manufacturing technologies. Sensing primarily occurs in the noisy analog domain, while computation is typically performed digitally using traditional von Neumann computing architectures. Consequently, sensor terminals acquire large amounts of raw data locally and must then transmit it to local computing units or cloud-based systems. This presents significant challenges in terms of energy consumption, response time, data storage, communication bandwidth, and security. This data diffusion imposes unique information processing requirements on sensor-rich platforms such as smart cars, autonomous driving, microrobots, and wearable electronics, as well as applications with stringent latency requirements, such as real-time video analytics and cooperative autonomous driving.

[0003] To overcome this problem, computing architecture needs to shift from a compute-centric approach to a data-centric approach, with a particular focus on performance and energy efficiency. One approach is in-sensor computing, where tasks including data generation, collection, and computation are performed internally within the sensor device in some cases. This approach aims to provide efficient sensor data processing and reduce or eliminate data transmission and conversion at the sensor / processor interface. While numerous in-sensor computing systems have been developed, they still require external bias voltages and have limited spectral response ranges (limited to ultraviolet to visible light), inevitably leading to high device power consumption and difficulties in operating in specific spectral scenarios (such as communication bands). Summary of the Invention

[0004] The purpose of this invention is to provide a position-sensitive Te nanowire photodetector, its fabrication method, and its application, in order to solve the technical problem that existing photodetectors need to operate under bias, resulting in high device power consumption.

[0005] To achieve the above objectives, the present invention employs the following technical solution: This invention discloses a position-sensitive Te nanowire photodetector, comprising an insulating substrate, a metal electrode, and a Te nanowire; the metal electrode is disposed on the surface of the insulating substrate; the Te nanowire is suspended above the metal electrode and forms an ohmic contact with the metal electrode.

[0006] Furthermore, the insulating substrate is a silicon substrate, an aluminum oxide substrate, a hafnium oxide substrate, a yttrium oxide substrate, or an aluminum nitride substrate with a silicon dioxide layer.

[0007] Furthermore, the metal electrode is a composite electrode composed of a chromium electrode and a gold electrode.

[0008] Furthermore, the thickness of the metal electrode is 35~55 nm.

[0009] Furthermore, the Te nanowires have a length of 300~600 μm, a cross-sectional shape of hexagonal prism, and a diameter of 10~20 μm.

[0010] This invention also discloses a method for fabricating the aforementioned position-sensitive Te nanowire photodetector, comprising the following steps: Electron beam lithography was used to prepare electrode patterns on the surface of an insulating substrate, and then metal electrodes were prepared according to the electrode patterns using electron beam evaporation deposition. A position-sensitive Te nanowire photodetector was obtained by transferring Te nanowires to the surface of a metal electrode using a dry transfer method and suspending them on the surface of the metal electrode.

[0011] Furthermore, the electron beam exposure dose is 3.2~8 mJ / cm. 2 The beam current is 2~8 nA.

[0012] Furthermore, the suspension distance between the Te nanowire and the metal electrode is 35~55 nm.

[0013] Furthermore, the Te nanowire is suspended 40 nm above the metal electrode.

[0014] This invention also discloses the application of the aforementioned position-sensitive Te nanowire photodetector in sensor-based computation.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention discloses a position-sensitive Te nanowire photodetector. By placing a two-dimensional semiconductor and Te nanowires on the surface of an insulating substrate, with the Te nanowires suspended above the two-dimensional semiconductor, the thermal conductivity of the insulating substrate itself leads to the attenuation of the photothermal-electric effect of the Te nanowires, resulting in a decrease in the photoresponse of the Te nanowires. This suspended design can minimize substrate parasitic interference to improve the collection efficiency of photogenerated carriers, and achieve rapid thermal isolation to precisely control the temperature gradient, ensuring efficient injection and extraction of photogenerated carriers. Furthermore, the position-sensitive Te nanowire photodetector of this invention has higher separation and does not rely on additional external conductivity control devices. Through the built-in artificial neural network (ANN) constructed by this system, and using position-adjustable responsivity as the weight of the ANN to achieve MAC operation for efficient visual information processing, it realizes in-sensor computation in the communication band (1550 nm) without external bias, solving the technical problem of high power consumption caused by the need for existing photodetectors to operate under bias.

[0016] Furthermore, due to the thermoelectric effect of the Te nanowire itself, photons in this band are absorbed by the Te nanowire and hot carriers are generated, resulting in a temperature gradient at both ends of the Te nanowire, thereby achieving current output (Seebeck effect), thus realizing the spectral response in this band. This invention utilizes the prepared thermoelectric material Te NWs to achieve a wide spectral response in the communication band. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the position-sensitive Te nanowire photodetector of the present invention; Wherein: a-Schematic diagram of the overall structure; b-Scanning electron microscope image; Figure 2 Aberration-corrected electron microscopy and selected area diffraction pattern of Te nanowires; Where: a-spherical aberration electron microscopy; b-selection of diffraction pattern; Figure 3 The photocurrent-power dependence curves of the position-sensitive Te nanowire photodetector of this invention in the 1310-1650 nm band, as well as the corresponding responsivity and detectivity; The following parameters are included: a - Power-photocurrent power-law fitting curves for 1310 nm lasers; b - Responsivity for different powers under 1310 nm lasers; c - Detectivity for different powers under 1310 nm lasers; d - Power-photocurrent power-law fitting curves for 1550 nm lasers; e - Responsivity for different powers under 1310 nm lasers; f - Detectivity for different powers under 1310 nm lasers; g - Power-photocurrent power-law fitting curves for 1650 nm lasers; h - Responsivity for different powers under 1650 nm lasers; i - Detectivity for different powers under 1650 nm lasers. Figure 4 The linear variation curves of photocurrent with power and the corresponding position-dependent responsivity curves of the position-sensitive Te nanowire photodetector of the present invention at different positions of the nanowire in the wavelength range of 1310-1650 nm are shown. Wherein: a - Power-law fitting of 1310 nm laser power and photocurrent at different positions; b - Power-law fitting of 1550 nm laser power and photocurrent at different positions; c - Power-law fitting of 1650 nm laser power and photocurrent at different positions; d - Linear fitting curves of 1310 nm laser power and photocurrent at different positions; e - Linear fitting curves of 1550 nm laser power and photocurrent at different positions; f - Linear fitting curves of 1650 nm laser power and photocurrent at different positions. Figure 5 This is a schematic diagram illustrating the in-sensor computation principle of the position-sensitive Te nanowire photodetector used for image recognition in this invention. Figure 6 Image recognition accuracy curves of the letters “K”, “H”, and “X” achieved using position-sensitive Te nanowire photodetectors. Detailed Implementation

[0018] To enable those skilled in the art to understand the features and effects of the present invention, the terms and expressions used in the specification and claims are explained and defined in general below. Unless otherwise specified, all technical and scientific terms used herein have the ordinary meaning understood by those skilled in the art regarding the present invention, and in case of conflict, the definitions in this specification shall prevail.

[0019] The theories or mechanisms described and disclosed herein, whether right or wrong, should not in any way limit the scope of the invention, that is, the contents of the invention can be implemented without being limited by any particular theory or mechanism.

[0020] In this document, all features defined by numerical ranges or percentage ranges, such as numerical values, quantities, contents, and concentrations, are for the sake of brevity and convenience only. Accordingly, descriptions of numerical ranges or percentage ranges should be considered as covering and specifically disclosing all possible sub-ranges and individual numerical values ​​(including integers and fractions) within those ranges.

[0021] In this article, unless otherwise specified, “contains,” “includes,” “containing,” “has,” or similar terms cover the meanings of “composed of” and “mainly composed of,” for example, “A contains a” covers the meanings of “A contains a and others” and “A contains only a.”

[0022] For the sake of brevity, not all possible combinations of the technical features in each implementation scheme or embodiment are described herein. Therefore, as long as there is no contradiction in the combination of these technical features, the technical features in each implementation scheme or embodiment can be combined arbitrarily, and all possible combinations should be considered within the scope of this specification.

[0023] This invention provides a position-sensitive Te nanowire photodetector for in-sensor computing. By suspending the Te nanowire, the efficiency of photogenerated carrier collection is improved, thereby enhancing photodetector performance. At the same time, by utilizing the position response characteristics dominated by the Seebeck effect, a linear positive and negative light response is achieved inside the channel, demonstrating the application of the device in in-sensor computing.

[0024] The position-sensitive Te nanowire photodetector provided by this invention comprises, from bottom to top: an insulating substrate 1, a metal electrode 2, and a Te nanowire 3; wherein, the substrate 1 is a silicon substrate with a silicon dioxide layer. The metal electrode 2 is a chromium (Cr) + gold (Au) composite electrode. By designing the growth thickness and surface treatment process, a high-quality ohmic contact is formed between it and the Te nanowire to ensure the efficient injection and extraction of photogenerated carriers. The Cr thickness is 5 nanometers and the Au thickness is 30 nanometers. The Te nanowires 3 are suspended on the surface of the aforementioned composite electrode by dry transfer. This suspended design can minimize substrate parasitic interference to improve the photogenerated carrier collection efficiency, achieve rapid thermal isolation to precisely control the temperature gradient, realize high-performance photoelectric detection performance, and enable a wide range of adjustable weights in sensor-in-computation applications.

[0025] This invention also discloses a method for fabricating the aforementioned position-sensitive Te nanowire photodetector, comprising the following steps: 1) Metal electrodes 2 are fabricated on insulating substrate 1 using electron beam lithography combined with thermal evaporation and stripping processes; 2) The Te nanowires 3 were transferred to the surface of the metal electrode 2 using a dry transfer technique.

[0026] The suspended nanowire structure used in this invention can minimize substrate parasitic interference to improve the photogenerated carrier collection efficiency, and can also achieve rapid thermal isolation to precisely control the temperature gradient, thereby achieving high-performance photoelectric detection.

[0027] Traditional photodetectors used for sensing computation often require large bias voltages to improve carrier separation efficiency or suppress dark current noise, resulting in extremely high power consumption. In contrast, this invention eliminates the need for bias voltages in current sensors used for in-sensing computation, achieving operation with extremely low power consumption.

[0028] Currently, most materials used in in-sensor computing systems are two-dimensional (2D) materials, resulting in non-uniform devices. Furthermore, the band gap of these materials limits their spectral response range, typically extending from the visible to the ultraviolet range, making it difficult to cover the more critical communication bands. This invention utilizes the prepared thermoelectric material TeNWs to achieve a wide spectral response in the communication band.

[0029] Unlike other photodetectors used for in-sensor computing, which typically use bias voltage, wavelength, etc. to control positive and negative conductance, limiting device separation and application in fixed-band environments, this invention uses a position-tunable positive / negative photoconductivity Te NWs in-sensor computing system. This system has higher device separation and does not rely on additional external conductance control devices. Through the built-in artificial neural network (ANN) constructed by this system, and using position-tunable responsivity as the weight of the ANN, MAC operation is performed to execute efficient visual information processing, realizing in-sensor computing in the communication band (1550 nm) without external bias.

[0030] The present invention will be further illustrated 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.

[0031] The following examples use instruments and equipment conventional in the art. Experimental methods in the following examples, unless otherwise specified, are generally performed under conventional conditions or as recommended by the manufacturer. All raw materials used in the following examples are conventional commercially available products with specifications conventional in the art. In this specification and the following examples, unless otherwise specified, "%" refers to weight percentage, "parts" refers to parts by weight, and "ratio" refers to weight proportion.

[0032] The position-sensitive Te nanowire photodetector described in this invention uses readily available and inexpensive materials, and its fabrication is controllable and easy to operate. Therefore, the fabrication cost of this invention is low.

[0033] This invention develops a photodetector based on the Seebeck effect of Te nanowires for in-sensor computation. By designing the Te nanowires in a suspended manner, high responsivity and high sensitivity are achieved across a wide spectrum. Utilizing the position-tunable positive and negative photoconductivity at zero bias, in-sensor computation with low power consumption is realized in the communication band (1550 nm).

[0034] Example 1 A method for fabricating a position-sensitive Te nanowire photodetector, the specific steps of which are as follows: Step 1: Select a suitable substrate and clean it. Select silicon dioxide as the insulating substrate 1; Step 2: The metal electrode 2 (Cr (5nm) + Au (30nm)) was fabricated using electron beam lithography with an exposure dose of 3.2 mJ / cm². 2 The exposure beam current was 8 nA, followed by development and fixing. The development time was 2 min and the fixing time was 3 min. The channel length of the metal electrode 2 was 60 μm, and the width of the metal electrode 2 in contact with the Te nanowire 3 was 5 μm. Step 3: Subsequently, the grown Te nanowires 3 are precisely transferred above the metal electrode 2 using a dry transfer method. At this point, the Te nanowires 3 are suspended on the surface of the silicon dioxide substrate, which can effectively reduce substrate parasitic interference to improve the photogenerated carrier collection efficiency, and also achieve rapid thermal isolation to precisely control the temperature gradient, thereby improving the photoelectric detection performance of the Te nanowires. The suspension distance between the Te nanowires 3 and the metal electrode 2 is 40 nm; specific schematic diagrams and electron microscope images are shown below. Figure 1 a, Figure 1 As shown in b.

[0035] Example 2 A method for fabricating a position-sensitive Te nanowire photodetector, the specific steps of which are as follows: Step 1: Select a suitable substrate and clean it. Silicon nitride is selected as the insulating substrate 1. Step 2: The metal electrode 2 (Cr (5 nm) + Au (50 nm)) was fabricated using electron beam lithography at an exposure dose of 3.2 mJ / cm². 2 The exposure beam current was 8 nA, followed by development and fixing. The development time was 2 min and the fixing time was 3 min. The channel length of the metal electrode 2 was 60 μm, and the width of the metal electrode 2 in contact with the Te nanowire 3 was 5 μm. Step 3: Subsequently, the grown Te nanowires 3 are precisely transferred above the metal electrode 2 using a dry transfer method. At this time, the Te nanowires 3 are suspended on the surface of the silicon nitride substrate, which can effectively reduce substrate parasitic interference to improve the photogenerated carrier collection efficiency, and can also achieve rapid thermal isolation to accurately control the temperature gradient, thereby improving the photoelectric detection performance of the Te nanowires. The suspension distance between the Te nanowires 3 and the metal electrode 2 is 40 nm.

[0036] Example 3 The specific steps in this embodiment are similar to those in Embodiment 1, except that the electron beam exposure current is 2 nA, and the width of the metal electrode 2 in this embodiment is 2 and the channel length is 0.8 μm.

[0037] Example 4 The specific steps in this embodiment are similar to those in Embodiment 1, but the metal electrode pattern is prepared by ultraviolet lithography, with an exposure time of 8 s, a development time of 20 s, and a fixing time of 1 min.

[0038] Example 5 The specific steps in this embodiment are similar to those in Embodiment 1, but the Te nanowires used in this embodiment have a length of 600 μm.

[0039] Thanks to the photothermal and electrical properties of Te nanowires 3, as well as their high crystallinity and low defects, and combined with the effective separation of Te nanowires 3 from the substrate, which reduces the effective thermal conduction between Te nanowires 3 and the substrate, the position-sensitive Te nanowire photodetector realized by this invention has excellent infrared detection performance and position response characteristics.

[0040] The Te nanowires 3 in this invention possess excellent photothermal and electrical properties, exhibiting superior photoelectric detection performance in the infrared band of 1310-1650 nm under zero bias, such as... Figure 3 As shown (lasers in this wavelength range are all irradiated within the channel of the Te nanowire), thanks to the Seebeck effect, when the laser irradiates the two ends of the Te nanowire 3, the resulting currents have different directions, such as... Figure 3 As shown in a, d, and g, the Te nanowires based on this invention achieved responsivity of 2.76 A / W, 19.09 A / W, and 3.71 A / W at 1310 nm, 1550 nm, and 1650 nm, respectively, and detectivity of 5.98 × 10⁻⁶. 9 Jones, 4.13×10 10 Jones, 8.02×10 9 Jones, as shown in Figures b, e, h, c, f, and i.

[0041] When lasers of different wavelengths (including 1310, 1550, and 1650 nm) move along the axial direction of the Te nanowire within the channel of a position-sensitive Te nanowire photodetector—that is, from one end of the Te nanowire in contact with an electrode to the other end in contact with the Te nanowire—the photocurrent exhibits a quasi-linear dependence on the length of the laser's movement, and displays positive and negative symmetry (the x-coordinate is calibrated to 0 when the Te nanowire photocurrent is at its minimum). Figure 4 As shown.

[0042] Based on the above characteristics, position-sensitive Te nanowire photodetectors can be used to realize letter image recognition applications. Their positive and negative linear response characteristics allow for adjustment of the convolution kernel during letter image recognition, and also allow for weight updates during the training of convolutional neural networks.

[0043] The algorithmic principles for implementing in-sensor computing applications are explained, such as... Figure 5 As shown, a position-sensitive Te nanowire photodetector based on position-tunable positive and negative photoconductivity is used to classify images of letters ('K', 'H', 'X'), where each letter is a 3×3 pixel image with a Gaussian noise level (σ=0.3). Figure 5 This schematically illustrates the implementation of a neural network classifier within an image sensing array. The input image is expressed in terms of light intensity (...). P j The encoding and output photocurrent follow the formula I ph =∑ R j P j ,in R j This represents the responsiveness of each pixel. R j The amplitude and polarity of the laser can be adjusted by the position of the laser on the nanowire, effectively acting as weights in the convolution kernel matrix. The specific process is as follows: the weight of each pixel ( R j ) by location P (Position) Linear setting; this algorithm can be updated using the error gradient descent algorithm. Output I ds Image input P j The simulated multiply-add operation results of the weight matrix obtained from (optical power) and Kirchhoff's laws. After the activation function, the output error is calculated, and then the weights are updated using the gradient descent algorithm. R j ,Should I ds This is then applied to the array of devices to perform new training. When the output error is sufficiently low (or the accuracy is sufficient), training stops and inference is performed using a hardware neural network classifier on the array of computing devices within the sensor. Figure 6 The study demonstrated that the in-sensor computing system achieved a classification accuracy of 85% after 14 training iterations.

[0044] In summary, this invention discloses a position-dependent optical response characteristic based on the thermoelectric material Te for in-sensor computational applications. Most importantly, this characteristic enables efficient and accurate image classification. This research provides insights for developing low-power visual sensing systems with a broad spectral response.

[0045] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A position-sensitive Te nanowire photodetector, characterized in that, It includes an insulating substrate (1), a metal electrode (2) and a Te nanowire (3); the metal electrode (2) is disposed on the surface of the insulating substrate (1); the Te nanowire (3) is suspended above the metal electrode (2) and forms an ohmic contact with the metal electrode (2).

2. The position-sensitive Te nanowire photodetector according to claim 1, characterized in that, The insulating substrate (1) is a silicon substrate, aluminum oxide substrate, hafnium oxide substrate, yttrium oxide substrate or aluminum nitride substrate with a silicon dioxide layer.

3. The position-sensitive Te nanowire photodetector according to claim 1, characterized in that, The metal electrode (2) is a composite electrode composed of a chromium electrode and a gold electrode.

4. A position-sensitive Te nanowire photodetector according to claim 1, characterized in that, The thickness of the metal electrode (2) is 35~55 nm.

5. A position-sensitive Te nanowire photodetector according to claim 1, characterized in that, The Te nanowire (3) has a length of 300~600 μm, a cross-sectional shape of hexagonal prism, and a diameter of 10~20 μm.

6. A method for fabricating a position-sensitive Te nanowire photodetector according to any one of claims 1 to 5, characterized in that, Includes the following steps: Electrode patterns were prepared on the surface of an insulating substrate (1) using electron beam lithography, and then metal electrodes (2) were prepared by electron beam evaporation deposition according to the electrode patterns. Te nanowires (3) were transferred to the surface of a metal electrode (2) using a dry transfer method and suspended on the surface of the metal electrode (2) to obtain a position-sensitive Te nanowire photodetector.

7. A method for fabricating a position-sensitive Te nanowire photodetector according to claim 6, characterized in that, The electron beam exposure dose is 3.2~8 mJ / cm. 2 The beam current is 2~8 nA.

8. A method for fabricating a position-sensitive Te nanowire photodetector according to claim 6, characterized in that, The Te nanowire (3) is suspended at a distance of 35~55 nm above the metal electrode (2).

9. A method for fabricating a position-sensitive Te nanowire photodetector according to claim 6, characterized in that, The Te nanowire (3) is suspended 40 nm above the metal electrode (2).

10. The application of a position-sensitive Te nanowire photodetector as described in any one of claims 1 to 5 in sensing-based computation.

Images