CN-122002921-A - Bionic vision adaptive phototransistor and preparation method thereof

CN122002921ACN 122002921 ACN122002921 ACN 122002921ACN-122002921-A

Abstract

The invention discloses a bionic vision adaptive photoelectric transistor and a preparation method thereof, wherein the photoelectric transistor comprises a substrate, a grid dielectric layer positioned on the substrate, a single-layer metal oxide nanofiber network functional layer positioned on the grid dielectric layer, and a source electrode and a drain electrode positioned on the single-layer metal oxide nanofiber network functional layer, wherein a semiconductor channel layer is formed between the source electrode and the drain electrode, the single-layer metal oxide nanofiber network functional layer has an intrinsic defect state, so that under the condition of zero grid voltage, when the incident optical power density is lower than a first threshold value, the output current of the photoelectric transistor is monotonically enhanced along with time to simulate dark adaptation behavior, and when the incident optical power density is higher than or equal to the first threshold value, the output current of the photoelectric transistor is firstly increased to a peak value and then decayed along with time to simulate bright adaptation behavior. The invention overcomes the defect that the prior art relies on complex heterojunction and needs external bias voltage driving.

Inventors

JIANG SHANSHAN
CHENG SHUO
HE GANG
GUO XIAOHUI
ZHAO YUNONG
WANG KECHENG
ZHANG HAINAN
REN YI
ZHU ZHENGXU

Assignees

安徽大学

Dates

Publication Date: 20260508
Application Date: 20251211

Claims (10)

1. A biomimetic vision-adapted phototransistor, comprising: A substrate; A gate dielectric layer over the substrate; a single-layer metal oxide nanofiber network functional layer over the gate dielectric layer; A source electrode and a drain electrode which are positioned on the single-layer metal oxide nanofiber network functional layer, wherein a semiconductor channel layer is formed between the source electrode and the drain electrode; The single-layer metal oxide nanofiber network functional layer has an intrinsic defect state, so that under the condition of zero grid voltage, the output current of the phototransistor is monotonically enhanced along with time to simulate dark adaptation behavior when the incident optical power density is lower than a first threshold value, and the output current of the phototransistor is firstly increased to a peak value and then attenuated along with time to simulate bright adaptation behavior when the incident optical power density is higher than or equal to the first threshold value.
2. The bionic vision-adapted phototransistor of claim 1, wherein the substrate is a P-type heavily doped silicon substrate and doubles as a gate of the phototransistor, and the gate dielectric layer is aluminum oxide or silicon dioxide.
3. The biomimetic vision-adapted phototransistor as in claim 1, wherein the material of the single-layer metal oxide nanofiber network functional layer is indium gallium zinc oxide, wherein the individual diameter of the nanofibers is 30nm to 50nm.
4. The biomimetic vision-adaptive phototransistor as recited in claim 1, wherein the source and drain electrodes are each comprised of an aluminum layer, the source and drain electrodes each have a thickness of 60nm to 100nm, and the source and drain electrodes each have a width of 1000 The lengths of the source electrode and the drain electrode are 100 The channel length of the semiconductor channel layer is 200 Channel width of 1000 。
5. The biomimetic visual adaptive phototransistor as recited in claim 1, wherein the adaptation index of the phototransistor is continuously varied between 0 and 1 with the incident optical power density, the adaptation index is a ratio of a difference between a photocurrent peak value and a real-time adaptive response current to a difference between the photocurrent peak value and an initial current, and the first threshold is 5.8 。
6. The preparation method of the bionic vision adaptive phototransistor is characterized by comprising the following steps of: S1, providing a substrate with a grid dielectric layer on the surface; s2, preparing a metal oxide semiconductor precursor solution; S3, depositing the metal oxide semiconductor precursor solution on the grid dielectric layer by adopting an electrostatic spinning process to form a non-compact nanofiber network precursor film; S4, baking treatment, ultraviolet irradiation treatment and high-temperature annealing treatment are sequentially carried out on the nanofiber network precursor film, so that a crystallized single-layer metal oxide nanofiber network functional layer is obtained; S5, preparing source electrodes and drain electrodes which are mutually spaced on the single-layer metal oxide nanofiber network functional layer, and finishing the preparation of the phototransistor.
7. The method according to claim 6, wherein in the step S2, the metal oxide semiconductor precursor solution is an indium gallium zinc oxide precursor solution.
8. The method according to claim 6, wherein in the step S3, the spinning voltage of the electrostatic spinning process is 12kV to 15kV, the distance between the needle and the substrate collecting plate is 13cm to 17cm, and the advancing speed of the metal oxide semiconductor precursor solution is 0.2mL per hour to 0.4mL per hour.
9. The method of manufacturing a biomimetic vision-adaptive phototransistor as set forth in claim 6, wherein the baking process is performed at a temperature of 120 ℃ to 150 ℃ and the high-temperature annealing process is performed at a temperature of 450 ℃ to 500 ℃ in step S4.
10. A method of visual information processing employing a biomimetic vision-adapted phototransistor as claimed in any one of claims 1 to 5, the method comprising: Converting gray information of each pixel in an image to be processed into an incident light signal with corresponding intensity, and inputting the incident light signal into the bionic vision adaptive phototransistor; The incident light signals are processed through the bionic vision adaptive phototransistor according to the self-adaptive response characteristic depending on the optical power density, the current response generated by the high-intensity incident light signals is automatically restrained, and meanwhile, the current response generated by the low-intensity incident light signals is enhanced; and outputting the current signal processed by the bionic vision adaptive phototransistor as image information with enhanced contrast.

Description

Bionic vision adaptive phototransistor and preparation method thereof Technical Field The invention relates to the technical field of bionic electronic information, in particular to a bionic vision adaptive phototransistor and a preparation method thereof. Background The artificial vision system is a leading research direction in the field of neuromorphic computation, and the core aim is to imitate the efficient and intelligent environment sensing and information processing capability of the biological vision system. Biological vision realizes dynamic adaptation from dim to bright ambient light, namely 'dark adaptation' and 'bright adaptation' through cooperation and switching of the visual rod cells and the visual cone cells. The biological mechanism is realized through hardware devices, and has important significance for constructing a bionic sensing and calculating integrated system with low power consumption and high energy efficiency. At present, the research of bionic photoelectric devices for realizing the visual adaptation function mainly depends on heterojunction or composite structures. For example, xie et al (adv. Function. Mater. 2021, 31, 2010655) designed a double layer CsPbBr 3/MoS2 phototransistor. The device successfully simulates various nerve behaviors including self-adaptive sensitivity by utilizing a charge capturing and releasing mechanism at a heterogeneous interface, but the scheme needs expensive quantum dot materials, and the photoelectric adaptation behavior of the device usually depends on external grid voltage driving, so that the problem of higher power consumption exists. Cheng et al (adv. Sci. 2025, 2417461) developed self-powered photo-memristors based on ZnO/WOx heterojunctions and further integrated with threshold switch memory, built adaptive neurons. The device can generate transient response to constant light stimulus, and simulate adaptation behavior based on pulse. However, such implementations still rely on interface effects of the heterojunction in nature and require integration with peripheral memory circuits to perform complex functions, with significant system complexity and integration difficulty. Kuang et al (adv. Funct. Mater. 2023, 33, 2209502) fabricated phototransistors with both synaptic and adaptive functions by controlling the defect state of the polymer-perovskite heterojunction interface. While this work achieves rapid visual adaptation in a single device, the environmental stability of the perovskite material is often poor, and the core function of the device still stems from the carefully designed heterogeneous interface, rather than the inherent properties of a single homogeneous material. In summary, the prior art generally adopts a strategy of "space-shifting function", i.e. implementing adaptive behavior through a multi-layer heterostructure or coupling with external circuitry. This results in complex device structure, high material system compatibility requirements, cumbersome fabrication process, expensive cost, and interface stability challenges. More importantly, most of these devices cannot realize dynamic adaptation behavior of light intensity dependence without continuous external electrical bias (pure light induction), and it is difficult to seamlessly integrate visual perception, adaptive information processing and neuromorphic calculation functions on the same hardware platform. Therefore, developing a convenient and efficient strategy to manufacture a multifunctional device with light-induced adaptation behavior and brain-like calculation function is regarded as a key step for realizing the construction of the brain-like artificial vision system with high hardware efficiency and excellent performance. Disclosure of Invention In order to solve the technical problems in the background technology, the invention provides a bionic vision adaptive phototransistor and a preparation method thereof. The invention provides a bionic vision adaptive phototransistor, which comprises: A substrate; A gate dielectric layer over the substrate; a single-layer metal oxide nanofiber network functional layer over the gate dielectric layer; A source electrode and a drain electrode which are positioned on the single-layer metal oxide nanofiber network functional layer, wherein a semiconductor channel layer is formed between the source electrode and the drain electrode; The single-layer metal oxide nanofiber network functional layer has an intrinsic defect state, so that under the condition of zero grid voltage, the output current of the phototransistor is monotonically enhanced along with time to simulate dark adaptation behavior when the incident optical power density is lower than a first threshold value, and the output current of the phototransistor is firstly increased to a peak value and then attenuated along with time to simulate bright adaptation behavior when the incident optical power density is higher than or equal to the first threshold value. Pre