01

2024

-

08

Two-dimensional MoS/black phosphorus heterojunction medium-wave infrared detector, integrating "sensing, storage and calculation"

Author:


With the rapid development of artificial intelligence and Internet of Things technology, mid-wave infrared (MWIR) target detection and recognition technology is becoming more and more important in the fields of advanced assisted driving, event monitoring and complex environment reconnaissance. Traditional mid-wave infrared photodetectors, such as mercury cadmium telluride (HgCdTe) and indium antimony (InSb) based materials, often require refrigeration to control temperature, which not only increases costs, but also limits the portability and energy efficiency of the device. Therefore, the development of high-performance, low-cost and non-refrigeration medium-wave infrared detectors has become a hot research topic.

Two-dimensional materials provide new possibilities for the development of photodetectors due to their unique physical and chemical properties. These materials have atomic thickness and rich material systems, which make them show great potential in photoelectric detection, storage and computing. In particular, the layered structure of two-dimensional materials and van der Waals (vdWs) contacts support high-density heterogeneous integration, which facilitates the realization of multifunctional integration. In addition, two-dimensional materials exhibit high photon sensitivity at room temperature, making them more responsive in mid-wave infrared photodetectors. However, existing mid-wave infrared detectors still face challenges in terms of power consumption, latency, and functional integration. To solve these problems, researchers are exploring combining sensors with memory computing structures to achieve "all-in-one" hardware. This device can not only perform photoelectric detection, but also store and process information, thereby reducing the power consumption of analog signal conversion and transmission.

Recently, a joint research team from Fudan University and the Shanghai Institute of Technical Physics of the Chinese Academy of Sciences has successfully developed a non-volatile molybdenum disulfide (MoS)/black phosphorus (BP) heterojunction medium-wave infrared photovoltaic detector using the heterogeneous integration characteristics of two-dimensional materials, which integrates near-infrared to mid-infrared photoelectric detection, storage and calculation (PMC) functions. Relevant research results are published in "Non-volatile 2D MoS/black phosphorus heterojunction photodiodes in the near- to mid-infrared region"Nature CommunicationsJournal.

1 Application of mid-wave infrared PMC device based on convolutional neural network (CNN) in night driving assistance

In this study, the researchers demonstrated a non-volatile MoS₂/black phosphorus heterojunction medium-wave infrared photovoltaic detector with a semi-floating gate structure design. The detector can not only carry out efficient photoelectric detection, but also integrate storage and calculation functions, and can realize high-response mid-wave infrared detection at room temperature. The device is a layered heterojunction that includes four layers of two-dimensional materials: black phosphorus, MoS₂, h-BN (tunneling layer), and graphene (Figure 2a).

Fig.2 Internal computing device of non-volatile mid-wave infrared sensor

The salient feature of this detector is the ability to store stable responsiveness, and this responsiveness is linearly related to the stored conduction state. Specifically, the stable response of the detector can be programmed with power consumption as low as 1.8 fJ, while the black body peak response can reach 1.68 A/W, which is suitable for the mid-wave infrared band. The relevant results are shown in Figures 3 and 4.

Fig.3 Infrared laser response of PMC device in different conductive states

Fig.4 Blackbody detection and mechanism analysis of PMC devices in different conductive states

Finally, the researchers verified the effectiveness of the detector in the actual scene through simulation experiments. In the CNN simulation based on the FLIR data set, the hardware responsiveness weight of the detector can reach the average accuracy index of 89% of the software feature extraction network weight. This shows that the mid-wave infrared photovoltaic detector has important application prospects in advanced infrared target detection and recognition systems.

5 Target detection and recognition based on mid-wave infrared scene

In summary, this research successfully developed a non-volatile MoS₂/black phosphorus heterojunction medium-wave infrared photovoltaic detector that integrates photoelectric detection, storage and calculation functions. The device not only exhibits highly responsive mid-wave infrared detection capability at room temperature, but also enables stable responsive storage through low-power programming. Through simulation experiments, the researchers verified the effectiveness of the device in actual scenes and demonstrated its application potential in advanced infrared target detection and recognition systems. The researchers explored the working principle of the device and found that by changing the conductive state of the device, its photoelectric responsiveness can be adjusted, thereby realizing the simulation of the neural network weight. This discovery provides a theoretical basis for the future application of the device in intelligent visual assisted driving systems. In addition, the researchers also discussed the potential of the device in large-scale production and integration, pointing out that by optimizing the material production, selective etching process and advanced transfer technology, large-scale fabrication of PMC arrays can be realized. In short, this research not only makes a breakthrough in technology, but also provides new ideas and solutions for the development of future intelligent vision systems. With the continuous progress and optimization of technology, this device, which integrates photoelectric detection, storage and computing functions, is expected to be applied in more fields and promote the development of intelligent vision technology.

Chongqing Chongfan Technology Co., Ltd. Optical platform film thickness meter OCT optical coherence tomography LabVIEW control