Limei Qi

In traditional optics, diffraction ringing induced by a finite system aperture is often treated as an artifact requiring suppression. In this work, we experimentally demonstrate that diffraction ringing can directly enhance the edges of terahertz images. The experiment uses a terahertz time-domain spectroscopy system coupled to a 4f imaging system to image sharp-edged objects, including both square and hexagram, at 0.6, 0.8, and 1.0 THz. Under a limited aperture, clear edge enhancement is observed. This enhancement shows a negative correlation with frequency, meaning that lower frequencies produce stronger oscillations. This result is supported by good agreement among theory, simulation, and experiment. The research demonstrates that edge enhancement can be achieved directly during the imaging process, requiring neither additional optics nor complex digital processing. This approach improves processing efficiency and system integration, thereby holding promise for applications in object identification and classification.

A terahertz (THz) metamaterial broadband perfect absorber featuring a simple sandwich structure with a top layer composed of a triangular-patterned graphene film is presented. The graphene pattern is designed to exhibit a pronounced surface plasmon resonance (SPR) effect, which locally enhances the internal electric field’s intensity, leading to broadband absorption of 2.8 THz above 90% and a peak absorption rate of 99.99% at 6.05 THz. The broadband tunability of the absorber was further investigated by modulating the Fermi level of the graphene, demonstrating an adjustment in the absorption rate from 6.18% to 99.99% via an external voltage. This study shows that the absorber demonstrates excellent angular tolerance by maintaining an absorption rate above 90% across incident angles ranging from 0° to 50°. The absorber’s broadband perfect absorption properties were examined using relative impedance theory. Additionally, to reveal the fundamental physics behind this absorption, detailed analyses of the electric field distributions were carried out. Consequently, the origin of the absorption peaks is elucidated. This absorber enables noise suppression for optoelectronic integration and THz communications.

With the popularization of social networks, fake news is also widely and rapidly spreading, which poses a great threat to the Internet. Therefore, how to detect fake news automatically and efficiently has become an urgent problem to be solved. However, the existing approaches mostly focus on the explicit features (images and text) and deep fusions, without considering potential features such as text emotion and image category. To find a solution to this issue, we propose a Potential Features Fusion Network (PFFN), which models the explicit and potential features at the same time. To exploit the potential image features, we introduce a mixture of experts structure to process the news image separately, which can best use the relationships between the news image category and fake news detection. Besides, we also extract emotion features as potential text features and fuse them with explicit text features. Finally, we establish an attention-based feature fusion network to fuse the potential features with the explicit features, which can obtain a multimodal fusion feature of a piece of news and thus further improve the performance. We make experiments on four public datasets (Weibo16, Weibo19, Twitter, and PolitiFact); the results compared with the baseline approaches demonstrate that our PFFN has a better performance. Our code is available at https://github.com/Wang-bupt/PFFN

The existing THz technology uses THz metasurface to design THz waves. In order to produce SPP effect and local surface plasma plasmon (LSP) effect on THz band, there are obvious defects such as large area, high loss, slow tuning speed and insensitivity to laser irradiation. To address these limitations, an external field is employed to modulate the designed optical metasurface, exhibiting a strong response to the infrared laser spectrum. And then, the altered carrier concentration and dielectric constant of the material directly influence the performance of the THz devices. This approach results in a substantial increase in the modulation amplitude of responsivity ( R_A ) and noise equivalent power ( NEP ). Based on this principle, we propose a Petal-shaped, ultra-fast, low-loss, all-dielectric optical metasurface THz sensor. The THz irradiated device generate the carriers, travelling to the external circuit, forming a THz current. The laser acts directly on the metasurface to actively modulate the device._Notably, the designed sub-wavelength configuration creates a cross-resonator, which induces a local field enhancement effect, leading to a significant increase in optical absorption efficiency and a large modulation amplitude. Experimental results demonstrate that the maximum modulation amplitude of R_A can reach 50% at a wavelength of 633 nm, while the maximum modulation amplitude of NEP can achieve 74% at a wavelength of 532 nm. Moreover, compared to traditional THz-modulated sensors, the laser-modulated THz sensor exhibits ultra-high speed characteristics. Due to its rapid tunability, large modulation amplitude, and straightforward fabrication process, this THz sensor is expected to have wide-ranging applications.

Pear ring rot caused by Botryosphaeria dothidea severely affects the quality and storage life of pear fruit. Plant endophytes are potential new sources of biocontrol agents due to their abilities to produce novel bioactive products. In this work, we focused on an endophytic strain B1, which was isolated from roots of Dendrobium huoshanense. Notably, strain B1 exhibited strongly inhibitory effects against the growth of Botryosphaeria dothidea with an inhibition rate of 73.2% after 1 week of co-cultivation. Furthermore, it displayed a broad-spectrum inhibitory activity. Strain B1 also effectively delayed the onset of ring rot caused by B. dothidea in pear fruit and reduced the lesion diameter by 73.4%. Strain B1 was identified as Bacillus velezensis based on core-genome phylogeny. Genome mining with antiSMASH revealed 13 potential gene clusters involved in antimicrobial metabolites. Three main groups of lipopeptides (surfactin, iturin and fengycin) were identified with MALDI-TOF-MS, and the lipopeptides in the inhibition zone were greatly upregulated by B. dothidea; especially, fengycin isoforms were detected with higher abundance. These results proved that lipopeptides produced by strain B1 had significant antagonistic effect on B. dothidea. In conclusion, our results suggest that strain B1 has potential applications in the biocontrol of B. dothidea in postharvest fruit.