For realizing efficient solar to hydrogen energy conversion based on photocatalytic technology, it is important to explore a photocatalyst with wide-range solar absorption and high electron–hole separation efficiency. With a built-in electric field, the recently synthesized Janus MoSSe is intrinsically beneficial for promoting the separation of photo-generated electrons and holes. Thus in this work, we examine the possibility of MoSSe as an efficient water-splitting photocatalyst and the effects of isotropic and uniaxial strains by the first-principles calculations. It is interesting to find that MoSSe exhibits pronounced visible-light absorption efficiency, proper valence and conduction band positions for initializing the redox reactions of H₂O, and high carrier mobilities. Moreover, the band gap of MoSSe is decreased and the direct–indirect band gap transition occurs upon tensile strain, which can not only extend the light absorption range, but also reduce the recombination of photo-generated carriers. Furthermore, H₂O molecules adsorb more strongly on the MoSSe monolayer surface than on the MoS₂ surface, which is also beneficial for the surface water-splitting reactions. These insights provide eloquent evidence that the Janus MoSSe monolayer is potentially an efficient and wide solar-spectrum water-splitting photocatalyst.

The electronic structures of TiO₂(001) films exhibit pronounced odd–even oscillations with the number of atomic layers in them.

Infrared imaging is widely used in astronomical observation, medical diagnosis, and military applications. In recent years, metasurface technology has provided an unparalleled platform for the development of miniaturized and integrated infrared imaging systems. However, metasurfaces normally have inevitable chromatic aberration due to the high phase dispersion of the building blocks, which makes broadband achromatic infrared imaging difficult to realize. In this paper, we propose a polarization-insensitive metalens with a numerical aperture of 0.38 that can eliminate chromatic aberration for unpolarized incidences with the wavelength ranging from 3 to 5 µm. The simulated results show that within the design bandwidth, the proposed device achieves near-diffraction limit focusing and can increase the fill factor of infrared focal plane array pixels by 2.3 times, from 11.1% to 36.4%, with an excellent optical crosstalk performance of about 2.72%. Our work may pave the way for the practical application of achromatic metalenses in mid-wave infrared imaging equipment.

The traditional ray tracing approach for infrared (IR) target simulation takes into account the radiance emitted from the target in the scene, resulting in an incomplete IR signature of a target. We present a practical method to enable the identification of the signature by considering radiation exchanges among all sources in the scene. Any source would contribute radiance to a target of interest. Therefore, the radiance of a target received by an IR imager includes self-emission and reflectance, where the reflectance is contributed due to radiation exchanges from other sources. Field experiments were performed to demonstrate our proposed method by comparisons of radiance profiles between our simulated images and the real scene images. It shows that our simulated images achieve a better match with the data than those generated by the traditional approach.

Recent advances in NeRF-based inpainting have enabled the completion of masked regions across multi-view images. However, two major challenges remain: generating accurate masks efficiently in the presence of complex multi-object interference and maintaining view consistency without floating artifacts in large-scale, unbounded 360° environments. To address these challenges, we propose a 3D inpainting framework with two principal components. First, a depth-aware mask-generation pipeline produces high-quality, view-consistent masks from sparse annotations. Second, spherical parameterization is combined with a joint objective comprising smooth inter-layer and original-scene constraints to suppress floating artifacts and content drift in unbounded scenes. We also introduce IM2360, a multi-object 360° dataset for evaluating 3D inpainting methods. Experiments show that our approach outperforms existing NeRF-based inpainting methods in PSNR, LPIPS, and FID. On a single NVIDIA GeForce RTX 4090, our method requires an average of 25.21 min of training per scene, compared with 98.14 min for SPIn-NeRF, 38.71 min for OR-NeRF, and 11.29 min for InFusion. These results indicate that the proposed method provides a favorable balance between reconstruction quality and computational cost for high-fidelity restoration of complex immersive scenes.