Giant enhancement of spin Hall effect of tunneling light (SHETL) is theoretically proposed in a frustrated total internal reflection (FTIR) structure with hyperbolic metamaterial (HMM). We calculate the transverse shift of right-circularly polarized light in a SiO₂-air-HMM-air-SiO₂ waveguide and analyze the physical mechanism of the enhanced SHETL. The HMM anisotropy can greatly increase the transverse shift of polarized light even though HMM loss might reduce it. Compared with transverse shift of transmitted light through a single HMM slab with ZnAlO/ZnO multilayer, the maximum transverse shift of tunneling light through a FTIR structure with identical HMM can be significantly enlarged by more than three times which reaches −38 μm without any amplification method.

Based on the phase transition of vanadium dioxide(VO₂), an ultra-broadband tunable terahertz metamaterial absorber is proposed. The absorber consists of bilayer VO₂ square ring arrays with different sizes, which are completely wrapped in Topas and placed on gold substrate. The simulation results show that the absorption greater than 90% has frequencies ranging from 1.63 THz to 12.39 THz, which provides an absorption frequency bandwidth of 10.76 THz, and a relative bandwidth of 153.5%. By changing the electrical conductivity of VO₂, the absorption intensity can be dynamically adjusted between 4.4% and 99.9%. The physical mechanism of complete absorption is elucidated by the impedance matching theory and field distribution. The proposed absorber has demonstrated its properties of polarization insensitivity and wide-angle absorption, and therefore has a variety of application prospects in the terahertz range, such as stealth, modulation, and sensing.

The introduction of “metamaterials” has had a profound impact on several fields, including electromagnetics. Designing a metamaterial’s structure on demand, however, is still an extremely time-consuming process. As an efficient machine learning method, deep learning has been widely used for data classification and regression in recent years and in fact shown good generalization performance. We have built a deep neural network for on-demand design. With the required reflectance as input, the parameters of the structure are automatically calculated and then output to achieve the purpose of designing on demand. Our network has achieved low mean square errors (MSE), with MSE of 0.005 on both the training and test sets. The results indicate that using deep learning to train the data, the trained model can more accurately guide the design of the structure, thereby speeding up the design process. Compared with the traditional design process, using deep learning to guide the design of metamaterials can achieve faster, more accurate, and more convenient purposes.

In this paper, we reported a theoretical study of a novel Surface plasmon resonance (SPR) biosensor composed of BK7 prism, gold (Au)/silver (Ag) bimetallic layer, silicon and two-dimensional (2D) materials. The bimetallic layer combines the advantages of Au and Ag and the high refractive index silicon layer enhances the electric field on the surface of the sensor, so that the sensor has a better overall performance in terms of sensitivity and figure of merit (FOM). Compared with ordinary dielectrics, 2D materials have excellent photoelectric properties, such as larger specific surface area, higher carrier density and stronger adsorption capacity, which improve the detection ability of the sensor. The sensitivity of the optimized sensor achieves 297.2°/RIU, 274°/RIU and 246°/RIU when the silicon layer is covered with graphene, MXene (Ti₃T₂C_x) and MoS₂, respectively. Compared with the traditional SPR sensor, the sensitivity of the structure has been significantly improved, and its excellent performance has broad application prospects in biosensing and other fields.