ZnFe₂O₄ microspheres were prepared by solvothermal method, and a novel ZnFe₂O₄@TiO₂ core-shell composite photocatalyst was prepared by ultrasonic (denoted as ZT-x) and mechanical stirring (denoted as ZTM-1.2). The morphology, structure, magnetic, and optoelectronic properties of the catalyst were investigated comprehensively, and the degradation performance of the catalyst was explored through the photocatalytic degradation of Rhodamine B (RhB) under UV light. The concentration of RhB was 10 mg/L, and the catalyst concentration was 0.3 g/L. ZT-1.2 showed the best photocatalytic degradation activity, and the degradation rate of RhB reached 97.3% within 60 min. The degradation ability of the catalyst was further evaluated by Methylene blue (MB), Methyl orange (MO), Phenol, and Ofloxacin (OFX). ZT-1.2 also exhibited excellent stability. The improved catalyst degradation performance was attributed to constructing a Z-type heterojunction. Moreover, the low-efficiency degradation of ZTM-1.2 was caused by catalyst agglomeration and low TiO₂ loading, confirming the superiority of the ultrasonic method and providing a new method for the preparation of magnetically recoverable TiO₂-based core-shell photocatalyst.

Diffraction gratings are becoming a preferred option for waveguide head-mounted in–out coupling devices due to their flexible optical properties and small size and light weight. At present, diffraction waveguide coupling devices for AR head-mounted displays (HMD) have difficulties such as a long development cycle and complicated processing. In this paper, we first establish a set of two-dimensional (2D) grating ray tracing models, based on which we determine the initial architecture of the dual-region two-dimensional exit pupil expansion (2D-EPE) AR-HMD holographic waveguide diffraction system. Second, we propose a honeycomb coupled grating array and optimize the optical energy utilization and brightness uniformity of the holographic waveguide and use a custom dynamic linked library (DLL) function to implement the ray tracing of the 2D grating and add a probabilistic splitting function to the DLL, which reduces the single simulation time from 11.853 min to 1.77 min. We also propose a holographic lithography device composed of holographic optical elements (HOEs) and a method for preparing HOEs. Finally, in order to obtain the diffraction efficiency preoptimized by the above DLL for the uniformity of the exit pupil brightness and light energy utilization, we inverse design with the preparation process parameters as the optimization variables and develop the adaptable electromagnetic calculation program Holo-RCWA. Using Holo-RCWA with nondominated sorting genetic algorithm II (NSGA-II), we inverse design to obtain the process parameters satisfying the diffraction efficiency distribution, and the optimization time of the whole system is reduced from 2–3 days to 10 h. This work is of great significance for AR/VR applications.