Abstract: Addressing the challenges of comprehending three-dimensional spatial arrangements in crystal structure instruction and the high barriers to classroom use of conventional 3D modeling software, this study employed generative artificial intelligence (GenAI) to develop unit-cell site diagrams, coordination polyhedron illustrations, and interactive teaching resources. These materials were implemented in classroom practice within the undergraduate course “Fundamentals of Materials Science” for materials science majors. The results demonstrated that these resources facilitated students’ understanding of A/B/X sites, BX₆ octahedra, and corner-sharing connectivity, while also enhancing their ability to convert between different structural representations. Student feedback indicated that the visualized resources improved classroom intuitiveness, while also underscoring the need for reinforced instruction on verifying AI-generated content. The findings suggest that AIGC can serve as an effective auxiliary tool for crystal structure instruction and offer a valuable reference for developing teaching resources in materials science and related science and engineering disciplines.

Key words: Generative artificial intelligence, Perovskite, Crystal structure, Visualization-based teaching, Teaching resource development