Abstract: To address the current gaps in incorporating cutting-edge scientific content and fostering students' innovative capacities in university chemistry experiments, this study developed a comprehensive teaching module that seamlessly integrates supramolecular chemistry advancements with fundamental physical chemistry experiments. Using the host-guest complexation of rhodamine B with α-, β-, and γ- cyclodextrins as a model system, we employed UV-Vis spectrophotometry to monitor molecular interactions. Departing from conventional verification experiments, this work introduces an innovative AIassisted data processing approach: students were guided to develop MATLAB-based nonlinear fitting programs through interactive AI learning, employing a 1:1 binding model to accurately determine thermodynamic parameters including bonding constants (K) and Gibbs free energy changes (ΔG) from absorbance-concentration data. This pedagogical design facilitated students' complete engagement in the scientific research cycle: hypothesis generation → experimental validation → program development → data analysis → results interpretation → reflective reporting. The implementation demonstrated that this approach not only effectively combines classical instrumental analysis with modern computational methods, but also systematically enhances students’ computational thinking, data analysis skills, and independent research capabilities through authentic scientific problem-solving. Furthermore, it establishes a practical framework for advancing experimental teaching reform through the incorporation of frontier research achievements.

Key words: Supramolecular chemistry, Reform of laboratory instruction, Bonding constant, Innovation capability cultivation