Abstract: Current instrumental analysis experiments in nuclear magnetic resonance (NMR) spectroscopy frequently overlook the crucial aspects of teaching and training in spectral interpretation. This study presents the independent development of “AI NMR Assistant”, an intelligent NMR spectral interpretation system based on the DP5 software package, designed for evaluating undergraduate NMR spectral analysis training. The system introduces an innovative “AI + NMR instrumental analysis” teaching methodology. AI NMR Assistant combines four core functionalities: automated NMR spectrum recognition and peak annotation, user practice in spectral interpretation, AI-assisted result validation, and visual feedback mechanisms, thereby establishing a comprehensive platform for undergraduate spectral analysis training. Students utilize NMR spectra obtained either experimentally or from literature to propose molecular structures through the system. DP5 subsequently performs computational analysis, providing error quantification for each carbon atom in the proposed structure. Through iterative refinement and structural adjustment, students progressively deduce the correct molecular configuration, thereby gaining deeper insights into NMR spectral interpretation. The novel “spectral database selection-practical interpretation-personalized solution” teaching paradigm not only enhances students’ analytical skills but also employs AI to streamline teaching assessments. With further development, this approach holds potential for extension to other instrumental analysis experiments.

Key words: NMR, Instrumental analysis experiment, DP5, Artificial intelligence