Advances in band assignment in near-infrared spectroscopy: Principles, methods, and applications

Near-infrared spectroscopy has gained extensive application across numerous fields owing to its distinct advantages in efficiency, non-destructiveness, and real-time online monitoring capabilities. However, its inherent characteristics of severe overlap of spectral bands and weak absorption signals have made the precise analysis of the correlation between spectral features and molecular structure a core challenge in this field. This review systematically examines research progress in near-infrared spectral band assignment. Building on the theoretical foundation of anharmonic vibration mechanisms, it categorizes three mainstream methodological approaches: experimental perturbation analysis, which separates overlapping bands by modulating external conditions such as temperature and concentration; data-driven methods that integrate chemometrics and explainable artificial intelligence to extract physically meaningful characteristic wavelengths from macroscopic spectral data; and theoretical computation, which employs quantum chemistry and anharmonic vibrational algorithms to achieve precise spectral predictions directly from molecular structures. The article further discusses representative applications of band assignment in quantitative composition analysis, process monitoring, material identification, and biomedical diagnostics, while illustrating how this approach can elevate spectral analysis from "black-box" prediction to the level of mechanistic insight. The article concludes by proposing future development avenues focused on the deep fusion of multi-paradigm approaches, the enhanced integration of artificial intelligence, and the advancement of miniaturized and interdisciplinary applications. Through the synergistic advancement of theory, computation, and data, the field of NIR spectral band assignment will continue to enable deeper breakthroughs across scientific and industrial domains.