Caffeine is an alkaloid with a purine structure and has been well known for centuries due to its presence in popular drinks-tea and coffee. However, the structural and spectroscopic parameters of this compound, as well as its chemical and biological activities, are still not fully known. In this study, for the first time, we report on the measured oxygen-17 NMR spectra of this stimulant. To support the assignment of our experimental NMR data, extensive quantum chemical calculations of NMR parameters, including nuclear magnetic shielding constants and indirect spin-spin coupling constants, were performed. In a theoretical study, using nine efficient density functionals (B3LYP, BLYP, BP86, CAM-B3LYP, LC-BLYP, M06, PBE0, TPSSh, wB97x), and in combination with a large and flexible correlation-consistent aug-cc-pVTZ basis set, the structure and NMR parameters were predicted for a free molecule of caffeine and in chloroform, DMSO and water. A polarized continuum model (PCM) was used to include a solvent effect. As a result, an optimal methodology was developed for predicting reliable NMR data, suitable for studies of known, as well as newly discovered, purines and similar alkaloids. The results of the current work could be used in future basic and applied studies, including NMR identification and intermolecular interactions of caffeine in various raw materials, like plants and food, as well as in the structural and spectroscopic characterization of new compounds with similar structures.
Keywords: 17O NMR; DFT; caffeine; molecular modeling; molecular structure.