Abstract
A computer software system for structure identification using nuclear magnetic resonance (NMR) spectra is described. For this purpose search files are created from full spectral patterns. In the search files peak information is stored, including positions and intensities as well as other search items such molecular formula, molecular, weight, etc. in order to assist in identification. The1H-NMR area intensity is used rather than the peak height intensity which is used in other spectroscopic methods. This enables us to utilize, in addition to peak positions, the ranges of peak positions for the complex patterns characteristic of1H-NMR spectra as input data for the identification. The problems and limitations of the structure identification by1H-NMR spectra relating to the frequency dependence are discussed. A brief description is given regarding structure identification by13C-NMR spectra in connection with1H-NMR spectra.
