Prediction of chemical structure of aromatic hydrocarbons by pattern recognition of spectral lines obtained by fluorescence spectrometry in a supersonic jet

A technique based on pattern recognition of data obtained by supersonic jet spectrometry is employed for the prediction of chemical structure. The degree of similarity is evaluated quantitatively by calculating a cross correlation factor between sample and reference molecules. A probability density function is determined by fitting the data to a specified equation. The functional group and its position are also predicted by a similar technique. The pattern recognition provides a method for prediction of the chemical structure and is applicable to samples that have not been examined by supersonic jet spectrometry.