A procedure for decreasing uncertainty in the identification of chemical compounds based on their literature citation and cocitation. Two case studies

An identification procedure connected with selection of candidates for identification according to high values of their literature citation and cocitation is suggested. The citation rate of the compound is the number of different literature units, such as papers, patents, etc., recording its name. The cocitation rate is the number of such units mutually recording the names of two corresponding compounds or the compound and the sample/matrix. General citation of a chemical compound is assumed to be related to the prior probability of its being contained in a sample to be analyzed. This citation measures abundance and popularity of the compound. Cocitation of a compound with a known/identified constituent of a sample is related to their mutual similarity in structure or properties, origin, use, etc. This data processing method is validated by counting citations and cocitations for detected impurities in pure n-hexane and naphthalene, polycyclic aromatic hydrocarbons in waste gas, as compared with counts for similar or dissimilar compounds that are absent in the samples. The analytes are preidentified by combined gas chromatography and mass spectrometry techniques. A median and a mean value of citation and cocitation are always higher for the group of unambiguously identified compounds. A difference between identified and similar compounds in citations or cocitations may be rather insignificant, with combined evaluation of both indicators distinguishing these groups. Chemical dissimilarity results in a large difference in cocitation values.