抄録
Identification of mass spectra by searching a file of known spectra employing the computer (IBM360/44)was discussed. Mass spectra of about 6000 compounds were stored into a disk. To take an advantage of random access for the disk, 50 similar spectra for each unknowns are sorted by pre-search using the 7 strongest peaks, and then the more similar ones are searched from them using the 10 strongest peaks. The names and chemical formulas are printed out with the similarity index in decreasing order down the list, and mass spectra of unknown compounds and searched ones are plotted by computer if desired. Time taken for the identification of an unknown was reduced to about a minute by reading in up to 10unknowns at once and keeping separate lists for each comparison. This sytem combined with GC/MS on-line data reduction system allowed in principle automatic compound identification.
