The invention relates to an automatic calibration method related to spectra of a spectrometer for examining substances in the pharmaceutical, chemical, cosmetic, dye, plastics, rubber, and foodstuff industries.

A method is disclosed to reduce the time needed to analyze data generated from very similar samples, emphasizing the differences obtained, from the combined technique of liquid chromatography/mass spectrometry, a technique used to analyze the composition of materials. The method quickly compares each mass chromatogram between similar samples and displays the limited number of mass chromatograms that are different to the analyst.

Spectra obtained from spectrographic readings from a sample can be filtered for artifacts, e.g., distorted data points arising from cosmic ray interference, by subtracting one spectrum from another to obtain a difference spectrum; smoothing the difference spectrum; and then calculating the difference between the smoothed and unsmoothed difference spectra to obtain a noise spectrum. The resulting noise spectrum, which represents localized differences between the original spectra, can then be reviewed for readings which exceed the norm by some predetermined amount (e.g., readings which exceed the average level of the noise spectrum by some percentage). These excessive readings constitute distorted data points, and the corresponding points on the spectra can have their values adjusted to eliminate the excessive readings, thereby removing the artifacts.

Method and system for processing and evaluating data from combination chromatographic and mass spectrometric data. The methods and system comprise data smoothing and determination of an quality factor for a chromatogram of a data set based on a determined entropy value. In one embodiment, the methods and system further comprise data baseline correction prior to entropy value determination. In a further embodiment, a correlation is determined between chromatograms processed according to the invention and one or more chromatograms of one or more data sets. Preferably, said correlation is performed using a multivariate analysis.

An unknown polymer is pyrolyzed by a pyrolizer and a produced mixture of pyrolytic products is separated into the pyrolytic products by a separating column. The separated pyrolytic products are detected by a mass spectrometer to generate a chromatogram. Ion intensities contained in all total ion intensities making up peaks in the chromatogram are added, for each ion of the same mass, to plot a combined mass spectrum of the unknown polymer. Combined mass spectra of a plurality of known polymers are generated in the same manner as with the combined mass spectrum of the unknown polymer, and stored as a library. The library is searched for the combined mass spectrum of one of the known polymers which agrees with a high probability percentage with the combined mass spectrum of the unknown polymer.