Data handling of complex GC-MS chromatograms: characterization of n-alkane distribution as chemical marker in organic input source identification.

The paper describes a signal method for processing GC-MS signals to extract usable information hidden in the chromatogram thus reducing the labour and time required to handle the data and increasing the quality and objectivity of the results. The method is focused on two relevant parameters for identification and characterization of the n-alkane series present in complex samples (in particular the C(14)-C(33) terms): the number of n-alkanes, n(max), and the Carbon Preference Index (CPI) describing the odd/even carbon-number predominance. This is a key diagnostic parameter to determine the biogenic and anthropogenic nature of n-alkane sources, useful as chemical markers in source identification and differentiation. The method is a further extension of the approach based on the AutoCovariance Function (ACVF(tot)): new mathematical equations have been derived and a new computation algorithm implemented to extract information on the n-alkane series--n(max) and CPI--directly from the EACVF(tot) computed on the acquired chromatographic signal. The procedure was validated on simulated chromatograms where the distribution of the terms of the series describing experimental GC signals was known: the obtained results prove that the parameters n(max) and CPI of the homologous series can be estimated with good accuracy and precision. The method applicability was tested on experimental chromatograms of real samples: gasoils and plant extracts were studied to identify n-alkane distribution patterns characteristic of petrogenic and natural samples.