Target vs spectral fingerprint data analysis of Iberian ham samples for avoiding labelling fraud using headspace – gas chromatography–ion mobility spectrometry

Abstract

The data obtained with a polar or non-polar gas chromatography (GC) column coupled to ion mobility spectrometry (IMS) has been explored to classify Iberian ham, to detect possible frauds in their labelling. GC–IMS was used to detect the volatile compound profile of dry-cured Iberian ham from pigs fattened on acorn and pasture or on feed.Due to the two-dimensional nature of GC–IMS measurements, great quantities of data are obtained and an exhaustive chemometric processing is required. A first approach was based on the processing of the complete spectral fingerprint, while the second consisted of the selection of individual markers that appeared throughout the spectra. A classification rate of 90% was obtained with the first strategy, and the second approach correctly classified all Iberian ham samples according to the pigs’ diet (classification rate of 100%). No significant differences were found between the GC columns tested in terms of classification rate.