Storage time of nut spreads using flash gas chromatography E-nose combined with multivariate data analysis

The quality assessment, in terms of lipid oxidative status, of food products stored in the long-term is of great importance, especially those with a high lipid content. Specifically, companies working in this sector need feasible, simple, and fast techniques that are suitable for quality or process control. Herein, a fingerprinting approach, based on headspace analysis carried out by flash gas chromatography electronic nose (FGC E-nose) and multivariate data analysis was applied to pistachio and gianduja spreads. These samples, differently packaged, were stored in climatic chambers at 40 °C for 180 days and their headspace fraction was analyzed periodically for a total of 15 sampling times. Principal component analysis showed a clear separation according to the packaging type for both pistachio and gianduja samples. Partial least squares regression models were developed to predict the storage time considering the aggregated data (R2 up to 0.985, RMSEP = 6.16 days) or separately (R2 up to 0.989, RMSEP = 5.71 days). Based on the obtained residual prediction deviation (RPD from 4.4 to 8.5 in prediction), the models can be considered suitable for use in quality control in an industrial environment.