Fast aroma profiling to detect invert sugar adulteration with zNose™

Rapid aroma profiling of food products is a potential technique for at‐line food quality evaluation. In this work the potential of zNose™, a surface acoustic wave‐based sensor, was tested for honey quality assessment. Buckwheat honey was purposely adulterated with different levels of beet and cane invert sugar, and its aroma profile was measured after different periods of headspace equilibration. PCA using the relative peak areas as well as the full zNose™ spectra resulted in a clear separation between honey, and beet and cane invert sugar adulterants in the mixtures. PLS models were developed for quantitative estimation of adulterants using the entire spectra as well as the relative peak areas. Better predictions were obtained with the PLS models based on spectra than with those based on relative peak areas. A correlation of validation of 0.98 was obtained between predicted and measured percentage of adulteration. This model was also successfully validated with an external set of honey mixtures, resulting in an average deviation of 3% adulteration between the predicted and reference values.