Physico-chemical characterization of EVOO adulterated with VOO deodorized in soft conditions

The adulteration of extra virgin olive oils (EVOO) with virgin olive oils (VOO) previously subjected to a soft deodorization process (to eliminate organoleptic defects from these latter lower quality oils), in up to a 1:1 ratio, in order to be subsequently fraudulently distributed as genuine EVOO represent a serious problem of quality assurance for the olive sector still in force today [1]. The downward evolution of the comparative prices between VOO and EVOO, together with the difficulty of detecting the presence of deodorized oils in adulterated mixtures by means of the quality, purity and sensory analyses contemplated in the legislation, has awakened the ghost of this type of fraud among the main olive oil producing countries, with Spain at the head. To serve as a starting point for the development of a reliable methodology for the unequivocal detection of adulterations of EVOO with deodorized VOO, fraudulent mixtures are needed in order to characterize them in the search for differential quality markers. Likewise, generating fraudulent mixtures requires prior obtaining soft deodorized VOO to be used as adulterants. For obvious reasons, this type of oil is not directly available for purchase in the market so, in the present study, a lab-scale pilot plan was set up to carry out deodorization experiments. An experimental design methodology was used to optimize soft deodorization process conditions in the pilot plant throughout a two-stage procedure. In the first stage, optimization was tackled as a linear design problem considering temperature and deodorization time as the main process variables involved. The perceived intensities of the residual negative attributes that still retained the lampante olive oil used as raw material once deodorized under a certain set of operating conditions were evaluated by an expert sensory panel. A two-level full factorial design with two factors was selected for the analyzed sensory defects (vinegary, fusty/muddy sediment, metallic, mould/humidity and rancid). Design resolution provided a mathematical model that allowed to correlate the process variables with the response variable studied, i.e., to predicted the value of the response variable (degree of organoleptic defect removed) as a function of deodorization conditions. Then, in a second stage, the steepest ascent method was applied to establish optimal conditions for VOO deodorization processes. Optimal deodorization conditions were selected as those for which all undesirable organoleptic properties were entirely removed. For each experiment carried out during optimization, a double sensory evaluation was performed: the first one served to determine the percentage of defect removed immediately after deodorization was completed, and the second one to assess oil stability over time.The optimally deodorized lampante oil obtained was subsequently used to generate adulterated blends with both a recently harvested (fresh) EVOO (arbequina) and a two-year-old EVOO (oliberus) in the range 0-50%. All adulterated EVOO samples thus prepared were submitted to the regulatory analyses required: acidity, peroxide value, UV spectrophotometric measurements and organoleptic evaluation [2-3]. A detailed analysis of the values obtained for the various quality and purity parameters assessed in the set of EVOO adulterated with deodorized VOO allowed to verify that these fraudulent oils were able to evade their detection in all the adulteration range studied, although they were not able to preserve their stability over time. The results of this study confirmed the need for developing and validating new analytical methodologies capable of efficiently detecting and quantifying these adulterations in routine controls at any point of the complex food chain in order to avoid the sale and consumption of products of inferior quality to the one declared