REPOSITORY > RESULTS

Abstract

Food fraud or economically motivated adulteration has been an issue throughout history. Even today, it remains a significant and growing problem, driven by globalisation, economic opportunity and the low probability and severity of punishment. Fraud is a concern for food producers, consumers, regulatory agencies and scientific organisations, and food fraud prevention is paramount to protecting consumer trust and maintaining fair and sustainable business practices. This doctoral thesis focuses on two fraudulent acts, adulterating natural flavours with cheaper synthetic counterparts and the mislabelling of the country of origin of food crops. In both cases, verification of the authenticity of flavourings and geographical traceability of products from the market can be achieved through the following steps: 1) the development of suitable analytical methods; 2) the establishment of databases; and 3) data processing using chemometric approaches. This thesis addresses all three. Finally, developed methods and databases are used to determine the authenticity of commercial products using fruit, vanilla and truffle flavourings, and selected fruit and vegetable crops: asparagus, garlic, strawberry, cherry, apple and kaki as exemplary commodities, presented in eight interrelated scientific papers/studies.
The first part of the thesis, which is devoted to developing, optimising, and validating robust analytical techniques for flavour authenticity, shows the advantages of coupling the HS-SPME extraction technique with the GC-IRMS method. Accurate and reproducible δ13C or δ2H values were achieved for key volatile organic compounds (VOCs) present in various fruits, vanilla and truffle samples, over varying concentrations measured in short analysis time, within the same run and without using solvents. In addition, isotope fractionation was not observed when using optimised measurement parameters. Moreover, the procedure that includes a multiple-point isotopic linear normalisation method with peak size/linearity correction significantly improves the measurement error of small peaks (below 1 nA) from 3 ‰ to 0.5 ‰. For selected fruit and vegetable crops, stable isotope and multi-elemental analysis were used for the geographical origin tracing.
Data interpretation in the selected cases of frauds requires extensive reference data set of authentic food samples, i.e. a database or databank against which a sample under investigation can be compared. This thesis establishes dedicated databases of authentic, sufficiently representative samples that cover the natural variation of isotopic and elemental values. Altogether, ten different extensive databases were established over the period 2017-2021. Four include isotope values of fruit, vanilla and truffle VOCs and the other six include stable isotope (C, N, O and S) values and elemental composition of selected fruits and vegetables. In this thesis, isotopic characterisation of 35 VOCs was performed for the first time.
Data analysis and the interpretation of results are also essential parts of this thesis. Comparing isotope ratios of VOCs collected in the flavour databases allowed the successful discrimination between synthetic and naturally produced VOCs, although the addition of δ2H data would further improve discrimination. However, a comparative analysis alone could not provide a definitive answer concerning the geographical origin of fruits and vegetables. For this reason, different chemometric approaches were explored. DD-SIMCA was chosen as most suitable method for determining whether the commercial product
complies with its declaration (i.e. Slovenian origin). The most important variables for classification were Sr, Ba Cs, S, Mo, Ni and Fe within elements and δ18O and δ13C within stable isotopes.
The developed approach represents an excellent foundation to verify the authenticity in real-word application. Therefore, the final part of this thesis is dedicated to verifying the naturalness of flavourings and the geographical origin of selected fruits and vegetables from the market. Results of commercial fruit, vanilla and truffle flavourings and their flavoured products imply that the authenticity can be questioned, most often within natural flavoured vanilla and truffle samples. Mislabelling of truffle species has also been identified. Also, 46 of the 124 (37 %) investigated fruit and vegetable samples did not correspond to their stated declaration.
In summary, this thesis shows that the developed analytical techniques used to establish a comprehensive set of databases and selected chemometric models ensure confidence in flavour authenticity studies and provide a sound basis for establishing an adequate traceability system for fruit and vegetables. Although the authenticity and traceability systems cover the needs of the food flavour industry and enforcement agencies in Slovenia, they can be readily transferred to other food commodities and countries. The thesis concludes by highlighting several new research questions and new avenues of research, especially in further developing HS-SPME-GC-IRMS for δ 2H determination of VOCs and more robust generalised DD-SIMCA models.