LCGC spoke to Giorgia Purcaro of the Department of Food Science at the University of Udine, Italy, and Chiara Cordero of the Department of Drug Science and Technology at the University of Turin, Italy, about the sensory evaluation of olive oil and the benefits of a comprehensive approach.
Q: Olive oil is classified by law according to both the chemical composition and the sensory evaluation. Is olive oil the only food product analyzed by sensory evaluation?
Giorgia Purcaro: Sensory evaluation of foods has been widely applied in industry to study consumer acceptance, which influences business decisions and guides product development to get closer to a benchmark product. In academia the correlation between sensory perception and chemical composition and texture has always been an important topic of study in food science. Sensory quality has been used as an additional requirement in several food certifications, such as PDO (product designation of origin). Sensory evaluation has never been used as a legal parameter to define any kind of food products before olive oil. Olive oil has been a very important product of the Mediterranean basin since ancient times, and it has always been recognized, from both a quality and a commercial viewpoint, as high value oil compared to other vegetables oils; therefore, it has been subjected to frauds by the addition of oils of lower quality and price.
Since the first European Regulation citing olive oil, which was enacted in 1966 (Regulation no. 136/1966) (1) and directed to market regulation on fats and oils within Europe, both chemical and sensory evaluations have been used for assessment of purity and quality of olive oil. Although the definitions of sensory quality were initially “blurry” (“perfect flavour”, “good flavour”, and “off-flavour”), they were elucidated in detail in the European Regulation no. 2568, which was enacted in 1991. In fact, this regulation introduced — along with a long list of chemical analyses with an accurate description of the analytical methods — a detailed explanation of the procedure to be performed in the training of panellists, how to carry out the panel session, and the attributes to be used to evaluate the sensory quality of olive oil (2).
Q: What does the sensory evaluation involve and what are the drawbacks?
GP: Sensory evaluation is fundamental in classifying olive oil within the specific commercial class, namely extra virgin, virgin, and lampante olive oil [Reg. CEE 2568/91, Reg CE 796/02; Reg CE 640/08] (2–4).
For example, extra virgin olive (EVO) oil, apart from respecting all the required chemical parameters, has to present a fruity aroma and no defects (muddy, fusty, rancid, musty, vinegary, or metallic) that have originated during olive or oil storage. Described in detail by both the European Regulation CEE 2568/91 (2) (and following modifications) and by several International Olive Oil Council (IOC) documents (COI/T.20/Doc. No 15/Rev. 6; COI/T.20/Doc. No 13; COI/T.20/Doc. No 4) (5–7), the sensory analysis is based on the judgement of a panel of assessors, composed of a minimum of eight testers (generally 12) headed by a panel leader, who has to motivate the judges, process the data, interpret results, and draft the final report.
The sensory assessment (both tasting and smelling) is performed according to codified rules, in a specific tasting room, and using a specific vocabulary (COI/T.20/Doc. No 4) (7). The panel has to be continuously trained and the performance constantly evaluated, usually using reference samples provided by the IOC, to verify the reliability of the results and harmonize the perception criteria. However, despite a rigid codification of all the aspects of sensory analysis, several drawbacks can be pinpointed: (a) Organization of a panel session is not easy since panellists are usually employed in other jobs; (b) each session is very tiring, consequently only a limited number of samples (maximum of six) can be analyzed for each session; (c) low reproducibility has been observed among different panels, mainly as a result of uneven training, related also to the difficult availability of reference standards — in particular when presenting only one defect at a time. Therefore, an objective sensory evaluation is costly and time-consuming (8).
Q: What analytical techniques have been used to analyze olive oil previously and what are the benefits of a comprehensive approach?
GP: To study the volatile profile of olive oil, the most suitable technique is gas chromatography (GC), usually coupled with a mass spectrometer (MS) detector for identification purposes. A few studies have reported the use of an olfactometric detector to describe the odour perception generated by specific compounds (9). Several sampling techniques have been used, mainly static (10), dynamic headspace (HS), and purge-and-trap techniques (9, 11) and more recently solid-phase microextraction (SPME) (12, 13). The latter technique represents an easy-to-use pre-concentrating method, alternative to the dynamic HS that provides a representative picture of the sample volatiles as a function of analyte volatility and relative polymer affinity.
Many studies (14–16) have been performed to unravel the complex mixture of olive oil volatiles and to understand their relationship with quality. Aldehydes, ketones, alcohols, esters, hydrocarbons, furans, lactones, and others still unknown are responsible for the whole aroma of olive oil. However, their exact composition and relative profile may be affected by several parameters, therefore complicating the identification of target compounds. Cultivar, geographic region, fruit ripeness, processing method, and storage may significantly affect the volatile composition (17). All of these variables affect the intensity and quality of the green and fruity perception, while the presence of defects is mainly a result of inappropriate manufacturing practices.
A few studies (9, 18) have focused on the determination of target compounds related to standardized defects using standards provided by the IOC, and mainly applying the one-variable-each-time approach. However, despite an extensive knowledge of the volatile profile of olive oil, any recognized analytical procedure has not replaced sensory evaluation yet. Several aspects to explain this situation have been recently discussed, and the need to improve the analytical determination, along with the pre-concentration procedure, both in terms of selectivity and resolution, has been highlighted as an important issue from the analytical viewpoint (19).
With this perspective, the use of comprehensive GC (GC×GC) coupled with MS, in combination with HS sampling and performed with different techniques, and implemented with advanced data elaboration tools, can be a valuable and more informative analytical approach. In fact, GC×GC is a relatively novel technique (20), which improves selectivity, using different column stationary phases in each dimension, resolution (theoretically corresponding to the product of the peak capacity of each column), and provides a useful support for identification of “unknowns” thanks to the formation of chemically-similar compound patterns when a homologous series of compounds are present.
Furthermore, depending on the kind of interface (modulator) used between the two columns, the comprehensive system can provide an important increase in sensitivity, eventually allowing the detection of informative volatiles present at low concentrations, but with a low odour threshold.
Moreover, the selection of a comprehensive set of samples (extra virgin, virgin, and lampante oil) to be analyzed together allows a wider set of informative data to be collected, and requires suitable chemometrics to derive useful information.
Q: How easy is comprehensive GC to implement in practice?
GP: The introduction of the comprehensive GC×GC system in 1991 by Liu and Philips (20) has been one of the greatest advancements towards increasing separation power in GC applications. The separation process is performed on two columns of different selectivity (with almost orthogonal discrimination principles) connected in series and interfaced by a modulator, which represents the heart of the system. Since the introduction of this technique, different kinds of modulators have been developed over the years, which can be classified on the basis of the specific operational characteristics in thermal and pneumatic modulators (21). The setting of a GC×GC system is a rather simple task. Essentially, it can be achieved using any commercially-available GC instrument by installing a modulator between two different columns. In our recent work (22), one of the two laboratories involved developed its own lab-made cryogenic (CO2) modulator, which performed comparably to the commercial one used in the other laboratory.
Although hardware configuration is quite straightforward, method optimization can be a cumbersome issue and is probably the main factor inhibiting the widespread use of GC×GC. Indeed, strong knowledge of chromatographic basic theory and experience in different branches of the GC field are of great help for rapid and effective GC×GC optimization (23, 24).
However, a bigger effort in method optimization is largely rewarded by the advantages obtained from such a technique. In fact, the introduction of GC×GC has not only allowed a better understanding of complex matrices, but it has also given a new insight into the so-called “well-known” samples.
However, since the authors are aware that such a powerful technique is not widespread yet, the main goal of our work in defining the blueprint of extra virgin olive oil will be to identify reliable target compounds through the support of advanced analytical techniques. Development of selective sample preparation techniques will follow, which can be used routinely in conventional, more widespread, and affordable GC techniques.
Q: You mention using different and complementary sampling techniques to obtain comprehensive results. Can you comment on this further?
Chiara Cordero: Sampling is a key step of any analytical method, and when the application deals with the profiling of volatiles and high volatiles it is also challenging. In particular, in the sensory quality assessment of olive oil the sampling should provide a consistent and meaningful picture of all sensory-informative analytes, including chemicals present at trace and ultra-trace level.
With this perspective, we have adopted a well-established miniaturized sampling technique (that is, SPME applied to the HS extraction of volatiles) that enables an ad hoc tuning of the extraction selectivity by modifying physico-chemical characteristics of the extractants and of the sampling conditions (time, temperature, and volume or mass of the extraction phase). SPME is also flexible in terms of extraction efficiency and capability, offering different commercial solutions that combine single or combinations of extraction polymers whose mechanisms are based on mild interactions that limit artifact formations. Last but not least, the possibility of full integration and automation of the extraction process, therefore including sample preparation as an additional dimension in the analytical platform, is another key aspect to take into consideration, especially in view of a routine quality control.
However, we have recently experienced the advantages of a “truly” complementary sampling on milk aroma profiling by collecting analytes from HS and in-solution sampling of milk samples (25). The sampling design has been inspired by the pathways aroma compounds follow to reach the regio olfactoria: orthonasal and retronasal. The characterization of volatiles should be consequently run through their extraction from the HS (orthonasal) and from the liquid sample that will be introduced in the oral cavity (retronasal). The results obtained have been really straightforward and, in our opinion, represent a bridge between high-throughput screenings with a complete and almost comprehensive profiling of volatiles related to the aroma perception. In addition, by tuning the extraction capability of sampling towards a wider range of polarities and volatilities, most of the sensory descriptors of the product can be monitored and subsequently related with corresponding odour perceptions. In such a context, the information potential of each analysis increases and an almost complete sensory profile can be objectively delineated.
Q: What are your next steps in olive oil analysis? Do you see the comprehensive approach totally replacing sensory evaluation?
GP, CC: Our study, although not conclusive, has emphasized the analytical advantages of GC×GC–MS in terms of sensitivity, reproducibility (evaluated by cross-validation between laboratories and instrumental setup) and information potential, and proposes a productive investigation strategy for a reliable quality assessment of EVO and VO oils.
The results obtained to date will be implemented by extending the number of representative samples to include in the classification model and by approaching the true quantitation of key targets to obtain reference blueprints of defects to adopt as decision makers.
In this context, the investigation of alternative high concentration capacity HS sampling techniques, (that is, headspace sorptive extraction – HSSE with different polar or apolar extraction polymers) will be decisive as a future perspective, which improves fingerprinting sensitivity and enables reliable quantitation of key-odourants (by multiple headspace extraction [MHE]) (25).
In our vision, a robust, reliable, and sensitive method able to provide an objective picture of key-odourants distribution would be a valuable tool to support, or even replace, sensory evaluation.
Q: Do you think this approach to analyzing olive oil could be useful in other applications?
CC: Actually, methods and approaches adopted in our investigation of olive oil chemical blueprint, are included in the analytical tools of sensomics (26). This intriguing discipline focuses analytical efforts on revealing sensorially-active compounds to link the chemical composition with food sensory quality (flavour). The investigation is extended to all possible stimuli of the multimodal perception (aroma, taste, texture etc.) by comprehensively treating sample constituents and related attributes (physicochemical properties, concentration in-the-matrix) together with their sensory properties (odour quality, odour threshold [OT], odour activity value [OAV]) (27, 28).
The multidimensionality of the approach represents a key aspect in this perspective, enabling some of the limits of conventional platforms and methods to be overcome by applying orthogonal discrimination in multiple analytical dimensions and combining: (a) the separation of analytes based on volatility, polarity, partition coefficient, and solubility; (b) identity assessment, provided by mass spectroscopy; (c) quantitation (true concentration and relative abundance); and (d) odour activity characterization when olfactometric detection is included.
Applications include studies on cocoa and chocolate aroma characterization, roasted coffee volatiles distribution, quality assessment of hazelnuts by aroma and technological markers distribution, and wine quality by aroma profiling and origin authentication.
Q: How important are the “-omics” in food analysis?
GP, CC: The role of academia has to be oriented towards advanced, innovative, and novel techniques within the field of the modern “–omics” disciplines, such as foodomics, flavoromics, and sensomics in food fields. Such techniques are extremely powerful and they allow a huge amount of data that requires advanced chemometrics to extrapolate the useful information to be collected. It may seem a rather complicated task and far from everyday practice, because, as is usually the case, the solution to a problem is simple and rather linear. However, when working with complex matrices such as foods, one has to isolate information deriving from informative analytes (such as compounds characterizing botanical origin, technological process, or sensory quality) from interferences to get close to a truly comprehensive practical solution that is able to act as “decision maker”.
A comprehensive approach is fundamental to have a complete overview of the sample to be studied and to provide a reliable indication on the variables to focus on to solve any kind of problem. We are aware of the industrial needs for easy and straightforward solutions to practical problems; we think that exploiting comprehensive approaches and extending the analytical investigation to the whole chemical pattern would increase knowledge of the effects of processing, storage, fermentation, and biotransformation on the overall quality of a food product.
In this context, academia and industry would work synergically towards a common objective and promote a truly technological transfer of knowledge and methods.
1. Reg. CEE n. 136/66 Off. J. Eur. Comm. L172 (1966) 3025.
2. Reg. CE n. 2568/91 Off. J. Eur. Comm. L347 (1991) 69.
3. Reg. CE n. 796/02 Off. J. Eur. Comm. L128 (2002) 8.
4. Reg. CE n. 340/08 Off. J. Eur. Comm. L178 (2008) 11.
5. International Olive Council (IOC). Document COI/T.20/Doc. No 15/Rev.6. Madrid, Spain, 2013
6. International Olive Council (IOC). Document COI/T.20/Doc. No 13. Madrid, Spain, 1996.
7. International Olive Council (IOC). Document COI/T.20/Doc. No 4. Madrid, Spain, 1987.
8. D.L. García-González and R. Aparicio, Grasas y Aceites 53(1), 96–114 (2002).
9. M.T. Morales, G. Luna, and R. Aparicio, Food Chem. 91, 293–301 (2005).
10. I. Marcos, J.L. Pérez, M.E. Fernández, C. García, and B. Moreno, J Chromatogr A. 945, 221–230 (2002).
11. A. Kanavouras, A. Kiritsakis, and R.J. Hernández, Food Chem. 90, 69–79 (2005).
12. J.-F. Cavalli, X. Fernández, L. Lizzani-Cuvelier, and A.-M. Loiseau, J Agric Food Chem. 51, 7709–7716 (2003).
13. S. Vichi, A.I. Castellone, L. Pizzale, L.S. Conte, S. Buxaderas, and E. López-Tamames. J Chromatogr A. 983, 19–23 (2003).
14. C.M. Kalua, M.S. Allen, D.R. Jr Bedgood, A.G. Bishop, P.D. Prenzler, and K. Robards, Food Chem. 100, 273–286 (2007)
15. F. Angerosa, Eur. J. Lipid Sci. Technol. 104, 639–660 (2002).
16. G. Luna, M.T. Morales, and R. Aparicio, Food Chem. 98, 243–252 (2006).
17. F. Angerosa, M. Servili, R. Selvaggini, A. Taticchi, S. Esposito, and G. Montedoro, J. Chromatogr. A 1054, 17–31 (2004).
18. G. Procida, A. Giomo, A. Cichelli, and L.S. Conte, J. Sci. Food Agric. 85, 2175-2183 (2005).
19. R. Aparicio, M.T. Morales, and D.L. García-González, Eur. J. Lipid Sci. Technol. 114, 1114–1125 (2012).
20. Z. Liu and J.B. Phillips, J. Chromatogr. Sci. 29, 227–231 (1991).
21. P.Q. Tranchida, G. Purcaro, P. Dugo, and L. Mondello, Trends Anal Chem. 30(9), 1437–1461 (2011).
22. G. Purcaro, C. Cordero, C. Bicchi, L.S. Conte, J.Chromatogr. A (under revision).
23. M. Adahchour, J. Beens, and U.A.Th. Brinkman, J.Chromatogr. A 1186, 67–108 (2008).
24. H.J. Cortes, B. Winniford, J. Luong, and M. Pursch, J.Sep. Sci. 32, 883–904 (2009).
25. C. Cordero, C. Cagliero, E. Liberto, L. Nicolotti, P. Rubiolo, B. Sgorbini, and C. Bicchi, J.Chromat. A. 1318, 1–11 (2013).
26. J. Kiefl, G. Pollner, and P. Schieberle, J. Agric. Food Chem 61(22), 5226–5235 (2013).
27. J. Charve, C. Chen, A.D. Hegeman, and G.A. Reineccius, Flav. Fragr. J. 26, 429-440 (2011).
28. W. Grosch, Chem Senses 26, 533-545 (2001).
Giorgia Purcaro obtained her Masters degree in food science and technology in 2002 with "summa cum laude" and she gained her PhD in food chemistry in 2008, both at the University of Udine (Italy) under the supervision of Professor L. Conte. In 2006, she was awarded the scholarship for international mobility within the Regional Operative Programme, Objective 3- European Social Fund 2000–2006, axe D, Measure D4 “Improvement of the human resources in the research and technological development field”, and she spent six months at the RMIT University in Melbourne (Australia) under the supervision of Professor P.J. Marriott, working on the development of rapid analytical techniques based on hyphenated chromatography methods (GC×GC) for food safety assessment. From 2008 to 2011 she worked in the laboratories of Professor L. Mondello at the University of Messina (Italy), where she developed comprehensive chromatography methods (GC×GC, LC×GC) with different interface and detection systems for the investigation of natural complex food matrices. Since 2012 she has worked as assistant professor at the Department of Food Science at the University of Udine.
Her research interests include development of conventional chromatography techniques (HRGC–FID, HRGC–MS, HPLC–UV, HPLC–FLD), as well as advanced multidimensional and comprehensive chromatography ones (GC×GC, LC–GC, LC×GC) to apply to food analysis.
Her work on the capability of a quadrupole MS for GC×GC detection was awarded with the prestigious Leslie S. Ettre Award in 2010 for the young scientist who presented the most interesting original research in capillary gas chromatography with an emphasis on environmental and food safety award during the 34th International Symposium of Capillary Chromatography & 7th GC×GC Symposium in Riva del Garda (Italy)
Chiara Cordero received her Masters degree in pharmaceutical chemistry and technology in 1998 with "summa cum laude" at the University of Turin (Italy), supervised by Professor Carlo Bicchi. Thanks to her curriculum studiorum she was awarded with the Silver Medal (Medaglia d'argento dell'Università di Torino) as best graduate student in pharmaceutical chemistry and technology for the academic year 1996–97. Since 2001 she has worked as assistant professor of Food Chemistry at the Dipartimento di Scienza e Tecnologia del Farmaco and Faculty of Pharmacy at the University of Turin.
In 2008, she was awarded the Leslie S. Ettre Award for the young scientist who presented the most interesting original research in capillary gas chromatography with an emphasis on environmental and food safety during the 32nd International Symposium on Capillary Chromatography (ISCC) and 5th GCxGC Symposium. The study focused on the application of GCxGC/qMS for group-type and fingerprint analysis approaches in the characterization of roasted food matrices.
Her research activities are focused on the development of innovative and advanced instrumental configurations for GC×GC coupled to MS applied to quali-quantitative characterization of complex food samples (roasted coffee, roasted hazelnuts, herbal extracts, and essential oils), also through the application of "omics" approaches (multidimensionality in sample preparation and separation) to molecular sensory science, followed by advanced data elaboration approaches.