PTR-MS applications in food

Mass spectrometry has been a powerful tool used in food analysis for many years. In fact, the technique as we recognise it today is almost 100 years old, having its roots in the work of Dempster (1918) and Aston (1919), and there is no better technique for identifying the wide range of molecules that are found in food (or any other matrix).

Hence, mass spectrometry already has many uses in food analysis, such as identifying the trace contaminants that gives rise to taints and offflavours, identifying chemical markers that are relevant to ingredient authenticity/purity, identifying molecules that are important in flavour, or that might be food toxins, and so on.

New developments

In the mass spectrometer, molecules are ionised (given an electric charge), and then identified on the basis of their absolute mass, and on the mass of their fragmentation products when the molecule is broken down into smaller pieces. The combination of mass data from any given molecule, whole and in pieces, is usually characteristic of that molecule and hence will lead to a positive identification. Two molecules may have the same absolute mass, they may even have the same constituent atoms, but when they have a different way of joining these atoms together (structure) the fragmentation products will be different. It is this difference that ultimately allows the original molecules to be distinguished and identified.

Incremental improvements in the technology have produced mass spectrometers that are now much more powerful and more sophisticated than the earliest devices, but the basic principle as described above remains the same.

One of the more interesting developments in recent years has been the introduction of Proton Transfer Reaction-Mass Spectrometry (PTR-MS), a variant not on the way that ionised particles are detected, but on the way they are produced. The PTR-MS technique is more ‘gentle’ in the way it applies a charge to molecules, and hence gives rise to fewer fragments. Whilst this makes PTR-MS less able to distinguish between molecules having the same molecular mass, (an issue that can be mitigated using a Time of Flight MS) it does mean that samples don’t need much preparation and importantly allows for real-time analysis of volatile organic compounds (VOCs) in the air.

This is significant for our industry because food aroma is a combination of VOCs that are released into the mouth and nasal cavity when food is consumed. However if the VOC is present within a particular ingredient, or formulation, but then is either not released (perhaps due to interactions with other ingredients), or released before the food even reaches the mouth (perhaps due to inappropriate packaging/storage), then this will have a negative impact on the flavour of the product.

So, it is how those flavour chemicals are released that is important, and until the development of PTR-MS and similar technologies (APCI-MS and SIFT-MS) it was not possible to measure the release of these chemicals in real-time. Sensory panellists could offer their expert opinion of the flavours perceived, but only PTR-MS can make a quantitative measurement of the chemicals responsible for those flavours and therefore give an understanding of why the panellist is perceiving the flavour. And it is not only in the area of flavour that PTR-MS has its uses. As shown below, various research groups also have also used the method in studies on food authenticity and in monitoring plant and animal health.

Applications of PTR-MS to flavour release

 Flavour analysis is clearly an important application area for PTR-MS. Various research groups have reported work where PTR-MS has been used to investigate flavour release in a diverse range of food products. Some of this research has been focussed on improving/modifying formulations, others on optimising process conditions.

Anne Saint-Eve et al[1] used the technique to identify and quantify physical mechanisms responsible for in-nose aroma release during the consumption of mint-flavoured carbonated beverages. They investigated the effects of two composition factors (sugar and CO2) or ‘modulators’ as they can also be termed on both the sensory and physicochemical properties of drinks. They were able to note that carbonated drinks release a greater quantity of aroma compounds in the nose space than non-carbonated ones, whereas sugar content only has an impact (increase) on aroma perception in the case of non-carbonated beverages.

The same principals[2] have also published work on the role of candy texture on the dynamics of aroma release. Four candy textures were established by varying gelatine content between 0 and 15% w/w and they were flavoured with three aroma compounds having different physicochemical properties and different sensory attributes. Using PTR-MS with a trained panel, they were able to identify the sample with the highest flavour release for all aroma compounds, and significantly, to monitor how the flavours were released and persisted over time.

Ground coffee is one of those foods where an appreciation of the aroma of the ingredient before use, and during production of the drink, is just as important to consumers as flavour of the actual beverage. The same might be true of strong cheeses. So it is not surprising that several groups have used PTR-MS to look at these product categories. In the case of coffee, PTR-MS has been used to examine VOC release during roasting, and how this differs between the same beans grown in different geographic regions, and between different beans. It has similarly been used to investigate VOC release during grinding, storage, extraction and consumption, offering potential for using PTR-MS both to understand and manipulate/modify flavour release at every stage, but also to use it for setting quality control parameters and product specifications.

Applications of PTR-MS to authenticity

One group[3] has used PTR-MS to investigate the geographical provenance of coffees. The characterisation of coffees – and many other foods – according to their origins is commercially important of course. In this study, the researchers investigated the aroma profiles of different batches of three monoorigin roasted Coffea arabica coffees (Brazil, Ethiopia and Guatemala). The PTR-MS measurements found significant differences in the volatile profiles of these coffees, and principal component analysis enabled the three coffees to be separated according to geographic origin.

The research group claims that partial least square regression-discriminant analysis (PLSDA) classification also showed completely correct predictions, and that the samples of one batch could be used as a training set to predict geographic origin of the samples of the other batch. This suggests the potential to authenticate the origin of further batches of coffee production by means of the same approach.

Similarly, another group reports using PTR-MS to authenticate the origin of olive oil[4]. Headspace PTR-MS analyses were carried out on 192 olive oils which originated from Cyprus, France, Greece, Italy, and Spain. The mass spectral data were subjected to PLSDA in order to estimate a classification model for the olive oil samples (classification by country of origin) with the research team concluding that cross-validation proved that 84% of the samples were correctly identified by their country of origin.

The group claims that high rates of correct classifications were observed for Greece (80%), Spain (87%), Italy (91%) and Cyprus (100%). Apparently, however, only 10% of the French olive oils were classified correctly. The research team noted that the 10 French samples chosen all originated from the Paca region in the Bouches-du-Rhone province, but were from seven different towns, and these oils have volatile characteristics in common with oils from Greece and Spain. Nonetheless, this study showed PTR-MS to be a promising technique for classification of olive oil samples by their geographic origin.

Conclusion

The above are just some of the examples of published work citing the use of PTR-MS. Other groups have investigated the aroma compounds released by fruits, cheese, milks (especially rancid milk) and meats. Indirectly linked to food, there have been environmental studies around the effects of chemical use (fertiliser, pesticides) on crops, and also on monitoring the breath, and other animal excretions, as an indicator of animal health. Clearly, there is a strong relationship between animal health and meat safety/quality, just as there is a relationship between crop health and ingredient safety/quality.

Where PTR-MS has an advantage over other analytical techniques is that there is very little need for sample preparation and results are available in approximate ‘real-time’. Given that there are MS instruments available that are portable, the technique does not even need to be restricted to the laboratory, although in the case of flavour studies aimed at improving formulations and processes, the laboratory is clearly the right place for any studies to be carried out.

No doubt the technique will continue to develop and more enhancements will be found to broaden its appeal to the food industry. That includes the industries associated with it, such as farming, environment, and even storage/transportation, since failings in these areas are often the root cause of the VOCs that result in taints and unpleasant odours in foods. For that matter, failings in the supply chain are often the root cause of other problems, whether they are associated with quality, safety or authenticity.

As such, any technique that helps us to understand what is right or wrong about our products, whether that be flavour or any other aspect of product composition, is a huge help to the industry.

References

1. Saint-Eve, A., I. Déléris, E. Aubin, E. Sémon, G. Féron, J. M. Rabillier, D. Ibarra, E. Guichard, and I. Souchon. Influence of composition (CO2 and sugar) on aroma release and Applications in Food Science perception of mint-flavored carbonated beverages. Journal of Agricultural and Food Chemistry. 57(13):5891-5898, (2009)
2. DÉLÉRIS I., SAINT EVE A., DAKOWSKI F., SÉMON E., LE QUÉRÉ J.L., SOUCHON I. 2011. The dynamics of aroma release during the consumption of candies with different structures. Relationship with temporal perception. Food Chemistry. 127:1615-1624.
3. Yener S1, Romano A, Cappellin L, Märk TD, Sánchez Del Pulgar J, Gasperi F, Navarini L, Biasioli F. PTR-ToF-MS characterisation of roasted coffees (C. arabica) from different geographic origins. J Mass Spectrom. 2014 Sep;49(9):929-35
4. Nooshin Araghipour, Jennifer Colineau, Alex Koot, Wies Akkermans, Jose Manuel Moreno Rojas, Jonathan Beauchamp, Armin Wisthaler, Tilmann D. Märk, Gerard Downey, Claude Guillou, Luisa Mannina, Saskia van Ruth. Geographical origin classification of olive oils by PTR-MS Food Chemistry Volume 108, Issue 1, 1 May 2008, Pages 374–383

About the authors

 Donna Byrom has recently joined RSSL with 30 years of experience in a variety of analytical techniques on varied matrices. Her experience includes method development, validation and special analysis on tobacco and related products; pesticide analysis on teas and various crops, and on pesticides for regulatory purposes. She was also a Forensic Scientist for 16 years in the field of drugs and toxicology.

Alex Webbe is the Laboratory Manager of the Investigative Analysis Department, which includes the Flavour, Taints and Off Flavours, Lipids and Analytical Chemistry teams, and has been at RSSL for over 4 years. Prior to joining RSSL Alex worked for 11 years in the field of Forensic Science, specialising in DNA and Body Fluid analysis.