Analyzing Volatile Organic Chemicals in Food: Emerging Trends and Recent Examples

The detection of volatile organic compounds (VOCs) is important for a variety of reasons, many of which stem either from the need to ensure that damaging chemicals are controlled or eliminated, or the desire to understand more about the chemicals that affect our enjoyment of the world around us. These applications range from environmental monitoring to material emissions testing, but one area in particular has recently seen large advances—the analysis of VOCs in food.

Analysis of food has long been important in order to evaluate its composition, monitor changes upon processing or cooking, and ascertain the nature of components that provide them with desirable (or undesirable) characters. Volatiles are just one aspect of the plethora of chemicals present in foods, but they have long been particularly challenging for the analyst. However, technological developments are now addressing some key issues in this area, and this article aims to highlight some of these trends by selecting a few recent examples from the extensive literature on the subject.

Eliminating complex sample preparation

Analyzing VOCs in food was once limited to the characterization of all sufficiently small organic molecules by methods such as solvent extraction followed by LC-MS. However, these methods tended to be laborious, and thus there has been a drive to develop sample preparation techniques that are less demanding of the analyst’s time.

One such method, headspace-solid-phase microextraction (HS-SPME), has become very popular over the last 10 years for its ease of use, and has been applied to the aroma profiling of a wide range of foods.1

The application of HS-SPME is exemplified by a comparative analysis of roasted pistachio aroma carried out by Montserrat Mestres and colleagues at Rovira i Virgili University (Tarragona, Spain).2 The authors concluded that HS-SPME was a good alternative to solvent extraction for obtaining representative profiles of pistachio aroma, having both better repeatability and requiring less sample handling. A downside was that HS-SPME was less efficient at extracting heavier compounds, although in this case it did not affect the assessment of the most odor-active compounds.

Despite the clear benefits of using HS-SPME, the small sample size can result in insufficient sensitivity for trace-level components, especially those toward either end of the volatility range. Such issues can be addressed by taking a larger sample and by using sorbent tubes, which can handle a wider range of analytes. An example of this is shown in Figure 1, where researchers at ALMSCO International (Llantrisant, U.K.) studying VOCs from potato chips used a microchamber to dynamically collect headspace onto a thermal desorption (TD) sorbent tube.3 Subsequent preconcentration and analysis by TD-GC-MS allowed identification of trace-level pyrazines important for flavor.

Figure 1 – Analysis of potato chip headspace using a microchamber sampling method and TD-GC-MS. Three pyrazines (2,5-dimethylpyrazine [T1], 2-ethyl-6-methylpyrazine [T2], and 2,6-diethylpyrazine [T3]) were identified, despite their presence at trace levels.3

Speeding up analysis

As mentioned above, part of the appeal of HS-SPME is its simplicity compared to solvent extraction methods, but when dealing with large numbers of samples, even this approach can be too time-consuming. Therefore, in certain fields (such as product screening), the development of rapid automated analytical methods has become a focus of interest.

This is highlighted by a study by Claudine Cognat and colleagues at The James Hutton Institute in Dundee, U.K., who investigated sampling methods for the analysis of off-flavor VOCs from oatcakes.4 Although they found SPME to be relatively inexpensive, they felt that sample preparation and collection was time consuming, limiting sample throughput. In contrast, sampling direct onto sorbent tubes was rapid, and subsequent analysis by TD-GC-MS “improved sample throughput in comparison with...the SPME method.” They found that it was possible to discriminate between fresh and rancid samples by considering changes in the relative abundances of the constituents (Figure 2).

Figure 2 – Comparison of the VOC profiles of fresh and rancid oatcakes analyzed by TD-GC-MS. (Reprinted from Food Chemistry4 [copyright 2012], with permission from Elsevier.)

Clearly, this issue of food safety is one in which rapid analysis and transportable equipment are paramount, and this is underscored by the launch of a research project to improve the safety and quality of ready-to-eat produce. The €4M ‘QUAFETY’ initiative, which is co-funded by the European Commission, is a collaboration between 14 organizations to develop predictive models that will allow VOCs indicative of microbial contamination to be rapidly identified.5 One of the researchers involved, Carsten Muller at Cardiff University, U.K., said, “At the moment...people working in the industry estimate shelf-life simply by judging the appearance of the product.” To address this, the researchers will use rapid sampling techniques in conjunction with TD-GC-TOF (time-of-flight) MS to identify changes in the VOCs emitted by ready-to-eat foods as they age, starting with melons, rocket, and fruit salads. One of the reasons for the choice of analytical method was the ability to detect trace-level components. A colleague at Cardiff University also involved in the project, Hilary Rogers, pointed out that “in terms of biological sensitivity and importance, minor compounds may be much more relevant.” It is hoped that the ultimate outcome of this project will be a set of objective criteria and an easy-to-use method for rapidly determining the safety and quality of fruits and vegetables.

Expanding analytical scope

The large number of compounds released from food and the need to separate and identify them presents a particular challenge for the analyst, and one that demands techniques with the ability to identify the widest possible range of compounds from a single sample. In this respect, sulfur compounds appear often in the literature, primarily because of their disproportionately strong odors, even at trace levels, and their propensity to degrade during analysis. They are often of key importance in food aromas, although not only as indicators of decay, since they can be contributors to the distinctive aromas of certain foods.

One such example is provided by a study conducted by Alain Chaintreau and co-workers from Firmenich SA (Geneva, Switzerland),6 who were able to identify with confidence 50 sulfur compounds in beef aroma, with seven being identified for the first time from this substrate. The method they chose was GC×GC-TOF-MS, which the authors said, “simultaneously offers an unrivalled peak resolution and a much greater sensitivity” than conventional GC-MS. The result, they concluded, is the ability to detect thiols in a nonderivatized form, which, when combined with olfactometry, gives a more representative picture of the aroma profile.

This is just one of the many examples applying advanced chromatographic techniques to identify and quantify previously inseparable compounds. Another example is enantioselective gas chromatography, which was used in a study led by Luigi Mondello from Messina University, Italy, to measure the enantiomeric distribution of compounds in strawberry-flavored foods.7 They found that whereas γ-decalactone in strawberries comprised 90–95% of the (R) enantiomer, four strawberry-flavored food products (two yogurts, candies, and ice pops) had values in the range 47–66%. The findings agreed with those determined using combustion isotope-ratio mass spectrometry, allowing them to infer the presence of synthetic strawberry flavoring.

Enhancing analytical scope can also be addressed with different sampling techniques. For example, Mitsuya Shimoda and colleagues at Kyushu University (Fufuoka, Japan) developed a low-density polyethylene pouch method to extract volatile flavor compounds from oils8—a type of substrate that has long presented difficulties for the analyst because lipid in the sample interferes with the extraction of VOCs. The authors concluded the method showed excellent performance for the selective extraction of volatile flavor compounds from butter oil, with recoveries of ca. 1.6% for C4–C16 fatty acids and ca. 11% for C10–C16 lactones—important for buttery/oily/waxy and fruity/creamy/buttery odors, respectively. In particular, they noted that δ-tetradecalactone and δ-hexadecalactone were extracted in high concentration compared to short-path distillation, and that the aroma character of the extract matched the odor character of butter oil, “indicating the promise of this extraction technique.”

Improving data analysis

The availability of large libraries of spectra and more powerful software to search them has over recent years transformed the process of compound identification. As recently as 2007, Sílvia Rocha and co-workers from Aviero University, Portugal, noted that in analyses of grape monoterpenoids, “deep analyses of the chromatograms frequently indicate that some peaks are the result of two or more co-eluting compounds...[meaning that]...reliable MS identification is not possible.”9

Although monoterpenoids undoubtedly remain a challenging target due to their spectral similarity, sophisticated algorithms are now available that greatly improve prospects for obtaining good spectral matches from coeluting components. When this is done, the need for more complex separation methods can be eliminated, making data-mining of this type a valuable first step when processing highly complex GC-MS datasets. In addition, if matching is fully automated, the analyst can be spared a considerable amount of tedious manual processing. Figure 3 shows the results of using such software in the analysis of cheese headspace, where almost 70 compounds were identified with confidence.10

Figure 3 – The application of target-matching software greatly simplifies the identification of VOCs released from cheese. In this case, the headspace from extra-mature cheddar was sampled using a microchamber apparatus, with analysis by TD-GC-MS.10

Another software development that can help food analysts is the application of algorithms that can carry out in-depth analysis of GC-MS data to identify peak profiles of individual components, allowing their separation both from each other and from the background. This results in much cleaner mass spectra, greatly improving the prospects for subsequent target matching.

The fact that such high-resolution processing is possible at all stems largely from the use of mass spectrometers that are capable of acquiring data at much higher rates, such as time-of-flight instruments. Increasingly, moving to new technologies such as these can have additional benefits, as found by Gerhard Horner and colleagues from ALMSCO International. By replacing a quadrupole mass spectrometer with a time-of-flight instrument, they were able to achieve a 30–50-fold increase in sensitivity to off-flavor components of orange juice, such as 3-methylbutanal (Figure 4).11 This and several other compounds were in this case thought to result from migration of oxygen through the packaging material during storage.

Figure 4 – Comparison of GC-quadrupole MS and GC-TOF-MS analyses for a sample of orange juice following storage in its original packaging for one year, and sampling using purge-and-trap. The improved sensitivity of TOF allowed identification of a number of low-concentration components.11


It is hoped that this short overview has shown that technologies for the analysis of VOCs in food have advanced substantially in recent years, with a range of new technologies and methods that allow more detailed information to be obtained. Discoveries resulting from these advances are certain to have a profound effect on food science, with implications for aroma profiling and product development, as well as studies of food safety, shelf-life, and packaging materials.


  1. Heaven, M.W.; Nash, D. Recent analyses using solid phase microextraction in industries related to food made into or from liquids. Food Control  2012, 27, 214–27.
  2. Aceña, L.; Vera, L. et al. Comparative study of two extraction techniques to obtain representative aroma extracts for being analysed by gas chromatography–olfactometry: application to roasted pistachio aroma. J. Chromatogr.A  2010, 1217, 7781–7.
  3. The application of TargetView software in the food industry—the identification of pyrazines in potato crisps. Application Note 002, ALMSCO International, April 2010.
  4. Cognat, C.; Shepherd, T. et al. Comparison of two headspace sampling techniques for the analysis of off-flavour volatiles from oat based products. Food Chem. 2012, 134, 1592–1600.
  5. Food safety and quality all wrapped up. www.americanlaboratory. com/617-News/121439-Food-Safety-and-Quality-All-Wrapped- Up/?catid=1414.
  6. Rochat, S.; de Saint Laumer, J.-Y. Analysis of sulfur compounds from the in-oven roast beef aroma by comprehensive twodimensional gas chromatography. J. Chromatogr. A   2007, 1147, 85–94.
  7. Schipilliti, L.; Dugo, P. et al. Headspace-solid phase microextraction coupled to gas chromatography-combustion-isotope ratio mass spectrometer and to enantioselective gas chromatography for strawberry flavoured food quality control. J. Chromatogr.A  2011, 1218, 7481–6.
  8. Chongcharoenyanon, B.; Yamashita, N. et al. Extraction of volatile flavour compounds from butter oil in a low-density polyethylene membrane pouch. Flavour Fragrance J. 2012, 27, 367–71.
  9. Rocha, S.M.; Coelho, E. et al. Comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry of monoterpenoids as a powerful tool for grape origin traceability. J. Chromatogr. A  2007, 1161, 292–9.
  10. Rapid aroma profiling of cheese using a Micro-Chamber/Thermal Extractor with TD-GC/MS analysis. Application Note TDTS 101, Markes International, May 2012.
  11. Gerhards, P.; Möller, M. et al. Development of a sensitive and reliable method for the measurement of volatile organic compound migration from food packaging—a comparison of GC/TOF MS vs. GC/quadrupole MS; 59th ASMS Conference on Mass Spectrometry, Denver, CO, June 2011.

Dr. David Barden is a Technical Copywriter, Markes International, Inc., 11126-D Kenwood Rd., Cincinnati, OH 45242, U.S.A.; tel.: 866-483-5684; e-mail: