The presence of compounds causing taints and off-flavours in food is a major concern to the food industry. However, the identification of these chemicals in foods presents an analytical challenge due to the complexity of the matrix, the need for sensitive methodologies and the commercial pressure for rapid results.
by Dr Kathy Ridgway
Taints and off-flavours in food can originate from many sources in all parts of the global supply chain [1]. Taints can be defined as contamination from an external source, for example during processing or storage. Off-flavours are defined as an atypical odour or taste resulting from internal deterioration of the food, such as lipid oxidation or microbial spoilage.
Identification of compounds causing a taint or off-flavour in food is critical to the food industry, to address consumer concerns as well as to prevent future quality incidents. As the levels of compounds responsible for taints are frequently very low, they rarely pose a health risk to the consumer. However, the presence of a taint or off-flavour in food leads to a reduction in consumer confidence, perception of poor quality and ultimately brand damage, which can be extremely costly for a food manufacturer. Therefore it is imperative that a robust analytical approach is followed to reliably identify the taint, which then enables root cause analysis to be carried out and appropriate risk reduction measures to be taken.
Analytical approach
The sampling procedures that are employed are important, as the chemical causing the taint may not be evenly distributed throughout a product or ingredient. Care must be taken at all stages to avoid any possible contamination from external sources during transportation or once within the laboratory.
Reliable sensory descriptors can often be the key to enabling more targeted chemical analysis and the descriptors provided by an experienced panel are preferable to reliance on a consumer’s odour or taste description. Hence in general two approaches might be taken. In the cases where it is possible to predict which compound is causing the taint from information provided or from sensory descriptors, the analytical method can be targeted from the outset. Where the cause of the taint is unknown, a more universal extraction method is required to try and identify differences in the volatile profiles of the tainted sample compared with a control/reference sample. An experienced analyst can then identify compounds based on mass spectral library searches (such as National institute of Standards and Technology, NIST). Where possible, analytical standards should be used for confirmation of chromatographic retention time, mass spectra and, if necessary, sensory descriptors for the taint in the matrix.
Methods of chemical analysis
Most compounds responsible for taints are volatile, so gas chromatography (GC) is the technique most frequently used for separating these compounds from other food components, generally with a mass spectrometric detector (GC-MS). Extraction methods tend to be based on solvent extraction, distillation, headspace procedures and more recently, sorptive extraction techniques such as solid phase microextraction (SPME) and stir bar sorptive extraction (SBSE).
Solvent extraction
Direct solvent extraction can be used for extraction of tainting compounds from food. However, solvent extraction methods often require subsequent concentration and clean-up procedures, potentially resulting in loss of analytes. Other solvent extraction techniques such as soxhlet, microwave and pressurised fluid extraction, as well as supercritical fluid extraction, have been reported for specific tainting compounds. In general, solvent extraction is only used for targeted analysis of specific tainting compounds where selective clean-up and sample enrichment procedures, such as solid phase extraction, can be utilised.
Steam distillation (Likens Nickerson) extraction
Steam distillation can be used to extract volatile tainting compounds from the non-volatile food components. Distillation can be performed under vacuum to enable lower temperatures to be employed for thermally labile analytes. Microwave-assisted steam distillation has also been reported for the extraction of certain taints.
Combined steam distillation and solvent extraction allows for a more exhaustive continual process and is widely used for isolation of volatile compounds. Variations to the original apparatus have been made [Figure 1], as first described by Likens and Nickerson [2], but the principle remains the same.
The sample is co-distilled between two solvents. Generally water containing the sample is boiled in one flask, while the extracting solvent (such as diethyl ether) is boiled in another. The mixing of the vapours enables extraction of the volatile compounds into the co-solvent. Relatively small volumes of solvents can be used, and the technique produces a relatively “clean” extract as the non-volatile food components remain in the sample flask. However, a concentration step is often still required for trace analysis, which can result in analyte losses. The main disadvantage of this technique is the need for specialist glassware, and the long extraction time, which is typically over an hour.
Direct static headspace
This is a very useful technique for analysis of volatile compounds as it requires little or no sample preparation. It can be used as a first step to detect differences between control/reference and tainted samples. As only the compounds in the vapour phase (gas) above the sample are analysed, it can be applied to most matrices without interference from the non-volatile components. However, the matrix will influence the partitioning of the analyte and since it is an equilibrium-based technique, accurate quantitation may require matrix-matched standards, the method of standard additions, or the use of an internal standard (preferably isotopically labelled). The main limitations of the technique are that it only detects the most intense taints; the compound responsible may have to be present at a relatively high level to enable detection and identification using mass spectral scan acquisition. More targeted analysis or selective headspace sampling that includes a concentration step (such as headspace solid phase microextraction, HS-SPME) can be used to help overcome this problem. Another approach is to use a dynamic headspace technique, such as purge and trap.
Dynamic headspace
Dynamic headspace enables concentration of the sample headspace and therefore offers a potential increase in sensitivity. Volatile compounds can be trapped onto sorbents such as tenax. However, traditional purge and trap devices can have problems with carry-over. Moreover, the limited selectivity means that matrix components are also concentrated alongside analytes, resulting in only small increases in signal-to-noise and therefore sensitivity, compared to direct headspace for most applications.
Solid phase microextraction
SPME involves a coated fibre, and can provide selective sampling and concentration in one step, resulting in potential increases in sensitivity for volatile compounds. The technique is most frequently used to sample the headspace above the sample matrix, although it can also be used for direct sampling of liquids or solutions. In headspace sampling mode, the method is used extensively to obtain volatile profiles of food products and ingredients. More recently HS-SPME has been used for quantitative analysis of volatiles and specific tainting compounds, such as chloroanisoles and chlorophenols. The SPME fibre coating (extracting phase) can be chosen to determine the selectivity of extraction, and as some compounds are not extracted, background noise is generally less than with direct headspace techniques. As with direct headspace techniques, the matrix must be considered and standard additions or the use of an internal standard are recommended. The sample matrix can be modified by pH adjustment or the addition of salt to increase extraction efficiency for specific compounds, and derivatisation can be used either in the matrix solution or on-fibre. The requirement to optimise extraction for each matrix limits the use of SPME as a screening method for determination of unknown taints. However, it can provide low detection limits and has been used for a range of tainting compounds [3], providing superior sensitivity compared with direct headspace analysis.
Stir bar sorptive extraction (SBSE)
Similarly to SPME, SBSE enables extraction and concentration in one step.
The extracting phase is coated onto a glass stir bar and analytes are then desorbed either thermally directly into a GC, or by solvent for subsequent analysis by GC or LC (Liquid Chromatography). Sampling can be performed directly in liquid samples or solutions; alternatively, the stir bar can be used to sample the headspace (headspace sorptive extraction HSSE). SBSE has been reported for targeted analysis of several tainting compounds and more recently as a screening procedure [4]. Using SBSE and GC-MS in scan acquisition, low limits of detection were achieved for most compounds as illustrated for the determination of 2,6-dichlorophenol in a soft drink sample [Figure2].
Future developments
Developments in sorptive extraction (SPME and SBSE) have resulted in more rapid procedures and enabled the low sensitivity required for some tainting compounds. However, for true unknowns, where selective sample preparation cannot be used, there is still a need for a rapid, sensitive, universal extraction procedure. Future developments in software and instrumentation are also likely to allow better profiling of food volatiles, taints and off-flavours. Software pattern recognition and background subtraction tools will enable more rapid comparison of control/reference and tainted samples. Instrumentation such as hyphenated chromatography (GC x GC) and time of flight mass spectrometers that provide accurate mass can also be powerful techniques for identification of unknowns.
Conclusions
The prevention of taints and off-flavours in foods is paramount to ensure food quality and potentially safety for the consumer. Analytical methods to identify causative compounds therefore need to be robust and accurate to enable root cause investigations and prevent future occurrences. The choice of analytical methods used depend on many factors and a flexible approach needs to be taken. An understanding of the limitations of current analytical methods and expertise in data interpretation are crucial for providing the rapid response required by the food industry. Moreover, further investigations and analysis to identify root cause and prevent re-occurrence can prove invaluable.
References
1. Ridgway et al. Food additives and contaminants 2010; 27: 146-148.
2. Likens ST, Nickerson GB. Am Soc Brew Chem proc 1964; 5: 13-34
3. Boutou S, Chatonnet P. J Chromatogr A 2007; 1141: 1-9
4. Ridgway K et al. Analytica Chimica Acta 2010; in press
The author
Dr Kathy Ridgway
Technical Specialist, Flavour
Investigative Analysis
Reading Scientific Services Ltd.
Tel +44 118 986 8541
e-mail: kathy.ridgway@rssl.com
www.rssl.com
