High Throughput Method for Meat Spoilage Analysis: SPME with CombiPAL Automation

---

BY SUSANNA WONG, JOHN O’REILLY, PHD AND JANUSZ PAWLISZYN, PHD

Food and beverage manufacturers continually strive to extend the shelf life of their products. The storage time is often limited by the release of off-flavours and off-odours generated through degradation of the food as it ages. While production of these unpleasant aromas and flavours is not always a health issue, it can cause consumer rejection of a product, which is a major concern for the food industry. Because of this, there is great demand for analytical methods to accurately determine the freshness and quality of foods.

Food degradation can be observed through changes in texture, colour, smell (production of off-odours) and flavour (production of off-flavours). In the case of meat, lipid peroxidation is the major cause of off-flavours and off-odours in foods. The lipid peroxidation process creates a large number of compounds, including carboxylic acids, ketones, alcohols and aldehydes. Notably, hexanal has been used as a biomarker for lipid peroxidation in the meat industry1,2,3. This article presents a brief description of popular analytical methods in hexanal detection in foodstuffs — mainly meat products — and discusses a protocol that uses automated solid-phase microextraction (SPME) coupled with gas chromatrogaphy-mass spectrometry (GC-MS) that is ideal for high throughput analysis.

The thiobarbituric acid-reacting substances (TBARS) assay is currently the benchmark method in the food industry for determining the hexanal content in various matrices. However, a drawback of TBARS is interference from other non-carbonyl compounds such as lipids, carbohydrates, proteins, pigments and metal ions4. It is also a lengthy test that makes it difficult to meet high throughput requirements.

High performance liquid chromatography (HPLC) and GC methods serve as popular alternatives to the TBARS assay because they allow shorter analysis times, and reduced solvent usage. The volatility of hexanal allows the use of headspace analysis, where the compound is sampled from the gaseous phase above the tested material. An aliquot of this gas can be taken, injected directly into the gas chromatograph for analysis, and the results related back to the amount present in the sample. However, the amount of hexanal in the headspace may be lower than the detection limit of the method.

To solve this problem, SPME can be used. SPME is a solvent-free technique that combines sampling and preconcentration of analytes in a single step5. Commercial SPME devices consist of a fused silica fibre coated with a polymer into which analytes are extracted. The fibre is attached to a plunger and installed in a syringe-like holder. The plunger moves the fibre out of a hollow needle for sampling, and in again for storage. The coating type, on the basis of polarity and volatility, dictates what type of analytes will be extracted. The SPME fibre can be placed directly into the liquid sample or in the headspace for sampling, and can be desorbed directly into the injection port of chromatographic instrumentation. Aldehyde quantification using SPME coupled to GC, has been used in various matrices such as vegetable oils6,7, drinking water8, milk9, alcoholic beverages10, pork2, turkey3 and chicken11.

Existing SPME-GC protocols for hexanal detection have generally involved manual sampling. This is not an ideal situation since it requires regular attention from the user, which is often impractical in industry, where numerous samples are tested daily. However, SPME-GC methods are ideally suited to automation. In our lab, we used a CombiPAL robotic arm system, developed by CTC Analytics AG (Zwingen, Switzerland). This system can automate the extraction, injection and subsequent fibre bakeout processes.

Agitation, which is an important part of SPME analysis to improve the extraction rate of the analytes, can be performed continuously using the PAL without the use of a magnetic stir bar. In manual techniques, sample agitation is commonly accomplished with magnetic stir bars, which can be a nuisance during cleanup. The CombiPAL system has many other add-on accessories, such as a temperature-controlled sample tray, which can simulate refrigeration conditions. The flexibility of this system is ideal for the food and beverage industry for quality control and R&D purposes.

Method Development
Prior to analysing meat samples, the testing protocol was developed using a simplified matrix of saline water spiked with pentanal and hexanal — two aldehydes commonly found in meat samples. Method development using SPME technology involved the following parameters:

fibre coating selection
extraction conditions (time, temperature) and method (headspace, direct sampling, membrane)
injection conditions (time, temperature, depth of injection) and post-injection conditions (bakeout of fibre)
sample preparation
agitation conditions
Methyl pentanal was used as an internal standard due to its similarity to hexanal (in terms of structure and physical properties), the ease with which it separates from pentanal and hexanal by GC, and because it is absent from the analysed meat samples3. Internal standardization was used for all calibration curves and quantification to account for the variance of the testing protocol between each sampling. More importantly, internal standardization allowed correction of the data to account for partial suppression of the analyte signal caused by adsorption of hexanal by the chicken-sample matrix.

Based on the polarity and volatility of hexanal, we tested the 100-micrometre PDMS poly(dimethylsiloxane) or 65-micrometre PDMS/DVB (poly(dimethylsiloxane)-d-vinylbenzene co-polymer) fibre. (The measurements refer to the coating thicknesses.) The 23-gauge needle size was used for both fibres because a thicker needle was more resistant to breakage — an important parameter for the automated agitation process.

During initial experiments, the PDMS/DVB fibre often showed non-linear behaviour in the tested concentration range. This was caused by a displacement effect, a result of the limited number of adsorption sites on this predominantly porous solid fibre coating. Displacement effects are not an issue on the PDMS fibre because it is a liquid polymer and has an absorptive mechanism. Accordingly, the linear range of this fibre (two to 5,000 nanograms per millilitre) was much greater than that of PDMS/DVB. Based on these results, the PDMS fibre was used for all remaining experiments.

Several groups3,11 have employed the use of an antioxidant to arrest the production of hexanal during sampling. We experimented with BHA (2(3)-tert-butyl-4-hydroxyanisole) and ascorbic acid as antioxidants, with little success. The use of BHA caused, on average, a 20% increase in signal for all three aldehydes investigated, while ascorbic acid favoured the release of hexanal into the headspace with methyl-pentanal remaining unaffected. This created an overestimated value of hexanal in the system. As a result of these problems, use of these antioxidants was discontinued.

After numerous parameters were tested, a preliminary protocol was ready for testing on chicken samples. Cooked chicken thighs were minced and placed in a vial and mixed with saline water. Reproducibility with chicken samples was worse than with standard aldehyde solutions, which is not surprising since the composition of the meat is not uniform. Reproducibility was also influenced by carry-over of analytes on the fibre, an effect that could be minimized by decreasing the agitation speed during extraction. It is thought that the carry-over was caused by contamination of the fibre with sample matrix using the more rigorous agitation conditions. Sampling from the chicken matrix also resulted in a more complex chromatogram, which interfered with quantification. This problem was solved by i) altering the column oven temperature programming to avoid co-elution of analytes, and ii) using single-ion extraction MS detection.

Three-day Monitoring of Chicken Degradation
We simulated a continual monitoring of chicken samples over the course of three days. All samples were placed at 30 C to mimic an accelerated degradation study. Six chicken samples were prepared, three of which (named 1a, 1b and 1c) were designated as control samples. The other three vials were special samples used to determine the effects of altering the matrix composition on the hexanal production of degrading chicken. Vial 2 contained large chunks of chicken instead of minced chicken. In vial 3 was minced chicken with ascorbic acid added as an antioxidant, and in vial 4, minced chicken with the atmospheric headspace replaced with nitrogen.

The results of the study are given in Figure 1. The control samples gave an average value of 20 micrograms of hexanal per gram of chicken (denoted µg/g) at the beginning of the run and 110 micrograms per gram at the end of 72 hours. Adding whole pieces of chicken instead of minced chicken gave a consistently lower hexanal value compared to the control. This indicated that the surface area of the chicken plays a significant role in the release of hexanal from the sample. Vial 4 with N2 headspace actually gave a higher level of hexanal production compared to the controls. The result is inconsistent with the hypothesis that based on the mechanism of lipid peroxidation, hexanal production should decrease in a low-O2 environment. The reason for this behaviour is unknown. As mentioned previously, addition of ascorbic acid gave an over-estimation of the amount of hexanal, hence no solid conclusion can be stated from this test.

In summary, we successfully demonstrated an automated method of hexanal quantification to monitor chicken degradation. However, this automated SPME-GC approach could be applied to almost any industry where high throughput sample analysis is essential.

References:
(1) Wilkes, J., E. Conte, Y. Kim, M. Holcomb, J. Sutherland and D. Miller. “Sample Preparation for the Analysis of Flavors and Off-flavors in Foods.” Journal of Chromatography A 880 (2003): 3-33.

(2) Fernando, L., E. Berg and I. Gruen. “Quantitation of Hexanal by Automated SPME for Studying Dietary Influences on the Oxidation of Pork.” Journal of Food Composition and Analysis 16 (2003): 179-88.

(3) Brunton, N., D. Cronin, F. Monahan and R. Durcan. “A Comparison of Solid-phase Microextraction (SPME) Fibres for Measurement of Hexanal and Pentanal in Cooked Turkey.” Food Chemistry 68 (2000): 339-45.

(4) Wong, J. and T. Shibamoto. Chemical Markers for Processed and Stored Foods. Eds. T.C. Lee and H.J. Kim. Washington: American Chemical Society, 1996. No. 631 in ACS Symposium Series.

(5) Pawliszyn, J. Solid Phase Microextraction: Theory and Practice. Toronto: John Wiley & Sons Inc., 1997.

(6) Keszler, A. and K. Heberger. “Influence of Extraction Parameters and Medium on Efficiency of Solid-phase Microextraction Sampling in Analysis of Aliphatic Aldehydes.” Journal of Chromatography A 845 (1999): 337-47.

(7) Doleschall, F., Z. Kemeny, K. Recseg and K. Kovari. “Monitoring of Lipid Degradation Products by Solid-phase Microextraction.” Journal of Microcolumn Separations 13.6 (2001): 215-20.

(8) Cancho, B., F. Ventura and T. Galceran. “Determination of Aldehydes in Drinking Water Using Pentafluorobenzylhydroxylamine Derivatization and Solid-phase Microextraction.” Journal of Chromatography A 943 (2001): 1-13.

(9) van Aardt, M., S. Duncan, D. Bourne, J. Marcy, T. Long, C. Hackney and C. Heisey. “Flavor Threshold for Acetaldehyde in Milk, Chocolate Milk, and Spring Water Using Solid Phase Microextraction Gas Chromatography for Quantification.” Journal of Agricultural and Food Chemistry 49 (2001): 1377-81.

(10) Wardencki, W., P. Sowinski and J. Curylo. “Evaluation of Headspace Solid-phase Microextraction for the Analysis of Volatile Carbonyl Compounds in Spirits and Alcoholic Beverages.” Journal of Chromatography A 984 (2003): 89-96.

(11) Goodridge, C., R. Beaudry, J. Pestka and D. Smith. “Solid Phase Microextraction-Gas Chromatography for Quantifying Headspace Hexanal Above Freeze-dried Chicken Myofibrils.” Journal of Agricultural and Food Chemistry 51 (2003): 4185-90.

For more information about the investigators, please visit www.science.uwaterloo.ca/chemistry/pawliszyn.