Headspace conditions and ingredients can affect artefactual benzene formation in beverages.

A headspace sampling-gas chromatography/mass spectrometry (HS-GC/MS) method using mild HS conditions (40 °C, 30 min) was established, validated in terms of specificity, linearity (1.75-87.65 ng mL-1), precision (0.3-9.1% RSD), and accuracy (81.1-117.7%); and applied for the monitoring of 900 commercial beverage samples of six different types. These mild (low-temperature) conditions were compared with 1) optimized (high-temperature) conditions and 2) a liquid-phase microextraction method involving no heat treatment. This method was desirable because a high equilibrium temperature induced artefactual benzene formation from benzoate and ascorbic acid. In a 2IV8-3 fractional factorial design, eight variables-ascorbic acid, benzoate, benzaldehyde, Cu2+, Fe2+, riboflavin, pyridoxine, and heat treatment-were tested as potential factors affecting benzene formation. All variables except Fe2+ and pyridoxine significantly affected benzene formation, both individually and interactively. The present study suggests an accurate and reliable method for benzene analysis and provides strategies to prevent unintentional benzene formation in beverages.