Confirmation of the full conversion of ethylene oxide to 2-chloroethanol in fumigated foodstuffs: possible implications for risk assessment

Abstract

This study describes the extension of a gas chromatography mass spectrometry (GC-MS) method, initially devoted to the analysis of ethylene oxide (EO) in ice cream, to a larger range of food items including herbs, spices, vegetables, inorganic salts, food supplements, thickeners, etc. Results are reported as EO_Total according to EC 2015/868 definition (expressed as EO equivalents as the sum of native EO and 2-chloroethanol (2-CE) after acidic hydrolysis) with a limit of quantification at 0.01 mg/kg regardless of the food item. Its ruggedness was demonstrated through fortification experiments on hundreds of samples. Re-analysis of 146 positive food samples without hydrolysis demonstrated that not EO but 2-CE is the predominant analyte detected in the different processed ingredients suspected to have been previously treated with EO. A series of eight contaminated dried herbs and spices were also re-analysed by four ISO 17025 accredited commercial laboratories making use of different analytical strategies for EO determination in foods. Each laboratory reported EO_Total levels within the same concentration range, but the resulting reproducibility ranged from 23% to 41% depending on the sample. Additionally, we show that results of free EO from methods based on conversion to 2-iodoethanol may lead to artefactual detection of native EO (false positive). An official method of analysis applicable for different food matrices would be useful to avoid discrepancies of results. Altogether, these data re-enforce the fact that in absence of native EO in food items, risk assessment of EO in foodstuffs should consider the predominance of 2-CE. A toxicological risk assessment using the food additive xanthan gum as a case study is discussed.

Keywords:

• Ethylene oxide
• 2-chloroethanol
• foodstuffs
• GC-MS
• risk assessment

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Acknowledgments

Authors would like to warmly thank Julien Alberteau and Remi Da Silva for their contribution to the method development and to the collection of data.

Disclosure statement

No potential conflict of interest was reported by the author(s).