Skip to main content
European Commission logo print header

Contaminant analysis in Food by Liquid Chromatography Mass Spectrometry: towards a generic extraction and screening for pesticides and veterinary drugs

Final Report Summary - CONFIRMS (Contaminant analysis in Food by Liquid Chromatography Mass Spectrometry: towards a generic extraction and screening for pesticides and veterinary drugs)

The use of pesticides and veterinary drugs has changed through time, going from more persistent to more polar, readily degradable products. Many multiresidue methods, for a wide variety of existing active compounds, are currently used in official routine laboratories to control their hazard residues in food. However, highly polar analytes cannot be usually included in these methods due to their different physico-chemical characteristics. Besides, the existing methods only collect data for targeted compounds, so other potentially harmful, non-targeted analytes are not detected although present in the sample. Therefore, the project’s general main goal was the development of generic extraction and analysis procedures for a wide range of target (highly polar) and non-target contaminants in food samples using liquid chromatography (LC) coupled to different complementary mass spectrometry detectors (MS) expected to address some of the challenges that food residue control faces nowadays.
14 aminoglycosides (AGs), an important family of highly polar antibiotics, were selected for target analysis. They were not retained in typical reversed-phase LC, so an alternative analysis with hydrophilic interaction liquid chromatography (HILIC) was tested. Five stationary phases with different mobile phases were evaluated at 40°C: four HILIC columns (ZIC®-HILIC, BEH Amide, Aminopropyl and Triazole) compared to ObeliscTM R with Liquid Separation Cell (LiSCTM) technology. Only zwitterionic phases (ObeliscTM R and ZIC®-HILIC) were able to achieve a proper chromatographic separation necessary due to crosstalk effect in low resolution MS. These columns present quaternary amine-like positive cations near the silica surface separated by a hydrophobic chain from head negative groups. ZIC®-HILIC needed highly concentrated ammonium buffer mobile phases for elution because the sulfonic groups are permanently charged. Therefore only ammonium counter-ions could compete with the retained AGs. on the other hand, response from the MS was significantly interfered because ESI could not use effluents with relatively high salt concentrations due to signal instability, poor peak symmetry and unacceptable low sensitivity. These problems were solved by the use of ObeliscTM R with a water/acetonitrile mobile phase with 1% formic acid. The negative ionic head groups were the carboxyl type and the low pH provided by the formic acid gave enough counter-ions to reduce the carboxyl ionization state suppressing ion-exchange interactions with AG amino groups for elution. The fellow visited the laboratory of the Public Health Agency of Barcelona (ASPB) in Dec. 2013 where she could confirm her results and acquired hands-on experience in the extraction of AG from complicated food samples. After, she adapted the extraction procedure for honey and kidney with solid phase extraction (SPE) silica-based weak cation exchange cartridges (WCX) developed at ASPB for liver, milk and honey with polymeric cartridges. The percentage of trichloroacetic acid was increased to 5% and a faster centrifugation in cold conditions was carried out in order to remove all proteins from milk and liver previously to SPE. Different brands of polymeric WCX cartridges were tested (Strata X-CW by Phenomenex, Oasis WCX by Waters, Taurus WCX by BGB Analytik) being the last one the best to provide high recoveries with a buffer elution at pH 3. The optimized final method was validated for honey, milk and liver samples according to Decision 2002/657/EC.
In addition to some of the previously selected highly polar pesticides (Chlormequat, Diquat, Difenzoquat, Mepiquat, Paraquat, Daminozide), Amitrole, Cyromazine, ETU (ethylene thiourea) and PTU (propylene thiourea) were also included in the project. They were separated and eluted from ObeliscTM R with a mixture of 50 mM ammonium formate at pH 2.5 and acetonitrile. A direct injection after a rapid sample pre-treatment with methanol 1% formic acid had been proposed for these analytes in fruits and vegetables. However, due to the low sensitivity of the MS detector available, a dispersive solid-phase extraction (d-SPE) based on QuEChERS procedure was further mandatory previous to injection. One mL of extract was mixed with different sorbents (PSA, C18 and GCB) but only Taurus WCX together with PSA and eluted with acetonitrile 3% formic acid improved analyte signal, mainly for paraquat and diquat.
Due to the extensive findings of quaternary ammonium compounds (QACs) residues on foodstuffs arriving to the EU during 2012 and considering their use as pesticide forbidden, they were also included in the project although it was not originally intended. DDAC-C10 and four BAC homologues (C10, C12, C14 and C16) were selected due to their predominance in the disinfection products in food processing, their toxicity and the current legislation. A C18 stationary phase with a water/methanol mobile phase containing 0.1% formic acid was used for QACs analysis by UPLC-MS/MS. QACs were extracted from plants with the buffered QuEChERS version (AOAC 2007.01) without clean-up because QACs were retained in the PSA sorbent. An alternative validation based on Document SANCO No. 12571/2013 was carried out in representative’s matrices, namely orange, lettuce and avocado. Five quantification methods were compared to compensate matrix effects being the two standard addition methods (SAM) the most accurate. Due to a non-linear relationship between the signal and the analyte concentration in the studied range, non-linear SAM extrapolations were used when the calibration dependence was of convex and slightly concave curvature while for concave curves a linear fitting was applied.
Regarding the non and post-target analysis objectives, the fellow made a stay in May-June 2013 at the “Laboratory of Pesticide Residue Analysis” at the University Jaume I of Castellon (Spain). She completed her training in the non-target analysis of water samples with very modern technology, UHPLC-APGC-QTOF/MS, and the post-target processing of samples. This technique allowed searching for a wide range of both expected and unexpected contaminants (pesticide, veterinary medicines, drugs of abuse…) with home-made data bases and improved by the high mass resolution and mass accuracy provided by the QTOF detector.
The overall work of this project was carried out at the “Laboratory of Pharmaceutical Analytical Chemistry” (PAC) at the University of Geneva. As part of collaboration, all the developed methods for highly polar analytes were transferred to the “Laboratory of Contaminants” at the Food Authority Control (SCAV) of Geneva to complement their own procedures. As a result, several campaigns to control food products in the canton of Geneva were developed: QACs in fruit and vegetable samples in spring 2013, AGs in honey, milk and liver samples in spring 2014 and highly polar pesticides in fruits and vegetables planned for autumn 2014. The fellow also helped completing the existing databases at the “Laboratory of Contaminants” at SCAV with the highly polar analytes included in the project. These screening techniques are currently applied at SCAV to food market samples in a post-target approach.
The obtained results for highly polar analytes were presented in three international conferences, with one poster in 6th RAFA 2013 and two oral communications in 7th HVDRA 2014 and 38th ISEAC 2014 respectively. The fellow also made an oral presentation in French for the Geneva's food authorities (27 May 2014) entitled “A safer food after the determination of highly polar veterinary drugs and pesticides” for all the staff from the General Direction of Health of the Canton of Geneva. The aim was to show the utility of the developed methods to control problematic highly polar analytes to official chemists, veterinarians and legislators.
In conclusion, the project’s general main goal was achieved with the development of three procedures for highly polar analytes that were directly implemented in an official routine laboratory in target and non-target approaches.