Analysis of synthetic endocrine-disrupting chemicals in food: a review.

This work focuses on a revision of analytical methodologies for the determination of industrial chemicals that have an endocrine-disrupting effect on food commodities. These food commodities have been divided into two major categories: crops and food of animal origin. The reviewed methods have been commented on in terms of sample preparation, analytical methods, and the occurrence of the studied compounds.

Determination of chlorothalonil in difficult-to-analyse vegetable matrices using various multiresidue methods.

The molecular characteristics of chlorothalonil can cause particular determination difficulties in some vegetable commodities such as leek or garlic. These difficulties are mainly related to the low recoveries obtained using common multi-residue methods (MRMs)--a consequence of the very high interaction level with natural components in the matrix. These shortcomings were pointed out in the last European Proficiency Test for Pesticide Residues on Fruits and Vegetables, where false negatives for chlorothalonil in leek were observed at around 50%. In this study we have evaluated the ethyl acetate, the Dutch mini-Luke and the QuEChERS MRMs to compare their capabilities for chlorothalonil determination using GC-MS/MS in both the electron impact ionization (EI) and negative chemical ionization (NCI) modes. Best recoveries (in the range of 100-120%, with an RSD below 20%) were obtained using the Dutch mini-Luke method. Lower values (52-70%) were obtained for ethyl acetate whereas no recovery was obtained when the QuEChERS method was applied. Furthermore, tomato matrix was also included in the experiments in order to facilitate the comparability of results. Two ionization modes, electron impact ionization (EI) and negative chemical ionization (NCI) in GC-MS/MS, were applied to evaluate their respective advantages and disadvantages for quantification and identification. As expected, NCI showed limits of detection (LODs) 5 to 10 times lower than EI. However, in both cases, the LODs were still below 10 µg kg(-1). The proposed optimal method was applied for chlorothalonil determination in leek and garlic with good results--in accordance with the European Union (EU) Analytical Quality Control (AQC) Guidelines for pesticides analysis.

Determination of selected pesticides by GC with simultaneous detection by MS (NCI) and µ-ECD in fruit and vegetable matrices.

A gas chromatography-mass spectrometry method in negative chemical ionization mode has been developed incorporating simultaneous detection using a micro-electron capture detector (µ-ECD) for the determination of pesticides in fruits and vegetables. This instrument configuration uses a three-way splitter device which divides the effluent from the analytical column between the two detectors with the split ratio 1:0.1 (MSD/µ-ECD) in each run. The µ-ECD was used for confirmation purposes. Validation of the method was performed on three matrices: tomato, apple, and orange. The ethyl acetate method was assayed; recovery studies were performed at 10 and 100 µg/kg. Recoveries between 70% and 120% were achieved and relative standard deviations lower than 20% (n = 5) were obtained for all pesticides and matrices studied. Limits of quantification lower than 10 µg/kg were obtained for 100% of pesticides in all of the matrices. Limits of quantification lower than 2.5 µg/kg were achieved for 77.8% of pesticides in the tomato and apple matrices, and for 72.2% of pesticides in the orange matrix. The method showed linear response in the concentration range tested (2.5-500 µg/kg) with correlation coefficients >0.99. Good repeatability and reproducibility results were obtained in all cases, with relative standard deviations lower than 16.7% and 20%, respectively. Finally, 20 incurred samples were analyzed using the proposed method. The simultaneous use of the two detectors was satisfactory for the analysis of these real samples. The total number of pesticides identified was 25. The number of samples which contained at least one pesticide was 15-this represented 75% of the total number of samples studied.

Determination of nicotine in mushrooms by various GC/MS- and LC/MS-based methods.

Due to the basic properties of nicotine, it is not easily integrated into commonly used multiresidue methods. The present work investigates the application of two commonly employed multiresidue methods-the QuEChERS method and the ethyl acetate method-for determining nicotine in mushrooms. Both methods are employed in a modified form and an unmodified form: the former to address the special properties of nicotine and the latter, combined with the use of isotopically labelled nicotine, to compensate for poor recoveries. The QuEChERS-based methods were followed by liquid chromatography-time-of-flight mass spectrometry and those based on ethyl acetate extraction were followed by gas chromatography-triple quadrupole-mass spectrometry. All methods were validated according to European guidelines (document no. SANCO/10684/2009). Recovery studies performed on mushroom spiked at 10 and 100 µg kg(-1) yielded average recoveries in the range 80-110% with relative standard deviation (RSD) values below 9%. The linearity of the response over two orders of magnitude was demonstrated (r(2) > 0.995) for all of the determination techniques employed. The limits of detection and quantification obtained were in the 0.7 and 10 µg kg(-1) range, depending on the technique, and thus below the maximum residue level established for this toxic alkaloid by current EU legislation. Good repeatability and reproducibility were obtained in terms of the RSD of the analytical methods (0.4-13.2%). The modified QuEChERS method was tested in a proficiency test on nicotine in dried mushrooms obtaining good results. The methods were successfully applied to 20 real samples.

Method development and validation for determination of thiosultap sodium, thiocyclam, and nereistoxin in pepper matrix.

This work reports a method for extraction and analysis of thiosultap sodium, thiocyclam, and nereistoxin in pepper. Different extraction methods were tested to attain the best recoveries. The final extraction method combines acetonitrile extraction in an acidic medium with ultrasonic extraction followed by a cleanup step with anhydrous MgSO(4). The analyses were performed on a Linear Ion Trap Quadrupole LC-MS/MS in negative mode for thiosultap sodium and in positive mode for thiocyclam and nereistoxin. Recovery studies carried out on peppers spiked at different fortification levels (20 and 200 µg∙kg(-1)) yielded average recoveries in the range 58-87% with RSD (%) values below 20%. Calibration curves covering two orders of magnitude were performed and they were linear over the concentration range studied (0.001-0.5 mg∙l(-1)). Instrumental detection limits were in the low µg∙kg(-1) range. Stability studies of thiosultap sodium in water were performed by evaluating a 100-µg∙l(-1) solution of this compound in water. It was analyzed over 7 days, after which more than 80% degradation of thiosultap sodium could be observed.

Validation and uncertainty study of a comprehensive list of 160 pesticide residues in multi-class vegetables by liquid chromatography-tandem mass spectrometry.

A rapid and sensitive liquid chromatography-tandem mass spectrometry method, in electrospray ionization positive mode, has been developed for the determination of 160 selected multi-class pesticides over a 33-min run time. Extracts were obtained using the acetonitrile-based QuEChERS (quick, easy, cheap, effective, rugged and safe) sample preparation technique. The validation study was carried out on tomato, pear and orange matrices following DG SANCO/2007/3131 of the European Quality Control Guidelines. These matrices represent high water, high sugar and high acidic content commodities, respectively. Matrix influence on recoveries and its effects on ionization were evaluated for the three matrices. Ten out of the 160 pesticides showed very low intensity, linearity and/or sensitivity problems. Linearity was studied in the 5-500 microg kg(-1) concentration range. Soft (<20%), medium (20-50%), and strong (>50%) matrix effects were obtained for 69%, 20%, and 11% of the studied compounds, respectively. Recoveries were investigated at the 10 and 100 microg kg(-1) levels, and depending on the commodity, 97%, 98% and 97% of the compounds in tomato, pear and orange, respectively, were in the 70-120% range. More than 90% of the investigated compounds had less or equal to a 5 microg kg(-1) limit of detection in the studied matrices. The relative standard deviations obtained exceeded 20% in only very few cases. The overall standard deviation obtained in the survey study (0.1551) was used for the method's uncertainty estimation. The expanded uncertainty resulted as being 0.3002 (coverage factor K=2, confidence level 95%). The method was applied on 59 real samples from 14 different kinds of fruits and vegetables. Thirty-three compounds were detected in 50 positive samples.