Research On Multi-class, Multi-residue Analytical Method For Screening Of Low-weight Organic Contaminants In Food Of Animal Origin

In this work, a series of comprehensive screening methods for multi-class small molecule organic contaminants in food of animal origin (e.g. raw milk, infant formula, pork, bovine, fish, chicken and pork liver) based on ultra performance liquid chromatography coupled with electrospray ionization tandem mass spectrometry (UPLC-ESI-MS/MS) were developed. The study consists of the following three chapters.In the first chapter, the review covering research of published multi-class methods between2005and2013provides an update overview of recent trends in sample preparation and analysis for the determination of trace residues and contaminants in food. In the last decade food safety of animal origin is still challenging due to numerous low-weight organic contaminants. There are increasing interests in method development for the simultaneous and rapid analysis of various classes of veterinary drug residues and other contaminants in food of animal origin with a wide diversity of polarities and pKa values. Recent development of mass spectrometry enables the possibility to simultaneously analyze many different compounds with very few method developments. Due to the high through-put and high sensitivity, liquid chromatography/mass spectrometry (LC-MS) has been playing a vital role in determination of multi-class/multi-residue trace residues and contaminants in food. The future perspectives and potential directions are also outlined. However, for food of animal origin with complex matrix, sample preparation is still the bottleneck, mainly in terms of purity, analysis time, and sources of error. The future perspectives and potential directions are also outlined.In the second chapter, universal, rapid and simple analytical methods which were able to identify veterinary drug residues and other contaminants in raw milk and infant formula were developed. For quantification, matrix-fortified calibration curves were performed to compensate the matrix effect and loss in sample preparation. For most of the target analytes, there is no significant interference on analysis with two sample matrix by using the optimized pretreatment processes. The limit of quantification (LOQs) varied from0.05to10μg/kg. Average recovery rates of the spiked standards in raw milk were in the range from63%to141%with associated RSD values from1%to29%under the selected conditions. Statistical evaluation revealed that average recoveries of the spiked standards in infant formula were in the range from51.7to157.1%. The reproducibility and inter-day precision were in range of1.1-28.8%and2.4-26.2%, respectively. For over80%of analytes in these two matrices, the average recoveries were from70to120%with the corresponding RSD below20%, which were acceptable and in agreement with the criteria of Commission Decision2002/657/EC.In the third chapter, all analytes were divided into three groups according to their polarities. The first group of analytes with very strong polarity (12aminoglycosides) were extracted and purified by binary phases (n-hexane and acetonitrile-ethanol aqueous). The competent linearity was found for all of target compounds with linear regression coefficients (R) higher than0.99; and the limit of quantification (LOQs) varied from5to40μg/kg. Average recoveries of the analytes spiked into pork were in the range from59.1%to106.4%with associated RSD values from3.3%to21.0%under the selected conditions. The intra-day precision was in the range of3.8%-24.7%.For the second group of analytes with polar to medium polarity, the pre-treatment was based on the extraction with binary phases (n-hexane and acetonitrile-ethanol aqueous) followed by the purification with low temperature clean-up and D-SPE (modified QuEChERS type). The limit of quantification (LOQs) was from0.05to10μg/kg. Average recoveries when spiked into bovine muscle, fish and chicken were 57.8%-158.5%,69.0%-143.0%and69.0%-136.9%, respectively. The relative standard deviations (RSDs) were1.0%-29.6%,1.8%-33.7%and1.2%-29.8%, respectively. The intra-day precisions were in the range of0.9%-29.4%,3.2%-29.5%and3.2%-27.9%, respectively.The third group of analytes with the very weak polarity were extracted and purified by two steps of D-SPE and low temperature clean-up. The limit of quantification (LOQs) in chicken and pork liver ranged0.05-1.0μg/kg and0.05-4.0μg/kg, respectively. Average recoveries were71.6%-122.4%(chicken),73.1%-133.4%(pork liver), with associated RSD values2.9%-26.%(chicken),3.9%-26.9%(pork liver) under the selected conditions. The intra-day precisions were in the range of2.5%-22.9%(chicken), and4.4%-22.9%(pork liver), respectively.For over85%of analytes, the average recoveries were in the range from70%to120%with the corresponding RSD below25%, which were acceptable and in agreement with the criteria of Commission Decision2002/657/EC.In conclusion, the research work was summarized with four innovations. In this work, a novel universal analytical method with analytes grouping by polarity was developed. Compared with former sample preparation methods, this method significantly improved extraction efficiency and detection coverage of residues, and thus more accurate and precise results could be generated. Also, a drastic reduction of both sample preparation complexity and analysis time in routine analysis could be achieved by using this method. Therefore, this method could greatly facilitate the routine screening of contaminants and residues in animal origin food to improve the food safety. The work also raised four potential research aspects for the future researches.