METHOD VALIDATION FOR DETERMINATION OF ORGANOCHLORINE PESTICIDE RESIDUES IN FOOD AND FEED

Validation of analytical methods ensures the reliability and accuracy of analytical results. To get a reliable result we performed a validation of the method taking into account all factors that may aff ect of the result. In this paper we optimized QuEChERS method for food and feed sample preparation as well as instrumental method using gas chromatography with mass spectrometry (GC-MS), and we obtained a method for successful determination of pesticides with calibration through the matrix. In this way, applying calibration through matrix, we satisfi ed the requirements for precision and reproducibility of the method being less than 20%, the accuracy rate in the range of 70-130% and method linearity throughout the range of interest.


INTRODUCTION
Pesticides are chemical compounds, which are used for the removal, suppression and destruction of plant and other pests. Unlike the majority of pollutants that are introduced into the environment without specifi c targets, pesticides are introduced with the intention to help the man, to increase nutrition and to protect the environment in the "fi ght" against harmful microorganisms and numerous pests (Stajkovac. 2009). Analysis of organochlorine pesticides (OCP) in food and feed samples comprises application of the number of methods in order to prepare samples for analysis and determination, depending on the type of the samples. Anastassiades et.al (2003) were the fi rst who developed Quick, Easy, Cheap, Eff ective, Rugged and Safe approach (QuEChERS) method for sample preparation, which gave fast, effi cient and reliable result of preparation in optimal time. To eliminate the infl uence of matrix, calibration through matrix that does not contain pesticide was performed as well. Identifi cation of pesticide residues is complex procedure, and moni-toring of low concentrations requires the use of highly sensitive instrumental analytical techniques,gas chromatography with mass spectrometry(GC-MS) or high performance liquid chromatography (HPLC).GC-MS predominates in the testing's due to its greater selectivity and sensitivity as compared to other analytical methods. Th e aim of this study was to validate the method for determination of organochlorine pesticides set down by regulations in Serbia ("Offi cial Gazette"RS. 29/2014) in food and feed by GC-MS technique aft er sample preparation applying QuEChERS method.

MATERIALS AND METHOD
Solvents and chemicals that were used in the validation process were HPLC gradient grade (Merck. Germany). Calibrant solutions were prepared using the pesticides mix of 20 pesticides (organochlorine pesticides mixture, manufacturer Ultra Scientifi c. lot CL-1069).Spiked samples were used to the purpose of internal control of the following parameters: recovery, precision, limit of quantifi cation (LOQ) and limit of detection (LOD). Pesticides mix that contains 19 pesticides (Chlorinated Pesticides-herbicides, AccuStandard. Inc. lot: 213091108) was used for spike preparation. In order to eliminate the infl uence of the matrix, calibration through matrix blank sample was performed as well (SANCO, 2014  Th is method of sample preparation is based on the extraction with acetonitrile in the presence of anhydrous magnesium sulfate (MgSO4) and anhydrous sodium acetate (CH3COONa). Sample (3 g) is measured and transferred into centrifuge tube, 3 mL of water is added and 3 mL of acetonitrile. Aft er intensive stirring on a vortex, 3 g of anhydrous magnesium sulphate and 1 g of anhydrous sodium acetate were added. Exothermic reaction occurred within 1 min aft er the intense stirring on vortex. Th e sample was then centrifuged until 5 min at 3000 rpm. 1 mL of upper acetonitrile extract is transferred into the 5 ml tube, which contained 150 mg of anhydrous magnesium sulphate, 100 mg of Primary and Secondary Amine (PSA) and 50 mg of C18 (Anastassiades et al. 2003). Th e tube content was centrifuged for 5 min at 3000 rpm. Aft er centrifuging, purifi ed and clear extract is obtained. Th en, 0.5 mL of the extract is evaporated in nitrogen and reconstituted with hexane. A sample prepared in this way is ready for the analysis on GC-MS (Agilent 7890B/5977A).
Organochlorine pesticides are separated on DB-5MS column (30 m· 0.25 μm· 0.25 mm). Sample volume of 4 μL (splitless mode) was injected at the constant pressure of 11.36 psi and and fl owed through the column of the carrier gas at the fl ow rate 1.2 mL/min. Th e list of analyzed OCP as well as retention time, molecule weight, ions important for the analysis of HCH isomer are shown in Table 3. Th e target and qualifi er abundances were determined by injecting the mixture of pesticide standards under the same chromatographic conditions using full scan with the mass/charge ratio ranging from 60 to 500 m/z. Standards were prepared in blank matrix extracts to counteract the matrix eff ect (SANCO. 2014). With the aim of obtaining more reliable results, further pesticide quantifi cation was performed in SIM mode. Pesticide quantifi cation was performed according to mass spectra and characteristic ions defi ned in SIM mode (Table 4), as well as the retention time of exit components, pesticides (Selvi et al. 2012). Th e processing of the obtained data was performed applying Mass Hunter Soft ware. Th e analysis of the method performance is performed in a calibration range from 0.005 to 0.1 mg/kg.

RESULTS AND DISCUSSION
Based on tests conducted on fi ve representative matrices and implemented to the internal controls, we obtained the results shown in Tables 5 and 6. Validation plan included determination of linearity, precision, reproducibility, accuracy and LOQ and LOD.
Linearity that was determined by setting a calibration curve was tested by regression analysis to establish the mathematical relationship between concentration and results in a set range of resultant values.
Th e precision of the method represents an agreement between values obtained in a series of repeated measurements of the same homogenous sample under the same determination conditions by at least 5 repeated measurements of the representative spiked matrix.
Th e reproducibility of the method represents matching results obtained by successive measurements of the same samples, but under many diff erent conditions, determination is accomplished by analyzing at least 5 spiked samples (representative matrix).
Th e accuracy represents the mean value of the obtained results and actual or accepted value of the results. It is expressed as the yield (recovery), calculated on the spiked sample in relation to the expected theoretical value of the results.
LOQ is the lowest concentration that can be determined with reasonable accuracy. It is calculated as the sum of the mean values of 10 repetitions and 3 standard deviations.
LOD , granica detekcije , je najniza koncentracija analita koja može biti dokazana ali ne i određena. Izračunava se kao zbir srednje vrednosti od deset ponavljanja na matrix spajku , koji odgovara prvoj tački kalibracije, i 10 standardnih devijacija. According to study of Maštovská et al. (2005), as compared to matrixmatched standardization, the analyte protectant approach off ers a more convenient solution to the problems associated with calibration in routine GC/ MS analysis of pesticide residues and possibly other susceptible analyte types in diverse samples. In a study on Alternative calibration techniques for coun-teracting the matrix eff ects in GC-MS-SPE pesticide residue analysis, Rimayi (2015) shows descriptive and inferential statistics proving that the matrixmatched internal standard calibration was the best approach for samples of varying matrix composition since it produced the most precise average mean recovery of 87% across all matrices tested. We demonstrated the same in our testing. Kartalovic et al. (2015) suggested application of gas chromatography with mass detector for the determination of pesticide traces, as it provides us with a confi rmation of result reliability by comparing the obtained spectrum with that from the library. GCMS analysis off ers good precision and recovery rate for determination of pesticides in hake fi llets when applying matrix calibration (Kartalovic et al., 2015b).

CONCLUSION
Based on the conducted research and appropriate preparation, calibration and verifi cation of the representative matrix we can conclude that the method for determination of pesticide residues in food and feed meets the eligibility criteria required by SANCO (2014). Th e method is linear in the range of 0.005 to 0.1 mg/kg. Th e linearity factor (R 2 ) is higher than 0.99. Th e precision and reproducibility rate for pesticide determination is a greater than 20%. Th e accuracy of the method is in the range 70-130%. Th e method can be successfully used for the determination of pesticide residues in food and feed.