3rd European Conference on Pesticides and Related Micropollutants in the Environment

37

SINGLE-LABORATORY VALIDATION OF AN ANALYTICAL METHOD FOR THE ANALYSIS OF VARIOUS PESTICIDES H. Karasali, G. Balayiannis, H. Anagnostopoulos and A. Hourdakis
Benaki Phytopathological Institute, 7 Ekalis str., Kifissia 14561, Greece E-mail: [email protected]

ABSTRACT
A new method was developed and single-laboratory validated for the quantitative determination of four pesticides (alachlor, chlorpyrifos-methyl, fenthion and trifluralin) in their commercially available EC (Emulsifiable Concentrate) formulations, widely used in Greece. The method enables the analysis of a number of active ingredients with the same chromatographic column and chromatographic conditions. Method validation is performed on the appropriate performance parameters, as indicated by the CIPAC (Collaborative International Pesticides Analytical Council) and EU (European Union) guidelines. The analytical system includes a capillary gas chromatograph equipped with an FID (Flame Ionization Detector) and a PTV (programmable Temperature Vaporizing) split injector, operated in CT (Constant Temperature) Mode. Two columns (CP-Sil 8Cb and DB-1701) of different polarities were used in order to establish the specificity of the method. The results presented are statistically processed.

INTRODUCTION
Several hundred pesticides of different chemical nature are currently widely used throughout Europe and USA. Because of their widespread use, pesticides are currently detected by determination of their residues in various environmental matrices, such as soil, water and air [1,-3]. In recent years, people have become more conscious of the risks and dangers arising from the intense use of pesticides. An important factor for minimizing these hazards and simultaneously achieving the desired level of pest control is the appropriate application of pesticide products of standard quality. The use of sub-standard formulations can result not only in ineffective pest control but also to several other problems such as pest-resistance, toxicity to mammals etc. As a consequence agencies for the monitoring of the quality control of pesticide products have been established. Monitoring of the quality of pesticide formulations is becoming more and more important, due to their widespread use in agriculture. The most important component of quality control of pesticides is testing their chemical composition. The majority of methods used for this is collaboratively tested and fully validated methods published by two International Organizations: AOAC (Association of Official Analytical Chemists) and CIPAC (Collaborative International Pesticides Analytical Council). These methods prescribe numerous specific analytical conditions, use a relatively large number of stationary phases, internal standard, eluent compositions and have increased demands on instrumentation. However, due to the great variety of pesticides to be monitored, the need for new methods with higher sample output and lower cost of analysis has become imperative. An answer to this need is the development of single-laboratory validated methods for pesticide analysis. Analytical laboratories must use validated analytical methods in order to ensure the reliability and quality of the data produced. Method validation may be described as a set of tests used to establish and document the performance characteristics of a method and thereby demonstrate that the method is fit for a particular purpose [4, 5]. A method can be fully or single-laboratory validated. Full validation of an analytical method is the examination of all the method characteristics in an inter-laboratory method performance study (collaborative study or collaborative trial). Internationally accepted protocols have been established for the full validation of a method of analysis by collaborative trials, most notably the International Harmonized Protocol [6] and the ISO Procedure [7]. Limiting factors for completing ideal multilaboratory validation studies include high costs, lack of sufficient expert laboratories available and willing to participate in such studies, and overall time constraints. In order to describe validation in one laboratory, in place of the previously used expression in-house the term single-laboratory validation was recommended by an AOAC/FAO/IAEA/IUPAC International Workshop on Principles on Method Validation [8] and was accepted by the Codex Committee on Methods of Analysis and Sampling. In the case of pesticide formulation analysis, the required performance characteristics have been defined in general by AOAC International [9], and specifically by CIPAC [10]. There are also a number of other requirements that have to be fulfilled, prescribed by national and international legislation, such as the Directive 91/414/EEC [11] in the case of the European Union, where methods are required to be robust, accurate and precise. There is a continuing need for reliable analytical methods for use in determining compliance with national regulations as well as international requirements in all area of analysis. Novel single laboratory

38

Analytical and Evaluation Techniques

validated methods, developed for monitoring purposes, have to determine the content of preferably more than one active ingredients in commercial pesticide formulations, under the same chromatographic column and elution system [12, 13]. Each sample is analyzed separately from the sample of another active ingredient. In one sample usually not more than two and known active substances are to be separated from the impurities of the technical material and components of formulations [13]. Therefore these methods allow the analysis of a large number of pesticides by a limited number of chromatographic columns, elution system and internal standards. The present method has been developed for the analysis of four of the most widely used pesticide products (alachlor, chlorpyrifos-methyl, fenthion and trifluralin) for the pesticide quality control in Greece.

MATERIALS AND METHODS

Analytical standard of trifluralin, fenthion, alachlor and chlorpyrifos-methyl were certified to be pure by 99.4%, 96.2%, 99.8% and 99.8% respectively. Dipentyl phthalate 99% pure was used as internal standard (I.S.). Each of the above mentioned compounds was supplied with a certificate stating the exact percentage purity, which has been determined by the ISO accredited supplier. For each active ingredient (a.i.) five different batches of commercial available EC formulations had been provided from their manufacturers, together with the Certificate of Analysis for each batch. An internal standard solution of 0.1512mg/ml was for preparing all standard (stock and working) and sample solutions throughout the whole procedure. Individual analytical standard stock solutions of alachlor, chlorpyrifos-methyl, fenthion and trifluralin (0.802mg/ml, 0.808mg/ml, 0.822mg/ml and 0.810mg/ml respectively) were prepared with the dilution of the appropriate amount of the respective analytical standard with the internal standard solution. The three working solutions were prepared by three independent dilutions of the stock solution with internal standard solution at concentration of about 0.8, 1 and 1.2 times of the nominal concentration of a.i. in the formulated product. The freshly prepared working solutions were used for establishing the chromatographic systems precision throughout repeatability testing and for defining the linearity of response for each individual component. Concentrated Samples of the formulations were prepared as follows: Appropriate quantity of formulation, containing approximately 80mg ( 5%) of active ingredient was diluted to the volume with internal standard solution. Appropriate quantity of this solution was diluted with the internal standard solution so that the final concentration fell within the calibration curve limits (Diluted Samples).

GAS CHROMATOGRAPHIC SYSTEM AND CONDITIONS

A ThermoFinnigan Trace Gas Chromatograph (GC) was used with the following chromatographic system: A Programmable Temperature Vaporizing (PTV) split injector operated in the Constant Temperature (CT) mode, a Flame Ionization Detector (FID) and an autosampler AS 2000. Two columns of different polarity were used: a 25m X 0.53mm, i. d. , 1m film thickness CP-Sil 8Cb low polar column and a 15m X 0.53mm, i. d. , 1m film thickness DB-1701 medium polar column. Operating conditions for CP-Sil 8Cb: Helium as carrier gas was set at 13kPa, 250 0C the injector temperature and 250 0C the detector temperature. Injection volume was 0.5l. The split flow set at 95ml/min and the split ratio was 13. Operating conditions for DB-1701: Helium as carrier gas was set at 45kPa, 250 0C the injector temperature and 250 0C the detector temperature. Injection volume was 0.5l. The split flow set at 45ml/min and the split ratio was 10. The air flow rate was set in both cases at 350ml/min whereas the split flow for hydrogen was set at 35ml/min. In both cases the temperature program for achieving good analysis of the compounds was the following: from 80 0C to 220 0C at a rate of 35 0C / min. The temperature remained isothermal for 1min at 800 C and for 8min at 220 0C. Identification of the peaks was achieved by comparing their retention time (RT) of each individual peak with the RTs of the reference standards. A computer program was used for the evaluation of GC runs.

3rd European Conference on Pesticides and Related Micropollutants in the Environment

39

RESULTS AND DISCUSSION

The GC determination was shown to be one of the major potential sources of the uncertainty of the analysis [14-16]. The performance of any chromatographic system is changing with time. These changes increase the uncertainty of chromatographic measurements. Therefore, before the instrumental analysis starts, the GC performance parameters should be checked to verify their suitability for the purpose of the analysis [16]. For evaluating GC column performance the plate number (Neff), the plate height (H), the peak asymmetry (As), the resolution (Rs) and the Thrennzahl separation number (Tz) were calculated. The parameters to be tested for the characterization of the methods applied for the determination of active ingredients and impurities in pesticide products based on CIPAC [10] and EU [11] guidelines.

2.2

2.0

Response Ratio, Yi

1.8

1.6

1.4

1.2

1.0

.8 1.0 1.5 2.0 2.5 3.0 3.5

Rsq = 0.9999

Ratio of Concentration, fa=Cas/Cis

Figure 1 Confidence Interval for Linear Regression of Trifluralin in CP Sil 8 Cb

40

Analytical and Evaluation Techniques

The performance parameters assessed in the present study were [10]: Linearity of response for the analyte in the method. After having performed the multi-point calibration (3 X 2 injections) as shown in the representative Figure 1 for trifluralin, correlation coefficient, slope and intercept with confidence limits and standard deviation (SD) of relative residuals were found to be acceptable since correlation coefficient is > 0.997 and the standard deviation of relative residuals is 0.01 [17]. Correlation coefficient (r), slope and intercept with confidence limits at 95% level were determined using ANOVA analysis. Standard deviation of relative residuals (sYrel) was also determined. An estimation of precision of the procedure. Data obtained from the analysis in duplicate of samples from five different batches of each pesticide was used for the calculation of the standard deviation (Sr) and the relative standard deviation (%RSDr). The acceptability of the %RSDr (coefficient of variation, SD) should be based on the modified Horwitz equation (Eq. 1), an exponential relationship between the among-laboratory relative standard deviation (RSDR) and the concentration (C). The calculated %RSDr values were correct and indicated that the repeatability was acceptable. RSDR = 2(1-0.5logC) For the estimation of repeatability (RSDr), Equation 1 is modified to: %RSDr = %RSDR X 0.67. (2) In that case the calculated %RSDr values were correct indicated that the repeatability was acceptable. A demonstration of the accuracy of the procedure. In the case of pesticide formulations the accuracy of the procedure can be determined by a number of samples containing a known quantity of the analyte. These could be either laboratory-prepared coformulant mixtures, in which known quantity of analyte (corresponding to the quantity demanded by the method) has been added [11], or Certified Reference Materials [4]. In the present study all the batches from all the formulations were accompanied with their Certificate of Analysis. The results obtained were compared with the reference values with the paired t-test. The results were found to be not significantly different at a significance level of a=0.05 (since tcalc is tcrit for all the cases). A demonstration of no interference from excipients. In the case of pesticide analysis concentrated blank formulation and concentrated samples were analysed in order to check that there is no interference with the analyte from any expected compounds in the sample, degradation products, metabolites or known additives. Concentrated blank formulation samples were prepared following the same sample preparation procedure with the commercial formulation samples, but they were diluted with acetone instead of internal standard solution and concentrated to half volume. In all cases lack of interference in the region of the pesticide and the targeted internal standard (see representative figure 2) demonstrated by the analysis of the concentrated blank and concentrated samples in two columns of different polarities. Comparison of the results obtained with two different columns The results obtained from the analysis in duplicate of the samples from five different batches of each pesticide, with the two columns of different polarity were compared with the paired t-test. As the results obtained from the two columns were not significantly different, the method is validated for the tested pesticides. (1)

3rd European Conference on Pesticides and Related Micropollutants in the Environment

41

CONCLUSIONS
The GC-FID generic system proved to be suitable for routine analysis of various pesticide formulations and the method performance characteristics meet the requirements of a quantitative method as specified in official Guidelines [10]. This method also is more efficient than the official methods for routine analysis of the pesticide formulations, as the operation cost and the total time of analysis are significantly reduced.

Figure 2 Interference of Fenthion in CP Sil 8 Cb

ACKNOWLEDGEMENTS
This study has been supported by the International Atomic Energy Agency (IAEA), Co-ordinated Research Project Quality Control of Pesticides. [Contract No 11772/RO].

REFERENCES
1. 2. 3. 4. 5. 6. 7. Barcelo D., Analyst, 116, (1991), 681-689. Barcelo D. and Hennion M.C., Analytica Chimica Acta, 338, (1997), 318. Harner T., Widerman J. L., Jantunen L. M. M., Bidleman T. F. And Parkhurst, Environmental Pollution, 106, (1999), 323332. Thomson M., Ellison S. L. R. And Wood R., Pure Appl. Chem., 74, (2002), 835-855. Eurachem Guide: The fitness for purpose of analytical methods: A Laboratory Guide to Method Validation and Related Topics, ed. LGC, Teddington. 1996. Horwitz W., Pure Appl. Chem., 60, (1988), 855-864, revised Horwitz W., Pure Appl. Chem., 67, (1995), 331-343. International Standard 5725, International Standardization Organization, Geneva, 1st ed., 1994.

42

Analytical and Evaluation Techniques

8. 9.

10. 11. 12. 13. 14. 15. 16. 17.

Principles and Practices of Method Validation, ed. A. Fajgeli and A. Ambrus, Royal Society of Chemistry, London, 2000, 253-295. AOAC Official Methods Program, Quick and Easy Associate Referees Manual on Development, Study, Review and Approval Process, AOAC International (1995), http://www.Aoac.org/vmeth/omamanual/omamanual.htm, accessed on Tuesday, 18 November. London. CIPAC, Guidelines on Method Validation to be performed in Support of Analytical Methods for Agrochemical Formulations. Collaborative International Pesticides Analytical Council, Document No. 3807, Black Bear Press, Cambridge, 1999. EEC Directive, 91/414/EEC. Methods of Analysis for Technical Material and Preparations (Annex IIA, point 4.1 and Annex IIIA, point 5.1). European Commission, Brussels (1991). Karasali H., Anagnostopoulos H., Ekonomopoulou H., Hourdakis A., Proceedings of the 2nd European Conference and Related Organic Pollutants in the Environment, (2002), 299-303. Karasali H., Anagnostopoulos H., Ekonomopoulou H., Hourdakis A., Intern. J. Environ. Anal. Chem., 84, (2004), 55-63. Bettencourt da Silva R. J. N., Santos J. R., Camoes M. F. G. F. C., Analyst, 127, (2002), 957-963. Alder L., Korth W., Patey A., van der Schee, Schoeneweis S., J. AOAC Int. 84, (2001), 1569-1575. Soboleva E., Ambrus A., J of Chromatography A, 1027, (2004), 55-65. J. C. Miller and J. N. Miller. Statistics for Analytical Chemistry, 3rd edn. Ellis Horwood, London (1993).