Page 16 - Volume 2, Issue 1, January 2009

Anagnostopoulos & Miliadis

16

ratories GmbH Germany) were used in this

study: bifenthrin, carbaryl, chlorpyriphos,

chlorpyriphos methyl, cyhalothrin-λ, delta-

methrin, diazinon, endosulfan-a, endosul-

fan-b, fenpropimorph, imazalil, iprodione,

kresoxyl-methyl, lindane, malathion, meth-

acrifos, parathion, penconazole, pirimicarb,

pirimiphos methyl, procloraz, procymidone,

propiconazole, thiabendazole, triadimefon,

triadimenol, triazophos and vinclozolin.

Acetone, 2,2,4-trimethyl pentane and

toluene were used for the preparation of

stock and working standard solutions. Ace-

tone, dichloromethane and petroleum ether

were used in the extraction procedure. All

solvents were pesticide residues grade, ob-

tained from Lab Scan (Ireland).

2. Stock and working solutions

Stock standard solutions at 1000 mg L

-1

were prepared in acetone for each pesti-

cide and stored at -20

o

C. Standard mixture

solutions of the compounds were prepared

in 2,2,4-trimethyl pentane/toluene (90/10)

at intermediate concentrations (1-10 mgL

-1

)

and stored at -20

o

C. In order to acquire the

retention time of each analyte, working so-

lutions containing only one analyte at 0.5

mgL

-1

were prepared and injected in the

chromatographic system.

Working standard mixture solutions for

measurement were prepared in an extract of

wheat flour, previously analyzed for the ab-

sence of compounds interfering with the an-

alytes. In the quantification of an unknown

sample one of the most serious problems is

the presence of unexpected interferences in

the matrix (6). The effect can be due to differ-

ent reasons e.g. the presence of a blank due

to solvent and/or reagents, or the presence

of a compound in the sample that contrib-

utes to the analytical signal (8). The detection

and correction of errors caused by matrix in-

terferences have been extensively studied for

a long time (2, 13). Matrix-induced enhance-

ment is a phenomenon commonly found in

the chromatographic analysis of pesticides in

food (3) that has been noticed in the analy-

sis of these contaminants by GC-ECD (7) and

GC-NPD (4). For this purpose matrix matched

standards (including matrix blanks) were

used.

The concentrations of the working solu-

tions were at 70 and 100% of the fortifica-

tion concentrations and quantification was

performed by bracketing. According to this

technique the peak area of the analyte in the

sample solution was bracketed between the

peak areas of two standard solutions, not

differing between them more than 20% (5).

3. Sample preparation

The sample processing according to the

applied method was the following (5, 12): An

aliquot of 10 ± 0.1 g of sample was weight-

ed into a 250 mL PTFE centrifuge bottle (Nal-

gene, Rochester, NY), 10 mL of water and

30 mL of acetone were added and stirred

for 1 min in an ultra-turrax homogenizer at

15.000 rpm, 30 mL of dichloromethane and

30 mL of petroleum ether were added and

the mixture was stirred for 1 min and then

centrifuged at 4.000 rpm for 2 min. Then, 25

mL of the supernatant liquid were evapo-

rated to dryness on a water bath at 65–70

o

C and 1 mL of 2,2,4-trimethyl pentane/tolu-

ene (90/10) was added. Another 15 mL of the

supernatant liquid were evaporated to dry-

ness on a water bath at 65–70

o

C and 3 mL

of 2,2,4-trimethyl pentane/toluene (90/10)

were added. The two extracts were placed

in ultrasonic bath for 30 sec and each was

transferred into a separate vial with a Teflon

stopper, ready respectively for NPD and ECD

chromatographic analysis. Simultaneous in-

jections were performed in the injectors

with the aid of 2 separate autosamplers.

4. Criteria for validation of the method

The accuracy was estimated by calcu-

lating the attained recovery. For validating

a method, mean recoveries of 70–120% are

considered acceptable, while in the case

of routine analysis, the acceptable recove-

ries are in the range of the mean recovery

± 2×RSD (5).

The precision of the method was evalu-

ated by assessing the relative standard devi-

ation (RSD) values under repeatability con-

ditions (same analyst, same instrument,