2. AOACRIChemContMethods-2018Awards

Pang et al.: J ournal of AOAC I nternational V ol. 98, N o. 5, 2015  1439

Table 3. Comparision of the deviation rate of calculated concentration and expected concentration at the minimum concentration point for LC/MS/MS matrix matched calibration curve for Laboratory 20

Oolong tea calibration curve

No.

Pesticide

Expected concn, μg/kg Calculated concn, μg/kg

Difference

Deviation ratio, %

1 2 3 4 5 6 7 8 9

Imidacloprid

18.0 20.0

14.2 29.3

3.8

21.1

Propoxur

−9.3 −0.9

−46.3 −10.9

Monolinuron

8.0 8.0 8.0 8.0

8.9 5.5 3.6 4.8 7.2 5.9 5.7 3.3 6.4 5.9

Clomazon

2.5 4.4 3.2 8.8 2.1 2.3 4.7 1.6 2.1 7.7

31.7 55.0 39.8 55.0 25.9 28.6 13.1 58.8 19.8 25.9 32.1

Ethoprophos Triadimefon Acetolachlor

16.0

Flutolanil Benalaxyl

8.0 8.0

10 11 12 13 14 15 16 17 18 19 20

Kresoxim-methyl

80.0

69.5

10.5

Picoxystrobin

8.0 8.0 8.0

Pirimiphos-methyl

Diazinon Bensulide Quinoxyfen Tebufenpyrad

24.0 40.0

16.3 49.0

−9.0

−22.4

8.0 8.0 8.0

6.9 9.0 7.3

1.1

13.7

Indoxacarb

−1.0

−13.0

Trifloxystrobin Chlorpyrifos

0.7

8.3

80.0

110.0

−30.0

−37.5 −26.9

Butralin

8.0

10.2

−2.2

the measured concentration of these pesticides in the sample was lower than the minimum concentration point calculated (Calc. Conc.) of matrix standard curve and did not report the result, thus causing the quantification results at the low end of the calibration curves to show large and variable within- laboratory SDs. Some laboratories (e.g., Laboratory 21) did not report the test results when they found pesticide concentrations in the samples to be lower than the minimum concentration point of the calibration curves. In hindsight, the Study Director thinks the laboratories that did not report the test results should have designed the experiments better to include a much lower minimum concentration point on the calibration curves (e.g., being 50% of fortification concentrations) than were designed in this collaborative study, and they should have tried to ensure that the concentrations of pesticide residues in the samples after extraction were still above the minimum concentration points of calibration curves. While these issues became apparent during the collaborative study, theywere not seen as an issue during the SLVof themethod in which very good recoveries and precision were obtained for samples fortified at concentrations where the minimum point on the calibration curves used for quantification was 80% of the fortification concentrations of samples ( see Table 8). Therefore, the Study Director hopes to pay more attention in any future organization of or participation in a collaborative study to this oversight and take appropriate measures to minimize its occurrence again in a future study design. (c)  Method extraction efficiency and reproducibility for aged samples .—The method efficiency parameters such as recovery,

RSD R , and HorRat values in Table 9 are summarized in Table 10 per sectors. ( 1 )  By GC/MS.— The results of the statistical analysis of data obtained from the analysis of the 20 pesticides in aged oolong tea samples in Table 10 show that for the 16 collaborating laboratories using GC/MS, all the pesticides demonstrated within-laboratory repeatability RSD r <8%, between-laboratory reproducibility RSD R <25%, and HorRat values less than 1.0. ( 2 )  ByGC/MS/MS.— For the 14 collaborating laboratories using GC/MS/MS, all the pesticides demonstrated within-laboratory repeatability RSD r  <15%, while only six of the 20 (30%) pesticides demonstrated between-laboratory reproducibility RSD R <25%. Seventy percent (70%) of the 20 pesticides had between- laboratory reproducibility RSD R >25%. The HorRat values for all the pesticides were less than 2.0. ( 3 )  By LC/MS/MS.— The within-laboratory repeatability RSD r for the 24 laboratories using LC/MS/MS was <15% for all the pesticides, and the between-laboratory reproducibility RSD R was <25% for only eight of the 20 pesticides. The other 12 (60%) pesticides showed between-laboratory reproducibility RSD R >25%. The HorRat values were <2.0 for all 20 pesticides. The probable explanation for the large variability in RSD R >25% for some of the pesticides in aged oolong tea in the GC/MS/MS and LC/MS/MS analyses may be traced to just how aged samples are prepared. They are prepared by spraying pesticides onto dry tea powders in advance, which are then mixed uniformly. In a certain period after sample preparation, pesticides in tea slowly degrade during storage and transit, so in our SLV at an earlier stage a two-phase study was conducted to measure the rate of decrease in concentration of pesticides , RSD r