47 / 71
Pang et al.:
J
ournal of
AOAC I
nternational
V
ol.
98, N
o.
5, 2015
1443
with time and to determine the effect it has on the quantitative
analysis of pesticides in tea samples. The studies concluded
that pesticides will decrease in concentration by approximately
4.3–31.1% over a 3 month period.
Accordingly, correlation equations were derived for the
degradation rate in concentration and have been defined
for each of the pesticides (Figure 2) under study so that the
concentration degradation factor can be considered in the
calculation of the concentration of a pesticide at any specified
time during a study. On the basis of stability results of
pesticides in aged oolong tea determined by GC/MS/MS or
LC/MS/MS in a continuous 3 month study, trend charts were
plotted in which the determination time (day) was the X-axis
and the difference between each concentration determined at
time t and the concentration of pesticide estimated at time
zero was the Y-axis. The logarithmic equations presented
graphically in Figure 3 were obtained by fitting the 3 month
determination results. From these equations, the degradation
value of any of the 20 target pesticides at any specific day
could be calculated and applied to the raw data generated for
that pesticide in that particular laboratory.
In this collaborative study, the total length of time for all
laboratories to complete the analysis, which began March
28, 2013 and ended June 27, 2013 was 120 days. Therefore,
based on the day of analysis, it would be possible to calculate
the degradation rate for each of the 20 pesticides in the study
from the logarithmic equation defined for each pesticide.
To eliminate the influence from the natural degradation of
pesticides, the rates of decrease in concentration for each
pesticide residue measured in each laboratory were pooled
together into a database from which correlation equations were
derived. The procedure was used to calculate corrected/revised
method efficiency parameters as shown in Table 11 for oolong
tea. Table 11 shows that of the method performance parameters
such as S
r
, RSD
r
, S
R
, RSD
R
, and HorRat, only the RSD
R
data
are impacted significantly by the degradation factor. So, only
the raw data and corrected data of RSD
R
as shown in Table 12
will be discussed.
In the previous discussion, it was noted that only six of the
20 pesticides analyzed by GC/MS/MS showed RSD
R
≤25%,
while the other 14 pesticides had RSD
R
≥25%. Application of
the degradation correction factor to the previous data results in
a revised RSD
R
with an average of 22.8%, and the RSD
R
of the
remaining 14 pesticides was RSD
R
≤25%. A similar trend was
observed for the 70% of the pesticides analyzed by LC/MS/MS
that showed RSD
R
>25%.
In conclusion, it is the way the samples are prepared rather
than any inherent weaknesses in the method that accounts for the
unexpected large RSD
R
data for some of the selected pesticides.
(d)
Method extraction efficiency and reproducibility for
incurred samples
.—The method efficiency parameters such
as Rec., RSD
r
, RSD
R
, and HorRat values in Table 13 are
summarized in Table 14 per sectors.
(
1
)
By GC/MS.—
The statistical results of green tea incurred
samples in Table 14 show that concerning GC/MS within-
laboratory repeatability: RSD
r
<8% accounting for 100%
illustrates that this method’s repeatability is very good in
different laboratories. Between-laboratory reproducibility
RSD
R
<25% making up 100% demonstrates that the between-
laboratory reproducibility is very good. HorRat values <1.0
accounting for 100% shows that the method has achieved
very good reproducibility under the corresponding analytical
conditions;
(
2
)
By GC/MS/MS.
—Concerning GC/MS/MS: within-
laboratory repeatability RSD
r
<8% accounting for 100%
shows that the method’s repeatability is very good in
different laboratories. Between-laboratory reproducibility
RSD
R
<25% making up 100% proves that the between-
laboratory reproducibility is very good. HorRat values <1.0
accounting for 100% reveals that the method has achieved
very good reproducibility under the corresponding analytical
concentrations;
(
3
)
By LC/MS/MS.—
Concerning LC/MS/MS: within-
laboratory repeatability RSD
r
<15% accounting for 100%
shows that the method’s repeatability is very good in different
laboratories. Between-laboratory reproducibility RSD
R
<25%
makes up 100%.
(
4
) To summarize the discussion above, the collaborative
study results show that the method efficiency is acceptable.
Table 7. Distribution ranges for recovery, RSD
r
, RSD
R
, and HorRat values for fortified tea samples
Parameters of
method efficiency
Recovery, %
RSD
r
, %
RSD
R
, %
HorRat
Range
<75 75–100 >100
<8
8–15 >15 <16 16–25
>25
<0.50 0.5–1.0 1.01–2.0 >2.0
GC/MS (16 laboratories)
Green tea
0
40(100)
a
0
20(100)
0
0 20(100)
0
0
17(85)
3(15)
0
0
Oolong tea
0
40(100)
0
20(100)
0
0 6(30)
14(70)
0
0 19(95)
1(5)
0
GC/MS/MS (14 laboratories)
Green tea
0
40(100)
0
20(100)
0
0 20(100)
0
0
8(40)
12(60)
0
0
Oolong tea
0
40(100)
0
20(100)
0
0 12(60)
6(30)
2(10)
3(15)
15(15)
2(10)
0
LC/MS/MS (24 laboratories)
Green tea
0
40(100)
0
18(90)
2(10)
0 19(95)
1(5)
0
8(40)
12(60)
0
0
Oolong tea
0
40(100)
0
16(80)
4(20)
0 1(5)
14(70)
5(25)
1(5)
18(90)
1(5)
0
Total
0
240(100)
0
114(95)
6(5)
0 78(65) 35(29)
7(6)
37(31) 79(66)
4(3)
0
a
Number (percentage): the number accordance with the range (percentage of the number accordance with the range in the total number of determined
pesticides).