16 Searching for Internal Standard for Chemical Routine Analysis. . .
191
Hep-PBIas an internal standard with regard to reducing the matrix effect. However,
a larger number of samples are required before drawing a final conclusion. The
sample size for Aza-1 results was too small to be handled statistically.
Discrepancies were observed between matrix-corrected and ISTD concentrations
in relation to DTX-1 and YTX using DHO. ESTD results of DTX-1 were in
agreement with the values from ISTD, in which no statistical significance was
detected (paired t-test, p < 0.05). However, both ISTD and ESTD results had
relative low recovery comparing to standard addition values, which indicated an
underestimation of the toxin amount. The concentration range of DTX-1 was
9.9–225 g kg
1 , and test samples in question were mussels only. One sample
contained DTX-1 above the regulatory limit (160 g kg
1 ) (Regulation (EC) No
853/2004), and recovery relative to matrix-corrected result for ESTD and ISTD
were acceptable; 82 and 87.4 %, respectively. The final toxin amount would cause
harvesting shutdown no matter how the data was quantified. The idea of using DHO
as an internal standard for DTX-1 because of the close retention time, is most likely
not feasible, although a larger number of samples are required in order to draw a
definitive conclusion.
Measurements of YTX (n D 12) ranged from 17.3 to 46.4 g kg
1 , were low
compared to the method limit of quantitation (LOQ) and EU regulatory limits
(Regulation (EC) No 853/2004). Average recovery of ISTD was 135.4 % (95 % CI:
123–148 %). Quantified toxin amount of YTX were significantly higher when using
internal standard than compared to standard addition, even though more samples
are needed to draw a final conclusion. Unfortunately, no samples containing higher
levels of YTX were available due to the seasonality of the sampling period.
In relation to PTX-2 (n D 4), two of four samples were scallops which clearly
suffered from major matrix effect of more than 40 % enhancement by using both
PBIs. The concentration range was wide, and extended from 8.9 to 155.7 g kg
1 .
For this reason, no comparisons based on the data were made. A high level of PTX-2
were measured in one of the samples (mussel), showing ESTD and matrix-corrected
results of 147.4 and 155.7 g kg
1 , respectively. ISTD values were measured to
<200 g kg
1 . Calculations done by the use of internal standard did not seem
to compensate for matrix effects affecting PTX-2, YTX or DTX-1, though more
samples need to be analyzed.
Response Drift
Response drift during analytical sequences were monitored by analyzing calibration
standards in the beginning, midway and the end of each analytical sequence.
Each calibration curve consisted of five levels of CRMs fortified in methanol. The
concentration levels of OA and DTX-1 ranged from 2 to 250 ng ml
1 , YTX; 10–
500 ng ml
1 , and Aza-1 and PTX-2; 2–100 ng ml
1 .
Two sets of calibration curves were constructed each day; one set using the toxin
peak areas (ESTD), and one set by the response ratios (ISTD). Four to five analytical
191
Hep-PBIas an internal standard with regard to reducing the matrix effect. However,
a larger number of samples are required before drawing a final conclusion. The
sample size for Aza-1 results was too small to be handled statistically.
Discrepancies were observed between matrix-corrected and ISTD concentrations
in relation to DTX-1 and YTX using DHO. ESTD results of DTX-1 were in
agreement with the values from ISTD, in which no statistical significance was
detected (paired t-test, p < 0.05). However, both ISTD and ESTD results had
relative low recovery comparing to standard addition values, which indicated an
underestimation of the toxin amount. The concentration range of DTX-1 was
9.9–225 g kg
1 , and test samples in question were mussels only. One sample
contained DTX-1 above the regulatory limit (160 g kg
1 ) (Regulation (EC) No
853/2004), and recovery relative to matrix-corrected result for ESTD and ISTD
were acceptable; 82 and 87.4 %, respectively. The final toxin amount would cause
harvesting shutdown no matter how the data was quantified. The idea of using DHO
as an internal standard for DTX-1 because of the close retention time, is most likely
not feasible, although a larger number of samples are required in order to draw a
definitive conclusion.
Measurements of YTX (n D 12) ranged from 17.3 to 46.4 g kg
1 , were low
compared to the method limit of quantitation (LOQ) and EU regulatory limits
(Regulation (EC) No 853/2004). Average recovery of ISTD was 135.4 % (95 % CI:
123–148 %). Quantified toxin amount of YTX were significantly higher when using
internal standard than compared to standard addition, even though more samples
are needed to draw a final conclusion. Unfortunately, no samples containing higher
levels of YTX were available due to the seasonality of the sampling period.
In relation to PTX-2 (n D 4), two of four samples were scallops which clearly
suffered from major matrix effect of more than 40 % enhancement by using both
PBIs. The concentration range was wide, and extended from 8.9 to 155.7 g kg
1 .
For this reason, no comparisons based on the data were made. A high level of PTX-2
were measured in one of the samples (mussel), showing ESTD and matrix-corrected
results of 147.4 and 155.7 g kg
1 , respectively. ISTD values were measured to
<200 g kg
1 . Calculations done by the use of internal standard did not seem
to compensate for matrix effects affecting PTX-2, YTX or DTX-1, though more
samples need to be analyzed.
Response Drift
Response drift during analytical sequences were monitored by analyzing calibration
standards in the beginning, midway and the end of each analytical sequence.
Each calibration curve consisted of five levels of CRMs fortified in methanol. The
concentration levels of OA and DTX-1 ranged from 2 to 250 ng ml
1 , YTX; 10–
500 ng ml
1 , and Aza-1 and PTX-2; 2–100 ng ml
1 .
Two sets of calibration curves were constructed each day; one set using the toxin
peak areas (ESTD), and one set by the response ratios (ISTD). Four to five analytical