16 Searching for Internal Standard for Chemical Routine Analysis. . .
193
different calibration series spread out in one analytical sequence containing samples
and controls, with a total number of 145 injections. All slopes were normalized to
the slope obtained from the first calibration curve (a D 0.20). The normalized slope
is shown in Fig. 16.1 as a dashed line.
The slopes of ESTD point toward a negative trend and deviate more from the
normalized slope than the ISTD slopes. The slope variation using Hep-PBI differs
in both positive and negative directions indicating that the internal standard response
was not affected to the same extent as the toxin. Hence, Hep-PBI compensated
for the negative trend of a decreasing slope which leads to overestimating of toxin
concentrations.
Response Variation
The response stability for the analytes and internal standards during analytical
sequences were tested in previously negative shellfish spiked with CRM (125 g
kg
1 ). For Aza-1 and PTX-2 a total of 98 crude shellfish extracts were studied, of
which 46 were mussels, 37 scallops and 15 cockles (Cerastoderma edule), analyzed
by the neutral method (Stobo et al. 2005) over 4, 2 and 3 days, respectively. OA,
DTX-1 and YTX were studied in 39 mussels and 37 scallops, analyzed over 2
days. The alkaline method (EU-Harmonised Standard Operating Procedure) was
used during the first day of analysis whilst the neutral method (Stobo et al. 2005)
was used during the second day. The determined peak areas were normalized to
the first sample in each sequence before the comparison of day-to-day variations of
response drift.
The results suggest that response variation depends on the species because their
matrix behaves differently. For the negative ionization mode, DHO followed the
response of OA (p > 0.05), and were not affected by different methods, instruments
or species. Peak areas for DTX-1 and YTX drifted in opposite directions (p < 0.05)
when considering the results without taking into account the different species (not
shown). If the results were split by species the response variations of DTX-1 in
mussels (Fig. 16.2) did not statistically differ from DHO. However, as seen in
Fig. 16.2, there were lower DTX-1 values on day one than day two due to the
difference in methodology and instrumentation, suggesting that a conclusion cannot
be drawn based on this study. YTX behaved differently from DHO regardless of
species, methodology and instrumentation.
The results for positive mode ionization demonstrated no statistical significance
in the response variation between Hep-PBI and Aza-1 considering all three species
together (p D 0.14) (not shown), or individually (example: scallops (p < 0.05),
Fig. 16.3). In fact, using Pent-PBI for Aza-1 quantification seems promising in
mussels (not shown) since their peak areas showed a linear relationship (p D 0.4).
Pent-PBI was included at a later stage in the study, and that explains why no data is
shown in Fig. 16.3 for the first and second day.
193
different calibration series spread out in one analytical sequence containing samples
and controls, with a total number of 145 injections. All slopes were normalized to
the slope obtained from the first calibration curve (a D 0.20). The normalized slope
is shown in Fig. 16.1 as a dashed line.
The slopes of ESTD point toward a negative trend and deviate more from the
normalized slope than the ISTD slopes. The slope variation using Hep-PBI differs
in both positive and negative directions indicating that the internal standard response
was not affected to the same extent as the toxin. Hence, Hep-PBI compensated
for the negative trend of a decreasing slope which leads to overestimating of toxin
concentrations.
Response Variation
The response stability for the analytes and internal standards during analytical
sequences were tested in previously negative shellfish spiked with CRM (125 g
kg
1 ). For Aza-1 and PTX-2 a total of 98 crude shellfish extracts were studied, of
which 46 were mussels, 37 scallops and 15 cockles (Cerastoderma edule), analyzed
by the neutral method (Stobo et al. 2005) over 4, 2 and 3 days, respectively. OA,
DTX-1 and YTX were studied in 39 mussels and 37 scallops, analyzed over 2
days. The alkaline method (EU-Harmonised Standard Operating Procedure) was
used during the first day of analysis whilst the neutral method (Stobo et al. 2005)
was used during the second day. The determined peak areas were normalized to
the first sample in each sequence before the comparison of day-to-day variations of
response drift.
The results suggest that response variation depends on the species because their
matrix behaves differently. For the negative ionization mode, DHO followed the
response of OA (p > 0.05), and were not affected by different methods, instruments
or species. Peak areas for DTX-1 and YTX drifted in opposite directions (p < 0.05)
when considering the results without taking into account the different species (not
shown). If the results were split by species the response variations of DTX-1 in
mussels (Fig. 16.2) did not statistically differ from DHO. However, as seen in
Fig. 16.2, there were lower DTX-1 values on day one than day two due to the
difference in methodology and instrumentation, suggesting that a conclusion cannot
be drawn based on this study. YTX behaved differently from DHO regardless of
species, methodology and instrumentation.
The results for positive mode ionization demonstrated no statistical significance
in the response variation between Hep-PBI and Aza-1 considering all three species
together (p D 0.14) (not shown), or individually (example: scallops (p < 0.05),
Fig. 16.3). In fact, using Pent-PBI for Aza-1 quantification seems promising in
mussels (not shown) since their peak areas showed a linear relationship (p D 0.4).
Pent-PBI was included at a later stage in the study, and that explains why no data is
shown in Fig. 16.3 for the first and second day.