Calibration curve, quality assurance/quality control (QA/QC)
Calibration standards were prepared by dissolving known
concentrations of standards and surrogate in acetone. Calibration
plots were constructed for each analyte by plotting the ratio
of peak areas (analyte area divided by internal standard area)
versus the ratio of target analytes concentrations to the internal
standard concentration, resulting in at least a 7-point calibration
curve. A linear un-weighted regression, forced through the
origin, was performed for each analyte, resulting in correlation
coefficients (R2) exceeding over 0.9996 for all compounds. The
resulting equation of the line provided response factor (RF)
relative standard deviation (RSD) between 6.5% to 15.0% and
was used to calculate the concentrations of the analytes. Deviations
from the calibration curve of ±15% RF RSD would cause
rerunning of the standards, and construction of a new calibration
curve, or replacement of the capillary GC injector liner.
Retention time variations were generally ±5 s, and peak widths
at half-height were about 3 s. On the day of analysis, a sequence
of injection: solvent blank, calibration standards, solvent blank,
sample extract, solvent blank, calibration standards was used.
No carryover of the target compounds was detected in the
reagent blank. The surrogate recovery was observed to be between
52% and 61%. An optimal surrogate should have very similar
physical–chemical properties as the target analytes. As the
properties of target analytes are different, variation in their
recoveries might be possible. However, in our earlier report
(Mottaleb et al., 2012) surrogate naphthalene d8 provided
recovery between 56% and 59% from fish liver matrix, while
recoveries of target analytes, synthetic fragrances/metabolites
ranged from 95% to 114% with 4–9% relative standard deviation
(RSD).