6. AOACSPIFANMethods-2018Awards

334

sequence. The blank sample serves to verify that no significant contamination of chemicals, devices or analytical instrumentation has occurred. The analyte signals detected in the reference sample serve to check analyte retention time shifts and appropriateness of signal abundance according to commonly accepted system suitability criteria. The reference sample is also used to evaluate the overall method performance. Therefore, the analyte contents of the reference sample should be evaluated according to accepted practices to ensure the concentration is within a defined confidence-level. (2) In addition, prepare a sample that serves for determination of the Transformation-factor (the “Tf” sample) by spiking a suitable blank sample (e.g., analyte-free milk powder or milk) with 100 µL of each of the standard working solutions 4d and 5 (d 5 -3-MCPD-1,2-dioleoyl ester, 10 µg/mL d 5 -3-MCPD equivalent in toluene and glycidyl oleate, 1 µg/mL glycidol equivalent in toluene). Do not spike Tf-samples with any other internal standards. For powdered Tf-samples, discard the polar fractions containing the free analytes since the transformation factor is based on conversion of the bound analytes to their corresponding free forms in the non-polar fraction. Also extract and prepare liquid Tf-samples for the bound analytes according to sections d and e . (g) Sample storage .—Because the bound analytes are suspected to be unstable in infant formula during storage at room temperature, powdered samples should be stored in the freezer at a maximum of –20°C. For the same reason, liquid samples should be stored in the refrigerator at a maximum of 6°C. F. Calculations In principle, all free or bound analytes are quantified via an internal 1-point calibration. Therefore, the concentration of any analyte in a single sample is calculated via the ratio of its signal response in relationship to the signal response of the corresponding penta-deuterated internal standard present in the same sample. It is recommended that an automated signal integration software is used, which quantifies chromatographic analyte signals based on the signals of the isotope-labelled internal standards using the following calculations: w Analyte X,Y,Z = Peak area Analyte X,Y,Z . w internal standard d5-X,Y,Z / Peak area internal standard d5-X,Y,Z where W = analyte quantity (in µg/kg sample); Analyte X = free or bound 2-MCPD, free or bound 3-MCPD, bound glycidol (present as 3-MBPD after sample preparation); internal standard d5-X,Y,Z = free or bound 2-MCPD- d 5 , free or bound 3-MCPD-d 5 , bound glycidol-d 5 (present as 3-MBPD-d 5 after sample preparation). For glycidol determination, this calculation yields a raw value that must be multiplied by the transformation factor in order to account for the proportional conversion of any 3-MCPD present in the sample to “induced” glycidol. Small portions of detectable 2-MBPD and 2-MBPD-d5 that will be generated during glycidol-transformation do not impact the accuracy of the method and do not have to be considered. The transformation factor, which accounts for the proportion of induced glycidol that results from 3- MCPD conversion during sample preparation, is determined for each analytical sequence by analysis of the Tf-sample using the following equations:

PA

w

×

5

cidol inducedgly

d

−

Tf glycidol −

w

=

d cidol 5

inducedgly

−

PA

Tf glycidol −