6. AOACSPIFANMethods-2018Awards

333

volatile solvents have been evaporated to a large extent. The solvent evaporation process can be accelerated as needed by heating a base plate between 40 and 50°C, but do not overdo this procedure in order to avoid altering the lipid phase. (3) Re-dissolve the lipid residue in 2 mL of a mixture of isohexane–tBME (4 + 1, v/v) and transfer the solution into a previously weighed 8 mL screw cap vial. Add an additional 2 mL of a mixture of isohexane– tBME (4 + 1, v/v) to the vial that contained the lipid residue and combine the organic phase with the first 2 mL aliquot in the weighed 8 mL screw cap vial. Evaporate the solvents under a stream of nitrogen using conditions specified in section d (2) . Further preparation of the non-polar isohexane–tBME fraction, which contains lipids and bound analytes, is described in section ( e ). (e) Powdered and liquid samples.—Sample preparation of the lipid fraction to determine bound 2- and 3- MCPD and bound glycidol.—(1) Weigh the resulting lipid phase (section a (6) and/or section d (3)), and by either selectively separating excess materials or by adding a blank matrix (e.g. analyte free cold-pressed oil) adjust the sample weight to 500 ± 50 mg. Dissolve the lipid residue in 3 mL tBME and store the solution for at least 15 min at –25 ± 3°C. (2) Initiate the base-catalysed transesterification at –25 ± 3°C by adding 1.4 mL of a methanolic sodium hydroxide solution (solution 1) (stored at –25 ± 3°C). Terminate the reaction after 15 to 18 h by adding 2.4 mL of the acidic stopping reagent (solution 3) that has been cooled to –25°C ± 3°C. Shake the mixture vigorously in order to achieve complete acidification of the sample. Carry out the base-catalyzed transesterification with diluted reagent solutions (solution 2 and 4) accordingly if the reaction is conducted over the weekend; this reaction time equates to 62 to 66 h. Place the reaction vessel under a nitrogen stream for at least 10 min to achieve evaporation of the volatile solvents and to convert glycidol into MBPD to a sufficient extent. (3) Wash the transesterified samples two or three times with 2 to 3 mL isohexane or an isohexane–tBME mixture (4 + 1, v/v) using Pasteur pipettes to separate and discard the organic phases. (4) Extract the meanwhile largely matrix-free aqueous phase three times using 2 mL diethyl ether or a diethyl ether–ethyl acetate mixture (9 + 1, v/v). Centrifuge the samples and separate the organic phases, combining them in an 8 mL screw cap vial containing a spatula tip of anhydrous sodium sulphate. Add 100 µL of phenylboronic acid solution (solution 7) and concentrate to dryness under a stream of nitrogen. (5) For the subsequent GC-MS analysis, re-dissolve the soluble residue in 300 µL isooctane and transfer a 200 µL aliquot into a2 mL GC vial containing a glass insert. Centrifuge the aliquots intended for GC-MS analysis. If sediment is present, decant the clear supernatant into a new vial insert prior to GC-MS analysis in order to prevent contamination of the injection system. Note: Moderately exceeding the time and temperature ranges specified in e (2) for alkaline catalyzed transesterification should not have a significant impact on the accuracy of the method because any changes in the transformation of 3-MCPD into glycidol is accounted for by determination of the batch- corresponding transformation factor. However, systematic deviations in the transesterification time and temperature have to be validated in order to evaluate method performance. (f) System suitability and determination of the Tf-factor; preparation of blank, reference and Tf-samples .— (1) In a similar manner according to the previous sections, prepare an analyte-free blank sample (e.g., spray-dried milk powder or milk) and a reference sample with known analyte contents with each analytical