Immunoaffinity column cleanup with liquid chromatography using post-columnbromination for determination of aflatoxins in peanut butter, pistachio paste, fig paste, and paprika powder: Collaborative study

A collaborative study was conducted to evaluate the effectiveness of an imm unoaffinity column cleanup liquid chromatography (LC) method for the determ ination of aflatoxin B-1 and total aflatoxins at European regulatory limits . The test portion is extracted with methanol-water (8 + 2) for dried figs and paprika, and with methanol-water (8 + 2) plus hexane (or cyclohexane) f or peanut butter and pistachios. The sample extract is filtered, diluted wi th phosphate buffer saline, and applied to an immunoaffinity column. The co lumn is washed with water and the aflatoxins are eluted with methanol. Afla toxins are quantitated by reversed-phase LC with post-column derivatization (PCD) involving bromination. PCD is achieved with either an electrochemica l cell (Kobra cell) and addition of bromide to the mobile phase or pyridini um hydrobromide perbromide. Determination is by fluorescence. Peanut butter , pistachio paste, dried fig paste, and paprika powder samples, both natura lly contaminated with aflatoxins and containing added aflatoxins, were sent to 16 collaborators in 16 European countries. Test portions of samples wer e spiked at levels of 2.4 and 9.6 ng/g for total aflatoxins which included 1.0 and 4.0 ng/g aflatoxin B-1, respectively. Recoveries for total aflatoxi ns ranged from 71 to 92% with corresponding recoveries for aflatoxin B-1 of 82 to 109%. Based on results for spiked samples (blind duplicates at 2 lev els) as well as naturally contaminated samples (blind duplicates at 4 level s, including blank), the relative standard deviation for repeatability rang ed from 4.6 to 23.3% for total aflatoxins and from 3.1 to 20.0% for aflatox in B-1. The relative standard deviation for reproducibility ranged from 14. 1 to 34.2% for total aflatoxins, and from 9.1 to 32.2% for aflatoxin B-1. T he method showed acceptable within-laboratory and between-laboratory precis ion for all 4 matrixes, as evidenced by HORRAT values <1, at the low levels of determination for both total aflatoxins and aflatoxin B-1.