On site, mycotoxin measurements shall enable rapid decisions on the acceptance or rejection of lots. Hence, results have to be available fast, easy to get and, first of all, reliable. An innovative approach using dust samples was tested for its fitness for on-site mycotoxin analyses of grain lots and compared to current practice in grain testing. To prove correlation between mycotoxin concentrations in dust and respective concentrations in grain, regression analyses were performed. To obtain data points, dust was sieved from grain and both samples were analysed. As the contamination of the overall sample and its dust particles correlated well (wheat: R2DON=0.85, R2ZEA=0.82; rye: R2DON=0.73), contaminations in the grain were predictable from concentrations determined in respective dust particles. For on-site analysis, common lateral flow devices (LFD) were evaluated for their suitability to detect deoxynivalenol (DON) in grain dusts. On site, grain and dust samples were taken during the unloading of trucks using a customised dust-sampler. In contrast to grain samples, no additional physical sample preparation or homogenisation step was needed for dust. Instead, the sample was directly extracted and analysed for DON using LFD. By means of the regression line DON concentrations in grain were predicted from dust results and compared to concentrations directly measured in grain samples. No false negative results were observed and a contaminated grain lot (<1000 ?g/ kg DON) could be clearly identified. Evidence for reduced measurement uncertainty compared to current practice at lower total measurement costs was given. In this way, the fitness for purpose of the new approach combining rapid analyses with dust sampling for on-site mycotoxin screening was shown. The innovative high-throughput technology has the potential to improve on-site mycotoxin measurements in terms of speed, sensitivity, manageability and reliability and thus is a promising tool for enhanced industrial self-control.

A novel reusable immunoaffinity cartridge containing monoclonal antibodies to aflatoxins coupled to a pressure resistant polymer has been developed. The cartridge is used in conjunction with a handling system inline to LC with fluorescence detection to provide fully automated aflatoxin analysis for routine monitoring of a variety of food matrixes. The handling system selects an immunoaffinity cartridge from a tray and automatically applies the sample extract. The cartridge is washed, then aflatoxins B1, B2, G1, and G2 are eluted and transferred inline to the LC system for quantitative analysis using fluorescence detection with postcolumn derivatization using a KOBRA® cell. Each immunoaffinity cartridge can be used up to 15 times without loss in performance, offering increased sample throughput and reduced costs compared to conventional manual sample preparation and cleanup. The system was validated in two independent laboratories using samples of peanuts and maize spiked at 2, 8, and 40 μg/kg total aflatoxins, and paprika, nutmeg, and dried figs spiked at 5, 20, and 100 μg/kg total aflatoxins. Recoveries exceeded 80% for both aflatoxin B1 and total aflatoxins. The between-day repeatability ranged from 2.1 to 9.6% for aflatoxin B1 for the six levels and five matrixes. Satisfactory Z-scores were obtained with this automated system when used for participation in proficiency testing (FAPAS®) for samples of chilli powder and hazelnut paste containing aflatoxins.