a b s t r a c tA reliable and cost-effective electrochemical method for the detection of deoxynivalenol (DON) in cereals and cereal-based food samples based on the use of a novel anti-DON Fab fragment is presented. The analytical system employed, Enzyme-Linked-Immunomagnetic-Electrochemical (ELIME) assay, is based on the use of immunomagnetic beads (IMBs) coupled with eight magnetized screen-printed electrodes (8-mScPEs) as electrochemical transducers.Using standard solutions of DON, a working range between 100 and 4500 ng/ml was obtained with an EC 50 of 380 ng/ml. The ELIME assay was employed to evaluate the cross-reactivity of the Fab fragment towards different trichothecenes revealing a good selectivity towards DON over other trichothecenes with the exception of 3-Ac-DON. The sensor was then applied to cereals and cereal-based food samples (wheat, breakfast cereal and baby-food) and a wide range of sample treatment procedures was tested. Within-laboratory precision (9-24% repeatability for breakfast cereals and 10-33% for baby-food) and recovery data (82-110% for breakfast cereals and 97-108% for baby-food) were calculated by analyzing blank breakfast cereals and baby-foods fortified with DON, demonstrating that the proposed method has the capability for use as a screening assay for DON in such products.

A method was developed for the determination of aflatoxin B1 in medical herbs (senna pods, botanical name Cassia angustifolia; devil's claw, botanical name Harpagophytum procumbens; and ginger roots, botanical name Zingiber officinale). The method, which was tested in a mini-collaborative study by 4 laboratories, is based on an immunoaffinity cleanup followed by reversed-phase high-performance liquid chromatography separation and fluorescence detection after post-column derivatization. It allows the quantitation of aflatoxin B1 at levels lower than 2 ng/g. A second extractant (acetonewater) was tested and compared to the proposed methanolwater extractant. Several post-column derivatization options (electrochemically generated bromine, photochemical reaction, and chemical bromination) as well as different integration modes (height versus area) were also investigated. No differences were found depending on the choice of derivatization system or the signal integration mode used. The method was tested for 3 different matrixes: senna pods, ginger root, and devil's claw. Performance characteristics were established from the results of the study and resulted in HorRat values ranging from 0.12 to 0.75 with mean recoveries from 78 to 91% for the extraction with methanolwater and HorRat values ranging from 0.101.03 with mean recoveries from 98 to 103% for the extraction with acetonewater. As a result, the method, with all tested variations, was found to be fit-for-purpose for the determination of aflatoxin B1 in medical herbs at levels of 1 g/kg and above.

An interlaboratory study was conducted to evaluate the effectiveness of an immunoaffinity column cleanup liquid chromatography (LC) method for the determination of aflatoxin B1 levels in corn samples, enforced by European Union legislation. A test portion was extracted with methanolwater (80 + 20); the extract was filtered, diluted with phosphate-buffered saline solution, filtered on a microfiber glass filter, and applied to an immunoaffinity column. The column was washed with deionized water to remove interfering compounds, and the purified aflatoxin B1 was eluted with methanol. Aflatoxin B1 was separated and determined by reversed-phase LC with fluorescence detection after either pre- or postcolumn derivatization. Precolumn derivatization was achieved by generating the trifluoroacetic acid derivative, used by 8 laboratories. The postcolumn derivatization was achieved either with pyridinium hydrobromide perbromide, used by 16 laboratories, or with an electrochemical cell by the addition of bromide to the mobile phase, used by 5 laboratories. The derivatization techniques used were not significantly different when compared by the Student's t-test; the method was statistically evaluated for all the laboratories. Five corn sample materials, both spiked and naturally contaminated, were sent to 29 laboratories (22 Italian and 7 European). Test portions were spiked with aflatoxin B1 at levels of 2.00 and 5.00 ng/g. The mean values for recovery were 82% for the low level and 84% for the high contamination level. Based on results for spiked samples (blind pairs at 2 levels) as well as naturally contaminated samples (blind pairs at 3 levels), the values for relative standard deviation for repeatability (RSDr) ranged from 9.9 to 28.7%. The values for relative standard deviation for reproducibility (RSDR) ranged from 18.6 to 36.8%. The method demonstrated acceptable within- and between-laboratory precision for this matrix, as evidenced by the HorRat values.

An interlaboratory trial for the determination of patulin in apple juice and fruit puree was conducted, involving 17 participants representing a cross section of industry, official food control, and research facilities. Mean recoveries reported ranged from 74 (10 ng/g) to 62% (25 ng/g) for apple juice and from 72 (25 ng/g) to 74% (10 ng/g) for fruit puree. Based on results for spiked samples (blind pairs at 2 levels), as well as naturally contaminated samples (blind pairs at 3 levels), the relative standard deviation for repeatability (RSDr) in juice ranged from 8.0 to 14.3% and in puree from 3.5 to 9.3%. The relative standard deviation for reproducibility (RSDR) in juice ranged from 19.8 to 39.5% and in puree from 12.5 to 35.2%, reflecting HORRAT values from 0.6 to 1.0 for juice and 0.4 to 0.9 for puree. The method showed acceptable within-laboratory and between-laboratory precision for each matrix, as required by current European legislation.

An interlaboratory study was conducted for the determination of deoxynivalenol in baby food and animal feed by high-performance liquid chromatography with UV detection. The study included 14 participants representing a cross section of industry, official food control, and research facilities. Mean recoveries reported ranged from 89% (at 120 g/kg) to 85% (at 240 g/kg) for baby food and from 100% (at 200 g/kg) to 93% (at 400 g/kg) for animal feed. On the basis of the results for spiked samples (blind duplicates at 2 levels), as well as those for naturally contaminated samples (blind duplicates at 3 levels), the relative standard deviation for repeatability (RSDr) in analyses of baby food ranged from 6.4 to 14.0% and in analyses of animal feed, from 6.1 to 16.5%. The relative standard deviation for reproducibility (RSDR) in analyses of baby food ranged from 9.4 to 19.5% and in analyses of animal feed, from 10.5 to 25.2%. The HorRat values ranged from 0.4 to 1.0 and from 0.7 to 1.3, for baby food and animal feed, respectively. The method showed acceptable performance for within-laboratory and between-laboratory precision for each matrix, as required by European legislation.