Combining nanoflares biosensor and mathematical resolution technique for multi-class mycotoxin analysis in complex food matrices.

• A nanoflares biosensor was designed for multi-target mycotoxin analysis. • EEM data of nanoflares biosensor was resolved by ATLD algorithm. • Spectral overlaps and unknown background can be solved by mathematical resolution. • Mycotoxins can be quantified at extremely low levels with the LOD of 0.02 μg kg−1. • The accuracy of this proposed strategy is far superior to current routine methods. A multi-functional nanoflares biosensor of spherical gold nanoparticle (Au NP) modified by fluorophore-labeled oligonucleotides (ONS) was designed for ultra-sensitive multi-target mycotoxin analysis in food. Au NP was densely modified with multiplex highly oriented hairpins of oligonucleotides (ONS), each ONS was hybridized to a reporter with a distinct fluorophore label and specifically affiliative to its corresponding mycotoxin target. The fluorescent signals of reporters were pre-quenched by Au NP based on ONS hairpin structures and recovered when exposed to ONS's targets. Excitation-emission matrix (EEM) fluorescence detection was performed in EX and EM wavelength of 200–800 nm. Heavily overlapping spectra of fluorophores, mycotoxins and backgrounds were resolved by alternative trilinear decomposition (ATLD) algorithm, pure spectra of specific fluorophore responding to mycotoxin target can be extracted out for quantitative analysis. Four mycotoxins (Aflatoxin B1, zearalenone, Fumonisins B1, ochratoxin A) were simultaneously quantified at extremely low level with limit of detection <0.02 μg kg−1, the average recovery accuracies were higher than 91.7 % in various matrices of cereals, nuts, edible oils. This study realized an important breakthrough of the application of nanoflares biosensor and maybe promising to be as an alternative strategy for onsite mycotoxins monitoring of food. [ABSTRACT FROM AUTHOR]