Detection of mite infestation in wheat by electronic nose with transient flow sampling

Abstract

A total of 71 wheat samples of 70 g each, to which were added 0–70 flour mites (Acarus siro L), were scanned, 1–2 weeks after preparation, by a conducting polymer-type electronic nose with transient flow sampling. The device uses an array of non-selective sensors to produce a digital fingerprint (odour profile) relating to the volatile composition of the sample. The results demonstrate the ability of the device to detect the mites in wheat at concentrations of relevance to the cereal trade, of the order of just 1000 A siro kg−1 (around 0.001% by weight). One week from preparation, discrimination between samples with no added mites and samples with 70 added mites achieved a classification accuracy of over 83%. Evidence is presented which suggests the known mite-produced volatile undecane as the origin of the electronic nose response. © Crown copyright 1999