System Of Fast Detection Of Mold In Food Based On The Smart Tongue

Food and feeds are often contaminated with spoilage or pathogenic microorganisms, which can reduce the edibility of many products or affect their taste by the production of undesirable flavours. In some cases, these microorganisms can produce toxic substances that are potent carcinogens, teratogens, genotoxins, and mutagens and pose a severe hazard to animal and human health. In addition, they are responsible for high levels of economic loss. Fungi are the major pathogens of plant diseases, and some can directly cause human and animal diseases. Some types of mold can infect food or animal feed, in which produce very toxic substances, namely mycotoxins. Aspergillus toxin in all animal species tested can cause cancer, and is one of the strongest cancer-causing compounds. At present, the quality control tests that are used include traditional microbiological and/or physico-chemical techniques (gas and liquid chromatography, mass spectrometry, and spectroscopic techniques), which are effective and accurate. Nevertheless, these methods have some typical drawbacks that make them unsuitable for frequent routine testing; these include high costs of implementation, long analysis times, low throughput of samples, and the need for highly qualified manpower. Rapid detection methods for measuring microbial activity are very important for the promotion of food safety and rapid diagnosis of food poisoning.A novel voltammetric electronic tongue, smart tongue was employed to detect the major mold of the moldy food. The sensor part of smart tongue--standard three-electrode system was consisted of several metallic working electrodes, an Ag/AgCl reference electrode and a pillar platinum electrode as counter electrode. The innovative aspect in this electronic tongue is the application of a new electrochemical method-----the multifrequency large amplitude pulse voltammetry (MLAPV) which is very useful for discriminating samples in the voltammetric electronic tongue by applying three frequency segments,1 Hz,10 Hz and 100 Hz. Multivariate data analysis (MVDA) such as Principal component analysis (PCA), and Soft Independent Modeling of Class Analogy(SIMCA) were often used to analyzing the data from the electronic tongue. With the advantages of rapid detection and simple sample preparation, it seems to be suitable for detecting the liquid cultures of the fungi. The main research work is as follows:1. Monitoring the Growth of four mold species with Smart tongueThe liquid cultrues of mold at different stages are different. Smart tongue that could reflect well the compositive information of liquid medium, combining with PCA was employed to monitor the growth of the four mold in liquid medium and differentiate different mold. Then the dispersion (σ) of the samples in the PCA score plots was done to review the monitoring ability of smart tongue. Dry weight and pH measurement of the mold were used as the refrence methods. The growth monitored by smart tongue was basically consistent with the results of dry weight measurement. Results indicated that the best electrodes and the corresponding frequency segements were as follows:The Pt, Au, Pd and W electrodes in all three frequency for Aspergillus; the Pt, Au and Pd electrodes in all three frequency and W in 100 Hz for Penicillium; the Pt electrode in 10 Hz, the Au electrode in both 1 Hz and 10 Hz for Mucor; while for Rhizopus, the Pt, Pd and Ti electrodes in the three frequency and the Ag electrode in 10 Hz.2. Recognition of four mold species with Smart TongueSmart tongue, combined with principal component analysis, was employed to recognize four mold species in liquid medium in this paper. A quantitative value of the separation was determined and a good separation term(dm;n) between two classes was studied. Results from smart tongue indicated that the best combinations of working electrodes with specific frequency segments for differentiation of four mold in liquid medium were as follows:both the Au and Pd electrodes are in 1 Hz,10 Hz and 100 Hz, and Ag in 1 Hz. A good separation was obtained when the separation term, d> 2.000.3. Intelligent models of Classfying the four mold species by Smart tongue and SIMCA Pattern Recognition Based on PCASmart tongue could reflect the compositive information of liquid samples. The liquid cultures of different mold are different, while those of the same mold represent their characteristics. Complying with SIMCA based on PCA, smart tongue was studied for classifying and identifying the four mold including four species of Aspergillus, Penicillium, Mucor and Rhizopus. The liquid cultures of mold culturing for 30 h were detected by smart tongue and divided into training set and testing set. The best models based on the training set for classifying were detected. The discrimination accuracy of the best models were 100% for both the training sets and the testing sets as follows:The Pt electrode in 1 Hz and 10 Hz, both the Au and Ag in 1 Hz for Aspergillus; excpt the Pt in 1 Hz, the other seventeen combinations are the best for classifying Penicillium; the Au electrode in 1 Hz,10 Hz and 100 Hz, Pd in 1Hz and 100 Hz, W in 10 Hz for Mucor, the Pd, W and Ti in all the frequency and Ag in 1 Hz for Rhizopus.4. Recognition of five common mycotoxin producing Aspergillus with Smart TongueThe smart tongue combined with PCA detected the liquid cultures for different training time (30h,40h,50h) of five common mycotoxin producing Aspergillus, including Aspergillus flavus(Af.), Aspergillus parasiticus s(Ap.), Aspergillus versicolor s(Av.), Aspergillus fumigatus s(Afu.) and Aspergillus ochraceus(Ao.). The five Aspergillums cultured for 30 h could not be differentiated by smart tongue, while these for 40 h had a trend of being distinguished and for 50 h distinguished well. As the longer culture time, the better differentiation for the five mold from 30 to 50 h. Results from smart tongue indicated that the best combinations of working electrodes with specific frequency segments for differentiation of five mold in liquid medium were as follows:the Ag electrode in 100 Hz and W in 100 Hz for distinguishing the cultures inoculating after 40 h; the Pt electrode in 100 Hz, Au in 1 Hz and 10 Hz, Pd in 1 Hz and 10 Hz, W in 10 Hz and 100 Hz, and the Ag electrode in 1 Hz and 100 Hz for these of 50 h.5. Intelligent models of Classfying the five common mycotoxin producing Aspergillus by Smart tongue and SIMCAComplying with SIMCA, smart tongue was studied for classifying and identifying the five common mycotoxin producing Aspergillus. The liquid cultures of mold culturing for 50h were detected by smart tongue and divided into training set and testing set. The best models based on the training set for classifying were detected. The discrimination accuracy of the best models were 100% for both the training sets and the testing sets as follows:The Pt and Au electrodes in 100 Hz, the Pd electrode in 10 Hz and 100 Hz, and W in 1 Hz for Af.;the Pt and Au electrodes in 100 Hz, and the W in all three frequency for Ap.; the Au electrode in 1 Hz and 100 Hz and W in three frequency segments for Afu.; the Au electrode in 1 Hz and 10 Hz, Pd in 1Hz for Av.; the Au electrode in 10 Hz and 100 Hz, and W in 100 Hz for Ao..