| Wheat is one of the main staple grains in my country.Due to its good storability,it is an important reserve grain,and its reserves rank first in my country.However,during wheat storage,mildew cause a huge loss of wheat resources.Therefore,the detection of wheat mildew is extremely important,which is conducive to reducing the loss of stored grain and effective implementation of mildew control measures.It is not easy to be noticed in the early stage of wheat mildew.It is necessary to not only detect wheat mildew,but also conduct early detection to achieve the purpose of early detection and early prevention.At present,the early detection of wheat mildew mainly depends on sensory evaluation methods and instrumental analysis methods,but these methods are highly professional,cumbersome,and time-consuming.They are not suitable for real-time in-situ early detection of wheat during storage.In summary,there is an urgent need to establish a mature early detection method for wheat mildew.In this study,wheat infected with Aspergillus glaucus,Aspergillus candida and Aspergillus flavus during storage was used as the research object.The colorimetric sensor combined with visible-near infrared spectroscopy technology and the nano-modified sensitization technology were employed to establish methods for identification of wheat mildew degree and quantitative detection of mold colonies.The details are as follows:(1)The detection and analysis of volatile gases from wheat mold infection.The experiment used headspace solid-phase microextraction technology combined with gas chromatography-mass spectrometry(HS-SPME-GC-MS)and plate counting method to detect wheat samples infected with Aspergillus glaucus,Aspergillus candida and Aspergillus flavus respectively.The volatile gases and number of mold colonies in the early mildew process was detected.Through the analysis of the correlation between the volatile gas with the number of mold colonies before the critical point of mildew,the results showed that the volatile gases gradually changed after mildew,mainly manifested in the increase of alcohols and hydrocarbons and the decrease of aldehydes;the wheat infected with different Aspergillus molds produced different volatiles during the process of wheat mildew;the relative percentage of 1-Octen-3-ol increased with the number of mold colonies.This phenomenon also occurred in wheat infected by three Aspergillus species.Therefore,1-Octen-3-ol was selected as a volatile marker reflecting the degree of mildew and the amount of mold.(2)Construction of nano-modified colorimetric sensor.Among the 30 chemically responsive dyes,5 chemically responsive dyes sensitively responded to 5 ppm volatile marker 1-Octen-3-ol,which were octaethylporphin OEP,tetraphenyl porphyrin TPP,tetraphenyl porphyrin B-TriP,8-(4-nitrophenyl)-4,4-difluoroboron dipyrrole methane NO2BDP and di[8-(4-Nitrophenyl)dipyrrole methane]nickel(NO2BDP)2Ni.Among them,the NO2BDP dye produced a more regular response to 1-Octen-3-ol in the concentration range of 1-100 ppm,and the difference in red,green,and blue(RGB)response decreased with increasing concentration.Then,using different surface modified porous silica nanospheres(PSN)to nano-modify the NO2BDP dye,nano-porous modified dye with high sensitivity and specificity was obtained.The RGB response to 1-Octen-3-of was about doubled,and the linear correlation of quantitative detection was strengthened,which realized the quantitative analysis of 1-100 ppm 1-Octen-3-ol in mixed volatile gases.(3)Early qualitative discrimination of wheat mildew based on colorimetric sensor array.Totally 23 chemically responsive dyes and 2 nano-modify dyes was used to establish a 5×5 colorimetric sensor array to detect 274 wheat samples.Using principal component analysis(PCA)and linear discriminant analysis(LDA)algorithms processed colorimetric sensor arrays to characterize different wheat samples to achieve qualitative discrimination of early wheat mildew.The PCA algorithm successfully distinguished the degree of mildew of most wheat samples,only some samples was close;the LDA algorithm successfully distinguished 100%wheat samples,and all samples had no overlap.It indicated that the colorimetric sensor array could achieve early qualitative discriminant analysis of the degree of wheat mildew(4)Quantitative analysis of wheat mold based on colorimetric sensing/spectroscopy technology.The experiment used chemically responsive dyes and nano-modify dyes sensitive to volatile markers to make colorimetric sensors to detect wheat samples carrying different numbers of molds.The visible-near infrared spectroscopy technique was used to capture the multi-dimensional spectral data of the colorimetric sensor.Partial least squares(PLS),joint interval partial least squares(Si-PLS)and genetic algorithm-joint interval partial least squares(Si-GA-PLS)algorithms were used to process the spectral data to establish quantitative prediction models.The quantitative prediction model established by the Si-GA-PLS algorithm had the best detection effect when predicting the number of three Aspergillus colonies.In the quantitative analysis of Aspergillus glaucus in wheat,training set Rc=0.9866,RMSECV=0.2893 1gCFU/g,prediction set Rp=0.9895,RMSEP=0.255 lgCFU/g;in the quantitative analysis of Aspergillus candida in wheat,training set Rc=0.979,RMSECV=0.367 lgCFU/g,prediction set Rp=0.9904,RMSEP=0.317 lgCFU/g;in quantitative analysis of Aspergillus flavus in wheat,training set Rc=0.9647,RMSECV=0.316 lgCFU/g,prediction set Rp=0.9821,RMSEP=0.302 lgCFU/g.The correlation coefficients of these quantitative prediction models were all above 0.95,and the mutual verification root mean square error RMSE were less than 0.5,proving that this method achieved rapid quantitative analysis of the number of mold colonies in the process of wheat mildew.This thesis analyzed the volatile gas produced by wheat infected with three kinds of Aspergillus and strengthened the correlation between wheat mildew and volatile gas.Using nanometer modification technology,a new type of nano-modified dye was synthesized,and the colorimetric sensor was successfully used to quantitatively analyze the early moldy volatile markers of wheat in the mixed gas.Using colorimetric sensor array technology,the early qualitative determination of the degree of wheat mildew was achieved before the wheat infected with three Aspergillus enters the critical point of mildew.Finally,combining colorimetric sensing with visible-near infrared spectroscopy technology,quantitative prediction models of the number of mold colonies in the process of wheat mildew was established.All in all,this paper established a rapid and non-destructive detection method of wheat mildew degree and mold colony number. |
