Research On The Detection And Identification Of Flour Additives Based On Raman Hyperspectral Imaging Technique

Currently,detection of additives in flour is mainly carried out by chromatography,but there are many disadvantages,such as complex pretreatment,long detection time and high cost,and are destructive to samples.Spectroscopy has shown good potential for application in the detection of flour additives.However,spatial range covered by single-point detection cannot meet the detection requirements of the whole sample,and can not be visually identified to obtain the spatial distribution information.Hence,Raman hyperspectral imaging technology with line scanning,imaging visualization and non-destructive detection was used in this research to detect and identify additives in flour.Ascorbic acid and azodicarbonamide were useded as the object of this research.The results of this work can provide a technical support to ensure the quality and safety of flour,protect the legitimate rights and interests of consumers,and also have important significance for standardizing the market order.The main research contents and conclusions are as follows:?1?Acquisition parameters of Raman hyperspectral imaging system suitable for this study were explored.By analyzing the Raman hyperspectral images of flour and ascorbic acid collected under different parameters,the acquisition parameters of the system were determined.The best Raman hyperspectral image can be obtained when the line laser output power is 10 W,the exposure time is 1000 ms,the spectral resolution is 0.54 nm and the spatial resolution is0.125 mm/pixel.The determination of collecting parameters provides the standard experimental conditions for the detection and identification of the additives in flour.?2?Raman spectra and images of the additives?ascorbic acid and azodicarbonamide?were analyzed.Raman hyperspectral images of the additives were obtained with the determined acquisition parameters.The vibrational modes of additives Raman peaks were assigned and their Raman characteristic peaks were found.Raman characteristic peaks of ascorbic acid were1658,1321,1258,1128,824 and 631 cm^-1,and Raman characteristic peaks of azodicarbonamide were 1335,1577 and 1121 cm^-1.Gray images corresponding to the Raman characteristic peaks are selected from the hyperspectral image to find the relationship with the Raman intensity.The results show that the brightness changes of the gray scale images are consistent with the intensity changes of the Raman peaks.This work provides a theoretical basis for the detection of additives in flour by using Raman hyperspectral technology.?3?Effective penetration depth of the line laser on different gluten flour?high,middle and low gluten?was studied.Two-layer samples with different thickness of flour layer on ascorbic acid layer were prepared.Raman characteristic peak method and spectral similarity method was used to analyze the data,and the penetration rate was used as the evaluation index to determine the effect of the line laser on different thicknesses of flour layer.The results showed that the penetration rate of the line laser is more than 99%when the thickness of the flour layer is 2mm.It can be considered that the line laser penetrates the entire flour layer.The difference of gluten degree?protein content?does not have a significant effect on the penetration of the line laser.The same brand of flour repeated three times and three different brands of flour on the results of the verification showed that the effective penetration depth of 2 mm is accurate and reliable.The results have laid a foundation for the effective detection of the additives in the subsequent mixture samples.?4?The detection and identification of single additive in flour was studied,and the quantitative analysis model was established.Different concentrations of flour-ascorbic acid and flour-azodicarbonamide mixture samples were prepared.Raman characteristic peak method?univariate method?and partial least squares regression coefficient method?multivariate method?were used to detect and identify the additives in the flour,and a linear relationship model between additive concentrations and the identification results was established.The results showed that the minimum detectable concentrations of ascorbic acid were both 0.01%by the univariate and multivariate methods,and the determination coefficient of the quantitative analysis model were 0.9861 and 0.9882,respectively.Minimum detectable concentrations of azodicarbonamide were both 0.01%by the univariate and multivariate methods,and the determination coefficient of the quantitative analysis model were 0.9873 and 0.9876,respectively.The identification results of univariate and multivariate methods are basically the same at the lower concentration.Data processing of the multivariate method is more complicated and time-consuming,and the univariate method is more convenient,which is beneficial to rapid analysis.?5?The detection and identification of various additives in flour were studied,and the quantitative analysis models were established.Mixture samples containing two additives in flour were ascorbic acid-benzoyl peroxide and azodicarbonamide-benzoyl peroxide.Mixture samples containing three additives in flour were ascorbic acid-azodicarbonamide-benzoyl peroxide.Significant Raman characteristic peaks that can represent each additive and have no influence on each other are determined.Raman characteristic peak method is used for the detection and identification of each additive,and they are combined into a chemical image that can identify various additives in the flour.A linear relationship model between the percentage of each additive pixel and its actual concentration was established.The results showed that the determination coefficients of the quantitative analysis models of ascorbic acid-benzoyl peroxide,azodicarbonamide-benzoyl peroxide in flour were 0.9847 and 0.9858,0.9865 and 0.9857,respectively.The determination coefficients of the quantitative analysis models of ascorbic acid,azodicarbonamide in flour were 0.9858,0.9868 and 0.9830 respectively.The quantitative results of various additives and the previous study of the same type of additives are basically the same,and the detection of different additives basically unaffected.