| As people’s health awareness continues to grow,food quality and safety issues are of great concern to consumers.Unscrupulous merchants or small food processing plants in order to make huge profits,will be high-priced food using poor-quality food substitution or in the food to add illegal additives for sale,and such adulteration problems often occur,seriously endangering the lives and health of consumers and disrupting the normal order of the food industry.There is no lack of trace components in food that are beneficial to human body,such as bioactive substances,etc.Conventional testing methods for the detection of bioactive substances,food additives and non-edible substances in food such as trace components are time-consuming and labor-intensive,expensive,complicated pre-treatment,and can not be achieved online testing and other deficiencies.In response to the current food safety situation and market demand,a highly sensitive,rapid and non-destructive detection method is needed for the detection of food trace components.Surface-enhanced infrared absorption spectroscopy(SEIRA)can reduce the detection limit of conventional infrared spectroscopy,and can also compensate for the shortcomings of some conventional detection methods.Therefore,in this study,glutathione,resveratrol and melamine in food were used as the detection targets,and surface-enhanced infrared absorption spectroscopy was combined with chemometric methods to achieve the efficient and rapid detection of trace components in food.The main research contents are as follows:1.Study on the SEIRA detection method of glutathione in dairy products.The study proposed a SEIRAS detection method based on silver nanosol,taking glutathione as the research object,firstly,silver nanosols(AgNPs)were prepared as enhanced substrates for SEIRAS and characterized by transmission electron microscopy and X-ray photoelectron spectroscopy.The SEIRA detection system was used to detect glutathione in pure milk and goat milk by SEIRA spectroscopy,and the anti-interference performance of glutathione and the stability of the enhanced substrate performance were evaluated.This study showed that glutathione was detected in both pure cow’s milk and pure goat’s milk within the detection interval of 0.02～0.12 mg/mL,and the predicted correlation coefficients(R2)were 0.987 9 and 0.983 3 for glutathione content,respectively.Meanwhile,the spiked recoveries for glutathione detection in actual pure milk and pure goat milk samples with standard addition were 101.3%and 92.5%.Therefore,this method provides a new way for the detection of glutathione in dairy products.2.The analysis of peanut oil adulteration and the study of SEIRA method for the determination of resveratrol.The application of infrared spectroscopy was studied for the detection of peanut oil adulteration problem.Firstly,peanut oil was used as the base oil and four edible vegetable oils,soybean oil,rice oil,rapeseed oil and palm oil,were used as the adulterant oil to build a binary mixture system;the influence of different spectral pretreatment methods on the spectral modeling was studied,and the attribution of the infrared characteristic absorption peaks was analyzed;the influence of different spectral characteristic bands on the spectral modeling was studied,and the results of each accurate peanut oil adulteration prediction model were compared and analyzed.A comparative analysis was conducted.The results showed that the prediction models constructed by using standard normal variational transformation processing combined with PLS-ANN could accurately predict peanut oil adulteration,with the selected characteristic wave numbers in the range of 1 900～700 cm-1,the corrected correlation coefficients RC and predicted correlation coefficients RP of 0.993 3 and 0.975 4,the corrected root mean square error RMSEC and predicted root mean square RMSEP of 0.016 7 and 0.033 6,respectively.The relative analytical error RPD was 8.460 2.The study proved that the PLS-ANN peanut oil adulteration model constructed by this method has good prediction accuracy.The SEIRA method combined with chemometrics was applied for the detection of resveratrol in peanut oil.Firstly,the SEIRA detection system was constructed by combining different concentrations of resveratrol in peanut oil with AgNPs-enhanced substrates;the effects of different pretreatment methods and spectral feature bands on the spectral modeling were investigated separately;the prediction results of each model established were compared and analyzed.The results showed that the pretreatment method of SG smoothing and standard normal variable transformation combined with the prediction model constructed by PLS-ANN could quantitatively predict resveratrol in peanut oil,with the corrected correlation coefficient RC and predicted correlation coefficient RP of 0.991 9 and 0.959 1,respectively,and the corrected root mean square error RMSEC and predicted root mean square RMSEP of 0.034 3 and 0.081 4,respectively,and the relative analytical error RPD was 7.873 4,and the average spiked recovery was calculated to be 114.3%.Therefore,the method is feasible for the determination of resveratrol in peanut oil.3.The electric field enhancement based on COMSOL simulation of equipartite excitation metal nano.The resonance characteristics of the local surface iso-excited elements of cylindrical and antenna-shaped gold nanoparticles and their dimers with one-dimensional periodic arrays were studied by using the discrete point dipole approximation method,and the variation of their extinction spectra with particle shape,size and spacing was analyzed;the variation of the maximum value of the electric field intensity with wavelength for the same geometric array type at different spacing of gold nanoparticles was calculated by using COMSOL software;the variation of the electric field intensity with wavelength for the same geometric array type was simulated.The electric field variation of the gold nano-antenna array is simulated,and the optical simulation and analysis of the equilibrium excitation metal nanoparticles are realized,which provides the theoretical and modeling basis for the subsequent optical analysis and optimization of the metal nanostructures.4.The study of NIR spectral enhancement method for the determination of melamine in milk.In this chapter,the application of surface-enhanced near-infrared spectroscopy(SENIRA)for the detection of melamine in milk was investigated.Different concentrations of melamine in milk were combined with AuNPs-enhanced substrates to construct the SENIRA detection system,and the effects of different pretreatment methods on the spectral modeling were studied;finally,the prediction results of the established model were compared and analyzed.The results showed that the prediction model constructed by the normalization method combined with PLS could quantitatively predict melamine in milk,with the corrected correlation coefficient RC and predicted correlation coefficient RP of 0.985 3 and 0.983 7,the corrected root mean square error RMSEC and predicted root mean square RMSEP of 0.005 9 and 0.006 6,respectively,and the relative analytical error RPD of 5.394 0,and the mean spiked recoveries were calculated to be 117%.Therefore,the method is feasible to be applied to the detection of melamine in milk. |
