Elemental Analysis In Chinese Dry Red Wine Based On Laser Induced Breakdown Spectroscopy

Detection of elements in wine is of great significance for production and circulation.The concentration of many elements in wine is stable in the finished wine,so it can not only reveal the brewing technology and quality level of wine,but also can be used as an important marker of wine origin traceability.Among them,copper and iron have the detection limits specified in the national standard,which are closely related to the health of consumers.However,the current chemical detection methods usually have some problems,such as complex sample pretreatment,unable to detect multiple elements at the same time,high cost of large-scale instrument detection methods and high operation requirements.Laser induced breakdown spectroscopy(LIBS)is an element analysis technology based on laser-induced plasma emission spectroscopy.It has the advantages of simple pretreatment,simultaneous detection of multiple elements,simple equipment and operation,and fast detection speed.It has great application potential in rapid detection of multiple elements in wine.This paper studies the detection methods of LIBS technology in wine elements in the following three aspects,in order to provide technical guidance and theoretical basis for LIBS technology in wine industry(1)The quantitative elements were screened through pre experiment,and Zn was selected as the target element of dry red wine detection method optimization.Combined with the National Institute of standards and Technology(NIST)database,330.3nm was selected as the characteristic wavelength of Zn.Six different pretreatment schemes were compared by signal-to-noise ratio(SNR)and relative standard deviation(RSD).The LIBS pretreatment scheme of dry red wine sample by hot surface deposition(HSD)was determined,that is,40 ml wine sample was placed in a constant temperature blast drying oven,maintained at 60?for 2 h to remove most of the moisture,and then adjusted to105?for 1 h to dry to constant weight.The settlement is completed on the surface of polycarbonate to form a settlement layer with a thickness of about 1 mm.The instrument parameters of LIBS are determined as follows:laser wavelength 1064nm,repetition rate10Hz,pulse width 10ns.The spectrum acquisition delay is set to 2?s.Exposure time 1?s.The laser energy is 100 m J and the spot diameter is 100?m.20 hits.On the surface of sample 5×Every 25 sampling points get a spectrum,and each spectrum is the sum of 500laser pulses.By using the pretreatment method and optimized instrument parameters,17elements in wine were identified qualitatively,and the maximum absorbance wavelength and signal-to-noise ratio were determined,which provided strong support for the data results of subsequent experiments.(2)In order to establish a universal method for the detection of Fe and Cu in wine and reduce the influence of matrix effect on the quantitative results of different wine types,the corresponding calibration curves of Fe and Cu elements in wine were established by single variable quantitative analysis method,and the influence of spectral noise and matrix effect on the quantitative results of single variable of Fe and Cu elements was analyzed;Then,the internal standard method and two multivariate analysis methods,pls and SVM,were used to establish the universal calibration curve of dry red wine,in order to reduce the decrease of accuracy and precision caused by matrix effect and experimental fluctuation,complete the correction of matrix effect,and optimize the quantitative analysis results.The results show that SVM can effectively reduce the negative effects of spectral noise and matrix effect on the detection of dry red wine.Compared with the national standard method,the average relative errors of Fe and Cu are 5.902%and 6.164%,and the relative standard deviations are 5.869%and 4.902%.In the quantitative analysis of Fe and Cu elements,the general calibration curve R~2established by SVM method is more than 0.999,which has good quantitative analysis ability.It is proved that the method of multiple analysis is effective and superior to deal with LIBS spectrum of dry red wine and can be used for quantitative analysis accurately.(3)In order to establish a wine origin traceability model based on LIBS technology,135 wines from 8 different regions in China were analyzed by LIBS.The experimental results show that PCA analysis can not effectively separate the origin of samples;After extracting the characteristic spectral lines of 17 elements,PCA analysis was carried out.The model can separate the samples from the northwest and Shanxi regions from other regions,and the recognition accuracy was 87.93%and 100%respectively.Based on the extracted characteristic spectral lines,the number of characteristic elements was reduced to9 to reduce the influence of exogenous factors.The separation ability of the optimized model has been greatly improved.The recognition accuracy of samples from Shandong and Hebei reaches 62.96%and 66.66%respectively,except for the northwest and Shanxi areas,which are 89.66%and 100%;The accuracy of sample recognition in Yunnan Province is69.23%.The preliminary exploration of LIBS domestic dry red wine modeling method was completed,which provided support for further research and wider application of LIBS domestic dry red wine modeling method.This study is the first time to apply LIBS technology to the detection of elements in domestic dry red wine,and deeply explore the pretreatment methods and instrument parameters.The feasibility of LIBS technology in qualitative and quantitative experiments of dry red wine was explored.The quantitative identification models of Cu and Fe and the qualitative identification models of different wine producing areas in China were established by multivariate analysis.