Application Of Laser-induced Breakdown Spectroscopy For Agricultural Products Analysis

The frequent occurrence of food safety incidents such as in fruit and vegetables has aggravated consumer concerns about the quality and safety of agricultural products.The traditional agricultural products analysis methods have the disadvantages of cumbersome preparation,long time-consuming,easy to introduce secondary pollution,and difficult to operate,which cannot meet the needs of modern agriculture with rapid,online and large-scale analysis.Laser-induced breakdown spectroscopy(LIBS)is an atomic spectrum analysis technology.It has the advantages of simple sample preparation,rapid analysis,convenient operation and non-contact detection,which has become a hotspot in the analysis of agricultural products.However,the inefficiency,low sensitivity and poor quantitative analysis precision make LIBS unable to meet actual requirements.It is the aim to conduct research deeply and find a suitable solution to the above issues in this thesis.The concrete research contents are as follows:Adulteration and inefficiency are very common in agricultural products.In this part,four different sample preparation methods and several chemometric algorithms were carried out to study the adulteration problem on rice geographic origins.20 kinds of rice samples from different geographic origins were used.The results showed that rice grain(RG)was the simplest and most efficient sample preparation method for distinguishing different geographical origins of agricultural products than the other methods,i.e.rice powder pellet with boric acid(RPPBA),rice powder pellet(RPP),and rice grain pellet(RGP).Classification accuracy and operation time were employed to evaluate the performances for different algorithm methods.LDA was found to be a more efficient tool for rapid,real-time and in-situ measurements due to the modeling efficiency,and SVM showed the best classification accurancy of 99.20%.In addition,the classifier was developed independently using Matlab 2015 b platform.Stability and sensitivity of toxic elements determination is still unsatisfactory in agricultural product using LIBS.A simple and low cost sample pretreatment method named ultrasonic-assisted extractions combined with solid-liquid-solid phase transformation(UE-SLST)method was proposed in this work.The target analytes of cadmium(Cd)and lead(Pb)from rice samples were prepared through ultrasound-assisted extraction in hydrochloric acid solution.Compare with conventional pellet method in rice analysis,the spectral intensity of Cd and Pb element were enhanced significantly using LIBS.The LODs of Cd and Pb were 0.0028 and 0.0437 mg/kg,which were accurate enough to satisfy the food quality standards for rice in China.Quantitative analysis of agricultural products using LIBS is usually suffering from matrix effects and nonlinear self-absorption.To overcome this problem,a hybrid quantitative analysis model of partial least squares-artificial neural network(PLS-ANN)was used to detect the compositions of agricultural products.Specifically,nine nutrient profiles(Mg,Fe,N,Al,B,Ca,K,Mn,P)were employed as the target elements for quantitative analysis.Fifty-eight plant materials were prepared to split into calibration,validation and prediction sets.Compared to the method of standard calibration,the prediction ability can be significantly improved using the PLS-ANN hybrid model.The mean prediction error(MPE)for Mg,Fe,N,Al,B,Ca,K,Mn and P analysis decreased from 5.66%,155.58%,137.08%,52.51%,57.19%,37.15%,9.12%,33.91% and 4.92% to 4.22%,13.29%,2.61%,19.02%,8.72%,12.37%,4.12%,7.91% and 3.87%,respectively.The law of element migration and transformation in agricultural products is not very clear.In order to obtain the law of element migration and transformation of roots,stem,and leaves of fruit and vegetables,laboratory samples of garlic seedlings and the soil,navel oranges collected in the field were taken as research objects.Garlic seedlings were planted at soil concentrations of 0,50,100,150,200,250,300 mg/kg and Cd at concentrations of 0,0.2,0.4,0.8,1,5,and 10 mg/kg.When the concentration of Pb was less than 200 mg/kg or Cd was less than 1 mg/kg,heavy metals promoted the growth of garlic seedlings,but there was no significant correlation with the concentration.When the concentration of Pb was more than 200 mg/kg or Cd was more than 1 mg/kg,heavy metals inhibited the growth of garlic.The absorptions of Fe,K and Mn were inhibited by Pb.Meanwhile,the absorptions of Zn,P and Fe were inhibited by Cd.Through the study of elemental absorption,the growth of navel orange can be effectively monitored in real time.Huanglongbing has an inhibitory effect on the absorption of navel orange.In order to improve the detection efficiency,LIBS coupled with SVM algorithms was used to distinguish healthy navel oranges and Huanglongbing-infected navel oranges.The classification rates were 100% and 92%,respectively.Compared with the traditional detection method,the detection efficiency of LIBS technology is significantly better than the polymerase chain reaction method,which provides a new means for the diagnosis of early Huanglongbing in navel orange.In summary,various methods combined with LIBS,such as sample preparation methods,statistical algorithms,UE-SLST method,and PLS-ANN hybrid model were carried out in this thesis.The accuracy and detection sensitivity of LIBS in the analysis of agricultural products were significantly improved through introduction the proposed methods.The studies on the law of migration and transformation of agricultural elements by LIBS provided a method and basis for exploring the correlation between element-plant growth and the diagnosis of Huanglongbing in navel orange.These researches will lay a foundation for promoting the application of LIBS in agriculture.