Preliminary Application Of Laser Induced Breakdown Spectroscopy Combined With Chemometrics In The Analysis Of Urban Atmospheric Particulate Pollution Sources

Atmospheric particulate matter is the most complex pollutant and the most harmful to humans in the atmosphere.The accumulation of metal elements in the atmosphere for a long time can cause serious damage to water,plants and human health.Rapid analysis of metal elements in atmospheric particulates can facilitate real-time monitoring and effective prevention of air pollution.Laser Induced Breakdown Spectroscopy(LIBS)is an emerging elemental analysis technique based on atomic emission spectroscopy.It has many advantages such as no complicated sample preparation,simultaneous multi-element analysis and remote detection and has a broad application prospect in environmental analysis.Due to the difference and complexity of the chemical composition of atmospheric particulate matter,a large number of complex LIBS spectra are generated.By using chemometrics,the effective feature information can be extracted from the massive spectral data,and the accuracy of the qualitative and quantitative analysis can be improved.In this paper,we researched the quantitative analysis of heavy metals and the preliminary discrimination of pollution sources in urban atmospheric sedimentation by the combination of LIBS technology and chemometrics.And this study can provide theoretical basis and technical support for accurate traceability and comprehensive prevention measures of composite atmospheric pollutants.The paper is divided into four chapters,the main research contents as follows:Firstly,the preliminary analysis of atmospheric pollution sources is the research purpose.The SEM-EDS is used to study the microscopic morphology of atmospheric particulate matter and the pollution source analysis,and the ICP-OES is used to study the quantitative analysis of 10 metal elements in atmospheric sedimentation.According to the different content of metal elements,a method for identifying pollution sources of atmospheric particulate matter has been established.Secondly,taking atmospheric sedimentation as the research object,a rapid quantitative analysis method based on LIBS technology combined with RF was established for six metal elements(Pb,Cu,Zn,Fe,Al and Ca)analysis in the atmospheric sedimentation.The LIBS spectra of the atmospheric sedimentation samples were collected.And the effects of different pretreatment methods on the RF calibration model prediction performance were investigated,and the input variables were optimized and selected based on the variable importance.Under the optimal pretreatment method and input variables,the PLS,LSSVM and RF calibration models for six metal elements of atmospheric sedimentation were constructed respectively,and the prediction performance of the three calibration models were compared.The results show that RF can obtain satisfactory quantitative analysis results compared with PLS and LSSVM methods.Thirdly,the preliminary discriminant analysis of pollution sources in different regions is taken as the research purpose,and the preliminary discrimination analysis method of pollution sources in different regions based on LIBS technology combined with RF is established.The effects of different preprocessing methods on the RF discriminant model performance were investigated.And the optimal pre-processed LIBS spectral data was used as input variables to construct PLS-DA,LSSVM and RF discriminant models respectively,and the prediction performance of the three discriminant models were compared.The results show that compared with the PLS-DA and LSSVM methods,RF can obtain satisfactory classification results,and it is an effective method for rapid analysis of atmospheric pollution sources.