Quantitative Acidity Analysis In Iron Ore Based On Calibration-Free Method And LIBS Technique

Iron ore is the basic raw material in metallurgical industry.The determination of its acidity is benefit to improve the utilization of raw materials and the quality of metallurgical products.Traditional analytical techniques have disadvantages of longer analysis time and complex sample pretreatment,thus a analytical technique with online and rapid advantages for acidity analysis of iron ore is desired.Laser-induced breakdown spectroscopy,(LIBS)is a new spectral analysis technique,which is based on atomic emission and have the advantages,such as no complex sample pretreatment,rapid analysis and remote detection advantages.Thus,its is widely used in metallurgical process and analysis.Calibration free is a newly quantitative analysis technique,which have the advantages such as without standard calibration and reducing the influence of matrix effect to some extent,but the accuracy of quantitative analysis is affected by many factors.This thesis investigated the main factors of CF-LIBS quantitative analysis and developed the combination of CF-LIBS and chemometrics to improve the quantitative accuracy of acidity analysis of iron ore.This study can provide a new thought and a new way for the rapid screening of metallurgical materials and process control.This thesis is divided into four chapters.The main contents are as follows:Firstly,comparative study on acidity analysis of iron ore based on the CF and PLSR technology was carried out.Iron ore powder was used to make tabletting and 20 LIBS spectrum was obtained.The optimized plasma temperature and calibration standard sample was searched by binary search algorithm(BSA)and principle of minimum relative error,and the parameter of the optimized calibration sample was used to calibration the other parameter of unknown sample.The acidity of the unknown samples were calculated by the parameter of the optimized calibration sample.In PLSR,the selected characteristic lines in CF was used for input variable,and then different PLSR models based on different input variable were constructed before correction and after correction.The model parameter in PLSR was optimized by 5 fold cross validation,the acidity in unknown sample is calculated by PLSR models based on different input variable.The obtained result showed that CF can get good performance with small root-mean-square error and good correlation coefficient.Secondly,method was developed for the acidity analysis of iron ore by the combination of CF with internal standard method.Firstly,the calibration curves is obtained by using internal element for different analyzed particles.The intensity ration can be obtained from the slope in the calibration curves.The concentration of the analyzed element in unknown sample are obtained by using the intensity ration and the concentration of the calibration element in the standard sample.Finally,the acidity of iron ore was calculated.The result showed that the relative error of the acidity is reduced greatly.Compared with traditional CF method,this method can overcome matrix effect.Finally,the acidity analytical method is established by the combination of CF with particle swarm optimization(PSO)technology.A weighted score as an evaluation index was defined to select analytical lines as the reference line with the small degree of self-absorption to correct analytical lines with self-absorption.The plasma temperature is optimized by using PSO technology.The acidity is calculated in unknown samples by using optimized plasma temperature and corrected intensity.The result showed that the calculated acidity by using the CF-LIBS method was close to the true value.This approach obtain accurate plasma temperature by using PSO technology,thus the accuracy of the quantitative acidity is improved.