Research And Preliminary Application To Fundamental Parameters Method Of Energy Dispersive X-ray Fluorescence Analysis

The mathematical correction model of fundamental parameters method is established by using Si-PIN detector and the portable high energy resolution X-ray fluorescence (EDXRF) analyzer with an embedded computer, and correct the absorption and enhancement effect of matrix elements with this method.First of all, this paper gives a review of the development and current situation of fundamental parameters method of the energy dispersive X-ray fluorescence analysis and introduces theory and experience fitting formula such as important physical constants and parameters of a higher degree of accuracy and fluorescence intensity used in the fundamental parameters method at present. After having done thorough research to the algorithms, we design and realize the calculation program of fundamental parameters method. Through preliminary application, results are as follows:(1)Because the excitation factor is related to fluorescence yield, relative emission rate and absorption jump factor and its source code of calculation program is complex, it is necessary to do some research to the excitation factor. Based on previous theory and a large number of experimental data, new research into fitting is done in this subject. The calculation formula of the excitation factor is put forward in this subject and the accurate degree has met the actual needs.(2)According to the actual calculation method of the fundamental parameters method and the calculation program of the fundamental parameters method based on C language, the calculation of values of parameters such as the first (primary) fluorescence intensity, the second (subordination) fluorescence intensity, the total relative fluorescence intensity, the mass attenuation coefficient, excitation factor, fluorescence yield, relative emission rate, the absorption jump factor, characteristic X-ray energy, critical stimulate energy could be achieved.(3)The quantitative analysis to the alloy samples prepared, some geological samples including copper, plumbum and zinc ore samples, and chromium and nickel ore samples was done with the calculation procedure made by myself.For the alloy samples prepared: the absolute value of maximal relative error of the content of the element Mn with the content below the 6% reaches 25.3% and it is not more than 10.5% with the content between 6% to 17%; the absolute value of maximal relative error of the content of the element Fe with the content below the 6% reaches 29.9% and it is not more than 11.2% with the content between 6% to 26%;the absolute value of maximal relative error of the content of the element Co with the content below the 6% reaches 12.1% and it is not more than 10.9% with the content between 6% to 13% and not more than 5.8% with the content between 13% to 30%;the absolute value of maximal relative error of the content of the element Ni with the content below the 11.01% reaches 12.6% and it is not more than 5.95% with the content between 11.01% to 44.32%;the absolute value of maximal relative error of the content of the element Cu with the content below the 13.45% reachesl3.9% and it is not more than 5.7% with the content between 13.45% to 20.12%;the absolute value of maximal relative error of the content of the element Zn is not more than 3.63% with the content between 17.16% to 54.79%.For geological samples, the absolute value of maximal relative error of the content of the element Cu with the content below 500μg/g reaches 56.44% and it is not more than 48.63% with the content between 500μg/g to 2000μg/g;the absolute value of maximal relative error of the content of the element Zn with the content below the 5300μg/g reaches 22.87% and it is not more than 19.96% with the content between 5300μg/g to 26600μg/g;the absolute value of maximal relative error of the content of the element Pb with the content below the 8700μg/g reaches 31.9% and it is not more than 13.96% with the content between 8700μg/g to 81500μg/g;the absolute value of maximal relative error of the content of the element Cr with the content below the 545μg/g reaches 20.04% and it is not more than 8.64% with the content between 545μg/g to 1507μg/g;the absolute value of maximal relative error of the content of the element Ni with the content below the 488μg/g reaches 54.9% and it is not more than 10.09% with the content between 488μg/g to 1090μg/g. We can conclude that the relative error of the high content is smaller, and the relative error of the low content is higher. However, its accuracy is basically close to the results of chemical analysis. So the calculation program is precise, and the method basically meets the needs of production.(4)In order to investigate the reproducibility of measurements, the precision experiment is done. That is to say, we analyze several types of standard geological mineral samples as well as a standard alloy sample prepared with this method. Under the same condition, we did 10 parallel tests to the same sample with this method and result is that its precision is higher, the maximal relative standard deviation of the element Ni is 5.78%, the greatest relative standard deviation of other elements is not more than 2.58 % and the stability is perfect.The research into this algorithm has achieved the desired aim. By analyzing alloy samples prepared and a number of mineral samples, the results show that its accuracy and precision can meet the on-line analysis of the material and finished product. This method could overcome the influence of absorption and enhance effect from base material components commendably and obtains the preliminary result of this application.