| Energy Dispersive X-ray Fluorescence(EDXRF),a commonly used rapid determination and nondestructive elemental analysis technique, is relatively fast, simple, inexpensive, it can measure various kinds of material elements without destroy the sample composition and the measurement results are stable and reliable and easy to implement on-line analysis and automatic measurement compared with other methods.In EDXRF analysis, the relationship between the exact concentration of the target element and spectral intensity obtained by measuring is not linear, the main interfering factor is matrix effects, Therefore, matrix effects correction has become a key issue.Since the 1950 s, influence coefficient method, fundamental parameter method and influence coefficient method and fundamental parameter of the combination are three main methods to solve the problem of matrix effect. Influence coefficient method use standard material for instrument calibration, establishing calibration equation to solve the peak area of interaction between elements problem. This method has not clear physical meaning, just use mathematical models to solve mutual influence. Fundamental parameter method based on physical parameters, eliminating the matrix effect, is a matrix correction method which been more concerned about.Currently NLXRF, MICRO-XRF and many other fundamental parameter method software are widely used. But most fundamental parameter method analysis developed for wavelength dispersive X-ray fluorescence(WDXRF), energy dispersive X-ray fluorescence analysis is not very applicable. First, energy dispersive X-ray fluorescence analysis mostly for field site work began measuring element from K,but not concern Si, Al and other macroelements, so there is "dark matter",For that case, the fundamental parameter method based on the scattering is presented and applied. Second, the fundamental parameter method theoretical formula and iterative process, the need to obtain the net peak area of the target element, but the energy dispersive X-ray fluorescence analysis of the energy resolution limitations, to obtain the net peak area is difficult, thus limiting the fundamental parameter method in application of energy dispersive X-ray fluorescence field.In the spectral analytical methods, generally workflow is first to deduct scattering background, and then the separation of overlapping peaks, and finally solve the net peak area. To solve the instrument spectral resolution problems, first of all high energy resolution X-ray detector is applied in the field of the energy dispersive X-ray fluorescence analysis, The most commonly used detectors are Si(Li) detector, electric refrigeration Si-PIN detector, electric refrigeration SDD detector, energy resolution at about 150 e V. In actual measurement instrument, the energy resolution at about 200 e V. At present, the best energy resolution of superconducting detector(energy resolution of about 10 e V) currently can not meet the need of field applications, just as physics experiments with high precision measurement of energy. Therefore, to solve the instrument spectral problem, many algorithms have been proposed, such as for background subtraction algorithm of SNIP, Fourier transform iteration, curve fitting, etc. For peeling problem of overlapping peaks, the most commonly used direct demodulation method, maximum likelihood estimation, Gaussian fitting and so on. Above algorithms to separate overlapping peaks have demanding applications:1ã€Required instrument characteristic X-ray spectrum in line with the almighty peak Gaussian distribution can be described by a simple mathematical function. But in geology and mineral resources and pollutants monitoring, because the measuring target element content is low, the measurement time is limited, the lower characteristic peak count rate peaks is not obvious, and the type and class of Gaussian peak far.2ã€Require neighboring energy difference is greater than a width at half or even more. In the actual energy dispersive X-ray fluorescence analysis, the characteristic energy interfering peak is less than a half-width, for example Fe.Kb and Co Ka-ray radiation, the energy difference 117 e V, less than half wide. Therefore, the above method can not completely solve the problem of X-ray fluorescence spectrum release. For this case, the release factor method is widely used, but the composition of the release factor is also related products.This paper is based on fundamental parameter method, combining the fundamental parameter method and overlapping peaks stripping process to form the content of calculation and quantitative analysis of spectral analytical integration, improve energy dispersive X-ray fluorescence analysis of quantitative precision low levels of the target elements in geological and environmental samples, the expansion of application fields energy X-ray fluorescence analysis.The main research contents and results are as follows:(1) Theoretical study of single standard FP methodAccording to the principles of FP method derived formulas single standard sample FP method, and the way to get each parameter of the formula. By making a number of single-element standard samples can be realized on energy dispersive X-ray fluorescence spectrometer for elemental content of scale, reducing the instrument calibration requirements for standard samples.(2) Monte Carlo simulation the primary level X-ray tube output spectrumUse Monte Carlo simulation technique to original stage output X-ray spectrum of the X-ray tube was simulated to obtain the primary X-ray spectral parameters EDXRF analysis excitation source for the realization of FP method provides parameters guarantee.(3) Improved FP method iterative method to achieveCombining FP method and elements characteristic X-ray spectrum release treatment of overlapping peaks together, proposed a new iterative algorithm, and the development of software modules, to achieve real-time processing and spectral resolution FP method.(4) Application of improved single standard FP methodAccording to the measurement of 16 national level geochemical reference materials for the measurement object verification showed improved single standard FP method can achieve multi-element complex quantitative analysis of geological samples. For Cu element content from 11.4μg / g to 390μg / g, the average element content and the recommended value Improved FP method to obtain the relative error is 22.05%. The influence coefficient method was 96%; Zn element for average element content and the recommended value of content from 31μg / g to 680μg / g, modified FP method to obtain the relative error is 15.14%, while the influence coefficient method was 34.77%. |