Study Of Photon Correlation Spectral Inversion Algorithm For Multi-Modal Particle Suspension

Photon correlation spectroscopy is a common method for measuring the particle size of submicron and nanoparticles in suspensions.In the process of particle size distribution inversion,it is necessary to solve the Fredholm integral equation of the first type,which is pathological and a typical undefined problem,and it is even more difficult when the measured sample is a multi-modal particle system.Therefore,in this paper,the research is carried out for the Sequential Extraction of Late Exponentials method,and the inversion of multi-modal particle system is transformed into the inversion of multiple single-modal particle systems.The main research contents include:1.Study on the optimal selection of data fitting windows for correlation functions in bimodal particle size inversion.The selection of the correlation function fitting window in the traditional Sequential Extraction of Late Exponentials method is the key to the bimodal particle size inversion,and the inversion results are different for different fitting windows.In this paper,we propose an improved algorithms named Attenuation Characteristics Sequential Extraction of Late Exponentials based on the relative attenuation characteristics of the correlation function of bimodal particle samples.The method defines the starting point of the fitting window for the correlation function of small particles,the interval point and the termination point of the fitting window for the correlation function of large particles,respectively.These three reference points are used as the selection criteria of the correlation function fitting window,and the fitting window is optimized to reduce the blindness of window selection,thus improving the accuracy of the particle size inversion results.2.Study on the optimal selection of initial values for the multi-modal particle correlation function of Sequential Extraction of Late Exponentials method.The inversion results of the multi-modal particle Sequential Extraction of Late Exponentials method are influenced not only by the fitting window of the correlation function but also by the initial values of the correlation curve fitting.The attenuation line width of the autocorrelation function and its distribution coefficient relational formula in the multi-exponential algorithm model,are combined with the traditional Sequential Extraction of Late Exponentials method for initial values optimization,the Multi-exponential Algorithm Sequential Extraction of Late Exponentials method based on the multi-exponential algorithm is proposed,so as to solve the problem of the dependence of the initial values on the fitting window of the correlation function.3.Study on the effect of random noise on the inversion results of Multi-exponential Algorithm Sequential Extraction of Late Exponentials method.In the traditional Sequential Extraction of Late Exponentials method,the initial values of the particle correlation function needs to be generated first according to the power expansion,however,since the power form of the multi-modal particle correlation function is determined by the fitting window of the correlation function data,the initial values also depends on the choice of the fitting window,and if the correlation function fitting window is selected in the noisy correlation function part,the inversion results are easily disturbed by the noise.For the proposed Multi-exponential Algorithm Sequential Extraction of Late Exponentials method,because of its initial values and the correlation function fitting window of the uncorrelated,to a certain extent to reduce the adverse effects of random noise on the inversion results of the multi-peak particle system.