Weighting inversion of dynamic light scattering based on particle-size information distribution character

Abstract

In particle sizing with dynamic light scattering (DLS) technique, the determination of particle size distribution (PSD), via inversing the autocorrelation function (ACF) of scattering light, is usually limited by the inherently low particle size information in ACF data and, the lack of targeted inversion on the noise restriction and the particle size information utilization. For the ACF data in DLS measurement, most of particle size information is centrally contained in the decay section and the larger noise is contained in the larger delay section. However, no consideration of the particle size information distribution in the ACF data for the routine inversion method increases the difficulty of the accurate PSD inversion, especially the broad and bimodal PSDs. Until now, it is still a difficult problem to obtain an accurate recovery of the broad and bimodal PSDs, specifically the bimodal PSD with a peak position ratio less than 2:1 and containing large particles (350 nm). In this paper, a character-weighted constrained regularization (CW-CR) method is proposed, in which, the particle size information distribution in the ACF as the base and the adjustment parameter as the exponent are used to weight the ACF. By using the weighting coefficients corresponding to the particle size information distribution along the delay time in ACF, the CW-CR method can enhance the utilization of the particle size information in ACF data, and effectively weaken the effect of noise at large delay time. With this method, the closely spaced bimodal PSD (with nominal diameters of m 350 nm:500 nm in simulation, m 300 nm:502 nm in experiment) is recovered successfully at a high noise level of 0.01. It shows that the CW-CR method, combined with the multiangle DLS (MDLS) measurement, can effectively make the best use of the particle size information hiding in the noisy ACF data, and improve the resolution of bimodal PSD as well as the capability of noise suppression. So it can make the advantages of MDLS more highlighted than the routine method in the recovery of the broad and bimodal PSDs.