Sizing Nanoparticles from Dynamic Light Scattering Measurements: I. A Bayesian Approach for Solving the Inverse Problem

CLEMENTI, L. A.; VEGA, J.R.; ORLANDE, H. R. B.; GUGLIOTTA, L.M.

Joao Pessoa

Simposio; Inverse Problems, Design and Optimization Symposium; 2010

A Bayesian method is used for solving the non-linear inverse problem of estimating the particle size distribution (PSD) in a colloidal system from multiangle dynamic light scattering (MDLS) measurements. The Markov Chain Monte Carlo sampling approach, implemented in the form of the Metropolis-Hasting algorithm, is used for solving the inverse problem. The method is tested through several simulated examples corresponding to polystyrene (PS) latexes of different PSD shapes, diameter ranges, and morphologies; and then experimentally validated through some PS samples (also characterized through independent techniques). For comparison, all examples are also solved through a non-linear Tikhonov regularization technique. For PSDs exhibiting high asymmetries and/or bimodal shapes, the PSDs estimated through the Bayesian method result more accurate than those obtained with the regularization technique.