Effects of normalization errors on size distributions obtained from dynamic light scattering data.

This paper presents a study of the influence of normalization errors on size distributions obtained from the analysis of intensity fluctuations by photon correlation spectroscopy. The effects of these errors are demonstrated by means of computer-generated autocorrelation functions simulating light scattered from a monomodal Schulz distribution of small, spherical, unilamellar lipid vesicles. The calculations show that even small errors in the baseline, modifying the data upon normalization systematically, will cause serious errors in the estimated size distribution. As it turns out this is due to the peculiar characteristics of normalization errors in data of the first order autocorrelation function. The errors introduced there are described in parts by functions of the delay time having positive exponents. Such components are not considered in the integral equations commonly used to analyze the measured data. The error's property to be a function of delay time in turn enables us to obtain the relative baseline error from the inversion of the data. The new method for its determination is described in some detail. Here, it has been realized with a modified version of the size distribution algorithm CONTIN.