NOTE
Determination of the Particle Size Distribution of Latex Using a Combination of Elastic Light Scattering and Turbidimetry: A Simulation Study nonabsorbing particles (5). In other words, the right tail of the distribution is more strongly weighted, and therefore a more precise estimation of the Turbidimetry at several wavelengths and elastic light scattering large particles should be expected. (ELS) at several angles can be used to determine particle size The problem of obtaining the PSD from ELS measurements using Mie distributions (PSDs) of suspended particles using Mie scattering theory was treated in Ref. (6) for homogeneous spherical particles. Good theory. The limited range of measurement in reciprocal space and
resolution was obtained in a range of diameters similar to the one used in the measurement noise always cause both techniques to have their (2) for polystyrene latex. In the Rayleigh-Debye-Gans approximation it own limitations. In this work, a way to combine turbidimetry and is possible to estimate an upper limit for the maximum diameter (D max ) ELS data to be processed together is proposed. In order to investithat can be found without error by an ideal scattering experiment (6). This gate the convenience of simultaneous versus individual processing, restriction, due to cutoff effects caused by the limited range of observation a set of computationally generated experiments performed on susin reciprocal space, is given by D max £ p/h min , where h min Å (4p/l) pensions of polystyrene particles is analyzed. The results clearly 1 sin(u min /2), l is the wavelength in the medium, and u min is the minimum scattering angle.
show that when the data are jointly processed the quality of the
The recovery of PSDs from ELS and turbidimetry using Mie theory PSDs obtained is highly superior. This fact can be explained by the requires in all cases the solution of Fredholm integral equations of the first complementary characteristics of both techniques. Turbidimetry is kind. The ill-posed nature of these equations in the continuous domain more accurate at larger sizes, whereas ELS gives more precise translates as ill-conditioned behavior in the discrete domain where the actual estimations at smaller sizes. This work shows that simultaneous computations take place. This situation was considered by the authors men- processing of experimental data from different sources seems to tioned above and proper regularization procedures were applied to obtain be a valid alternative for improving the quality of indirect measureuseful solutions. The fact that the experiments are corrupted with noise, ments.