Angle-dependent effects in DLS measurements of polydisperse particles

Abstract

Abstract Dynamic light scattering (DLS) is widely used for analyzing biological polymers and colloids. Its application to nanoparticles in medicine is becoming increasingly important with the recent emergence of prominent lipid nanoparticle (LNP)-based products, such as the SARS-CoV-2 vaccines from Pfizer, Inc.-BioNTech (BNT162b2) and Moderna, Inc. (mRNA-1273). DLS plays an important role in the characterization and quality control of nanoparticle-based therapeutics and vaccines. However, most DLS instruments have a single detection angle θ, and the amplitude of the scattering vector, q, varies among them according to the relationship q = (4πn/λ 0) sin(θ/2), where λ 0 is the laser wavelength. Results for identical, polydisperse samples among instruments of varying q yield different hydrodynamic diameters, because, as particles become larger they scatter less light at higher q, so that higher-q instruments will under-sample large particles in polydisperse populations, and report higher z-average diffusion coefficients, and hence smaller effective hydrodynamic diameters than lower-q instruments. As particle size reaches the Mie regime the scattering envelope manifests angular maxima and minima, and the monotonic decrease of average size versus q is lost. The discrepancy among instruments of different q is hence fundamental, and not merely technical. This work examines results for different q-value instruments, using mixtures of monodisperse latex sphere standards, for which experimental measurements agree well with computations, and also polydisperse solutions of physically-degraded LNPs, for which results follow expected trends. Mie effects on broad unimodal populations are also considered. There is no way to predict results between two instruments with different q for samples of unknown particle size distributions. Initial analysis of the polydispersity index among different instruments shows a technical difference due to method of autocorrelation analysis, in addition to the fundamental q-effect.