Search for a command to run...
Dynamic light scattering (DLS) is an essential technique for nanoparticle size analysis and has been employed extensively for decades, but despite its long history and popularity, the choice of weighting and mean of the size distribution often appears to be picked ad-hoc to bring the results into agreement with other methods and expectations <i>by any means necessary</i>. Here, we critically discuss the application of DLS for nanoparticle characterization and provide much-needed clarification for ambiguities in the mean-value practice of commercial DLS software and documentary standards. We address the misleading way DLS size distributions are often presented, that is, as a logarithmically-scaled histogram of measured relative quantities. Central values obtained incautiously from this representation often lead to significant interpretation errors. Through the measurement of monomodal nanoparticle samples having an extensive range of sizes (5 nm to 250 nm) and polydispersity, we similarly demonstrate that the default outputs of a frequently-used DLS inversion method are ill chosen, as they are regularizer-dependent and significantly deviate from the cumulant z-average size. The resulting discrepancies are typically larger than 15 % depending on the polydispersity index of the samples. We explicitly identify and validate the harmonic mean as the central value of the intensity-weighted DLS size distribution that expresses the inversion results consistently with the cumulant results. We also investigate the extent to which the DLS polydispersity descriptors are representative of the distributional quality and find them to be unreliable and misleading, both for monodisperse reference materials and broad-distribution biomedical nanoparticles. These results overall are intended to bring essential improvements and to stimulate reexamination of the metrological capabilities and role of DLS in nanoparticle characterization.