Dust discrimination in dynamic light scattering based on a quantile outliers detection method

Abstract

Dynamic light scattering (DLS) is a popular method for measuring particle size in the colloidal size range. However, like all light scattering techniques, the results can be problematic when dust or large contaminant particles are present in the sample. Dust particles are usually much larger than the colloids or nanoparticles of interest and, consequently, scatter light very strongly when they pass through the laser beam in the scattering volume. This biases the measured particle size to much larger values than expected. However, it also suggests a means to detect and, potentially, reduce the effect of dust particles on the measurement results. If the high count-rate data associated with dust can be detected and removed during the DLS measurement, the particle size results will be more reliable. An a priori electronic dust filtering method based on the long-time count rate statistics is proposed and demonstrated. This allows the identification of data sets which have an abnormal number of high count-rate events which are, presumably, associated with scattering from dust particles. These data can be rejected and an autocorrelation function calculated using the retained counts after removing unwanted outlier data. The particle size distribution is then recovered. By measuring a standard polystyrene suspension containing “dust” particles, the results show that the method can effectively discriminate and eliminate data contaminated by dust, thereby improving the measurement results.