Investigation and correction of size-dependent transport efficiencies of microparticles in SP ICP-MS analysis

Single particle inductively coupled plasma-mass spectrometry (SP ICP-MS) has emerged as a powerful technique for characterising nanoparticles (NPs) and is increasingly used to target microparticles (MPs) including microplastics. The accuracy of determined size distributions and number concentrations in SP ICP-MS is critically dependent on a precise knowledge of transport efficiency (TE), which describes the fraction of particles transferred from a suspension into the plasma. While for NPs TE can be assumed to be a constant, a sharp decline can be noted as sizes of particles increase. In this study, we systematically investigated the size dependency of TE by analysing polystyrene (PS) MPs within a range from 1 μm to 20 μm using SP ICP-MS and flow cytometry (FC). The latter was employed as an independent technique, which allowed counting of microparticles to determine particle number concentrations. We evaluated two nebuliser/spray chamber combinations, a conventional Scott-type and a total consumption setup. To improve the comparison of the same standard suspensions across the different setups, a novel multi-particle event filter was implemented to identify overlapping event signals and to improve particle counting accuracy. The shift of the TE from a constant to a size-variable parameter has significant repercussions for the accurate determination of both particle number concentrations and size distributions. Using the Scott-type spray chamber and the total consumption setup, the aerosol-based TEs were 5.5% and 47.3%, respectively. When targeting MPs with a size of 3 μm, TEs decreased to 1.3% and 30.0%. For larger MPs of 10 μm it further decreased to 0.1% and 3.3%, respectively. To address the size dependency of the TE and the resulting overrepresentation of small particles in size and frequency-based calibrations, we propose a correction method based on a sigmoidal TE model that retrospectively corrects size histograms using bin-by-bin frequency adjustment.