An investigation into the use of < 38 µm fraction as a proxy for < 10 µm road dust particles

Abstract

It is well documented that a large portion of urban particulate matters is derived from road dust. Isolating particles of RD which are small enough to be inhaled, however, is a difficult process. In this study, it is shown for the first time that the < 38 µm fraction of road dust particles can be used as a proxy for road dust particles < 10 µm in bioaccessibility studies. This study probed similarities between the < 10 and < 38µm fractions of urban road dust to show that the larger of the two can be used for analysis for which larger sample masses are required, as is the case with in vitro analysis. Road dust, initially segregated to size < 38 µm using sieves, was again size segregated to < 10 µm using water deposition. Both the original < 38 µm and the separated < 10 µm fractions were then subject to single particle analysis by SEM-EDX and bulk analysis by ICP-OES for its elemental composition. Dissolution tests in artificial lysosomal fluid, representative of lung fluid, were carried out on both samples to determine % bioaccessibility of selected potentially harmful elements and thus probe similarities/differences in in vitro behaviour between the two fractions. The separation technique achieved 94.3% of particles < 10 µm in terms of number of particles (the original sample contained 90.4% as determined by SEM-EDX). Acid-soluble metal concentration results indicated differences between the samples. However, when manipulated to negate the input of Si, SEM-EDX data showed general similarities in metal concentrations. Dissolution testing results indicated similar behaviour between the two samples in a simulated biological fluid.

Keywords: Bioaccessibility; Representation; Respirable particles; Size-segregated road dust.

Fig. 1

Separation schematic (https://chemix.org/)

Fig. 2

Particle size distributions from each fraction, where n_j/n_t indicates the fraction of particles present in that size range and N(D_p)/n_t indicates the cumulative distribution