Abstract
Questions still exist regarding which indicator better estimates worker’s exposure to diesel particulate matter (DPM) and, especially for ultrafine particles (UFP), how exposure levels and the characteristics of the particles vary in workplaces with different exposure conditions. This study aimed to quantify and characterize DPM exposures in three workplaces with different exposure levels: an underground mine, a subway tunnel, and a truck repair workshop. The same sampling strategy was used and included measurements of the particle number concentration (PNC), mass concentration, size distribution, transmission electron microscopy (TEM), and the characterization of carbonaceous fractions. The highest geometric means (GMs) of PNC and elemental carbon (EC) were measured in the mine [134 000 (geometric standard deviation, GSD = 1.5) particles cm−3 and 125 (GSD = 2.1) µg m−3], followed by the tunnel [32 800 (GSD = 1.7) particles cm−3 and 24.7 (GSD = 2.4) µg m−3], and the truck workshop [22 700 (GSD = 1.3) particles cm−3 and 2.7 (GSD = 2.4) µg m−3]. This gradient of exposure was also observed for total carbon (TC) and particulate matter. The TC/EC ratio was 1.4 in the mine, 2.5 in the tunnel and 8.7 in the workshop, indicating important organic carbon interference in the non-mining workplaces. EC and PNC were strongly correlated in the tunnel (r = 0.85; P < 0.01) and the workshop (r = 0.91; P < 0.001), but a moderate correlation was observed in the mine (r = 0.57; P < 0.05). Results from TEM showed individual carbon spheres between 10 and 56.5 nm organized in agglomerates, while results from the size distribution profiles showed bimodal distributions with a larger accumulation mode in the mine (93 nm) compared with the tunnel (39 nm) and the truck workshop (34 nm). In conclusion, the composition of the carbonaceous fraction varies according to the workplace, and can interfere with DPM estimation when TC is used as indicator. Also, the dominance of particles <100 nm in all workplaces, the high levels of PNC measured and the good correlation with EC suggest that UFP exposures should receive more attention on occupational routine measurements and regulations.
A predictive model for elemental carbon, organic carbon and total carbon based on laser induced breakdown spectroscopy measurements of filter-collected diesel particulate matter
