Artefacts in semivolatile organic compound sampling with polyurethane foam substrates in high volume cascade impactors

Polyurethane foam (PUF) is known to sorb gas-phase semivolatile organic compounds (SVOCs) from ambient air and is used routinely in conventional high volume filter-sorbent sampling of such pollutants. PUF rings have also been employed as impaction substrates in a high volume cascade impactor (HVCI) used as a sampler for the evaluation of particle toxicity. Though nonvolatile particles (e.g., trace metals, inorganic ions) have been the primary focus, the sampler has also been used to measure particulate SVOC concentrations in ambient air. The aim of this work is to investigate the validity of the latter approach. The results of three sets of experiments conducted in Canada and Denmark are reported herein. Model compounds included native and deuterated polycyclic aromatic hydrocarbons (PAHs). The experiments demonstrated that HVCI PUF substrates sorb gas-phase PAH compounds and that the sorbed mass is subject to mobilization through and out of the sampler. Particulate concentrations of low molecular weight and volatile PAHs are therefore prone to overestimation in samples that have been analyzed after extraction of whole PUF substrates. Sonication of collected particles in water before solvent extraction is effective at dislodging them from the PUF but also acts to redistribute their originally particulate PAH mass back to the PUF and to the sonication water. As a result, the PAH content of particles measured after sonication and subsequent filtration does not accurately represent their true values. These artefacts affect not only measured PAH concentrations but also the results of toxicological assays that are conducted to test the characteristics of particles collected using HVCI PUF samplers.