Particle transfer and adherence to human skin compared with cotton glove and pre-moistened polyvinyl alcohol exposure sampling substrates.

Measurement of skin exposure to particles using interception (e.g., cotton gloves) and removal (e.g., wiping) sampling techniques could be inaccurate because these substrates do not have the same topography and adhesion characteristics as skin. The objective of this study was to compare particle transfer and adherence to cotton gloves, cotton gloves with artificial sebum, and a pre-moistened polyvinyl alcohol (PVA) material with bare human skin (fingertip, palm). Experiments were performed with aluminum oxide powder under standardized conditions for three types of surfaces touched, applied loads, contact times, and powder mass levels. In the final mixed model, the fixed effects of substrate, surface type, applied load, and powder mass and their significant two-way interaction terms explained 71% (transfer) and 74% (adherence) of the observed total variance in measurements. For particle mass transfer, compared with bare skin, bias was -77% (cotton glove with sebum) to +197% (PVA material) and for adherence bias ranged from -40% (cotton glove) to +428% (PVA material), which indicated under- and over-sampling by these substrates, respectively. Dermal exposure assessment would benefit from sampling substrates that better reflect human skin characteristics and more accurately estimate exposures. Mischaracterization of dermal exposure has important implications for exposure and risk assessment.