Model-dependent variability in environmental exposure assessment of engineered nanomaterials: A comparison of material flow analysis and multimedia environmental fate models.

The predicted environmental concentrations (PECs) of engineered nanomaterials (ENMs) are determined not only by the quantity of ENMs released but also by how each fate model represents particulate behavior after release. Here, model-dependent variability was evaluated by integrating material flow analysis (MFA) with ChemCAN, used as a conventional chemical multimedia reference model, and SimpleBox4Nano, used as a nano-specific fate model. TiO2-containing coating/cleaning products were selected as the benchmark category because they generated disproportionate environmental releases despite representing only 3.1% of the inventory. ChemCAN strongly suppressed air and water PECs and retained TiO2 mainly in soil, whereas SimpleBox4Nano predicted substantially higher concentrations, with median SimpleBox4Nano/ChemCAN ratios of 5.5 × 107 for air, 4.7 × 103 for water, and 4.4 for soil. A 10,000-run Monte Carlo analysis showed that these cross-model differences persisted after uncertainty in inventory mass, transfer fractions, spatial parameters, and particle properties was propagated. Particle radius dominated SimpleBox4Nano uncertainty in air and water, whereas zeta potential had a minor effect. The results establish distinct model roles: MFA is appropriate for transparent release-scenario construction and source prioritization, while post-release interpretation of particulate ENMs requires a nano-specific fate model.