Atmospheric deposition of semivolatile organic compounds to plants

Abstract

This thesis describes how the deposition of semivolatile organic compounds (SOCs) to plant surfaces is affected by the characteristics of the plant. From a literature review, it was concluded that differences between SOC concentrations in different plant species are often very small (< a factor of 8), although sometimes also large differences (up to a factor > 50) are measured. In the experimental part of the study, the effect of plant properties on the deposition of polycyclic aromatic hydrocarbons (PAHs) to three species of Plantago (plantain) was studied. For these three species, the most important plant characteristics are the size of the plant and the presence of leaf hairs. For PAHs with a low molecular weight (MW £ 202), the deposition of gases is the dominating route of deposition, whereas for PAHs with higher MW (MW ³ 228) the deposition of particle-bound molecules is the main deposition mechanism. In contrast with the gaseous PAHs, which are mostly present in the cuticular wax and the remaining interior of the leaf, the particle-bound PAHs are present at the leaf surface. As a consequence, deposited particle-bound PAHs can be relatively easily washed off by e.g. rain. For the consumption of vegetables by humans, particle-bound deposition of SOCs is probably not a significant process, as particle-bound compounds are largely removed from the plants by washing. For cattle feed however, it is recommended to include particle-bound deposition in the model. Current models can only roughly predict the deposition of SOCs to plants, due to the high variation in plant characteristics and environmental conditions. For continuous emissions, the use of plant specific parameters in predictive models has little advantage, because (1) plant parameters are not constant, but dependent on environmental conditions and (2) environmental conditions also have a large, but quantitatively unknown effect on the SOC-concentrations of the plants. In case of incidents, the use of values for specific plants may improve the precision of the predictions. However, if accurate concentrations in plants are needed for risk assessment, concentrations should be measured and not calculated with a model.