Modeling the adsorption and desorption of munition constituents on soils using reversible and resistant components
Date
2015
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
The need to understand the fate of chemicals is important in performing risk assessments of chemical compounds. Munition constituents such as TNT and nitroglycerin, as well as newly developed compounds, are of significant concern due to the wide range of contamination across a variety of soil types. Currently, there are more than 15 million acres of potentially impacted soils where explosive and explosive related materials have been found in abundance. The bioavailability of these compounds and potential for transport to groundwater is controlled to a large degree by sorption onto soil. Studies have shown both reversible and resistant (hysteretic) desorption behavior for chemicals on soils. Observations of 5 munition constituents sorbed to 25 soils and 2 reference sorbents show that a constant fraction of sorbed chemical is released during desorption. However, an analysis using a fully reversible linear partitioning model yields accurate predictions of desorption for only short adsorption contact times and only for low organic carbon containing soils. A model that accounts for incompletely reversible sorption, the reversible resistant model, improves predictive capabilities. The model's partition coefficients can be correlated to soil properties, which allow predictions to be made for the sorption behavior of soils for which only the organic carbon and clay sized particle fraction is known. However the dependence of the degree of hysteresis on adsorption contact time remained to be explained. Using a variable strength binding site model, the site transformation model, shows that changes in binding strength are regular and predictable. Further, results suggest that the mechanism for hysteresis is related to soil organic carbon deformation during adsorption. To demonstrate the capabilities of the site transformation model, an a priori prediction of sorption behavior is presented. As a measure of implications for field study, a reversible resistant model is used to analyze soil column experiments. Results show the importance of accounting for hysteresis when applying bench scale results to field scale problems. A diffusion model, employing multiple particle sizes, is developed to describe hysteresis. Results indicate the need for substantial sorbent properties for accurate modeling. An n-particle model is presented and future work is suggested.