Predicting 2,4-dintroanisole (DNAN) sorption on various soil 'types' using different compositional datasets

Soil heterogeneity is a major contributor to the uncertainty in predicting the environmental fate of data-scare contaminants. For this paper, we focused on research designed to predict the soil environmental fate of the new munition compound, called 2,4-dinitroanisole (DNAN) -a compound increasingly employed by the U.S. and international militaries in the next-generation, insensitive explosive formulations. Here, we employed multivariate statistical correlation models to predict DNAN sorption among different soil “types” seeking to reduce the uncertainty common in all contaminant sorption models by using soil taxonomic designation as a calibrant. We collected composite soils classified under the Ultisols taxonomic Order in the U.S. National Resource Conservation Service soil classification system and quantified their properties via physical and chemical characterizations. Using multivariate statistical modeling modified for compositional data analysis (CoDa), we developed quantitative analogies of the Ultisols by partitioning the characterization data up into four different compositions: Water-extracted, Mehlich-III (referring to the weak acid) extracted, particle-size distribution, and solid-phase carbon‑nitrogen‑sulfur compositions. DNAN sorption was measured in batch soil suspensions and distribution coefficients (KD) were calculated using linear regression modeling. Prediction models testing the correlation of the DNAN KD values to the centered logratio -transformed compositions were calculated using CoDa-modified multilinear regression. Results showed that DNAN sorption was only predictable by dissolved organic carbon, pH, and the exchangeable cations Ca and K within the water-extracted composition. Analogies for DNAN sorption were the most discriminating at the Suborder level because of the inherent ambiguity in the Hapludults class at the Great Group level.