Evaluating the environmental impact of new munitions requires that estimates be made of their physical chemical properties and, more important, the parameters that determine their fate in the environment. This approach is not new to the science of environmental assessment. The Toxic Substance Control Act started this approach with the development of quantitative structure activity relationships. However, the available methods are not reliable for chemical classes outside their training set. In particular, chemicals with nitro- group functionality have large prediction errors using the conventional methods.
The objectives of this project were to (1) improve methods for the prediction of physical properties relevant for assessing environmental fate and transport, and (2) develop predictive models to assess the bioaccumulation of the chemical in relevant environmental receptors (plants and soil invertebrates). The models are intended to be used to assess the environmental impact of new munitions compounds.
The models were designed to predict the equilibrium steady state concentration of munitions compounds in terrestrial plants and soil invertebrates exposed via the soil. The models assume that the phases in the soil were in equilibrium with the phases in the organism which contain the accumulated munitions compound. The schematic below illustrates the framework. The arrows represent partitioning equilibrium between the phases and the partition coefficients associated with each arrow are the parameters that are required to estimate the contaminant concentration in the organism that results from exposure to the contaminated soil.
The Abraham polyparameter model was used to estimate the partition coefficients between the phases and water. The partition coefficient (KP) is the ratio of concentrations of a solute in the system phases, for example, the ratio of a munitions component concentration in soil organic carbon to the concentration in water. The model specifies the dependence of KP on the chemical properties of the solute and the system phases. The equation is:
The upper case variables are the Abraham solute parameters and the lower case variables are the system phase Abraham parameters. Each pair quantifies the chemical interaction between the solute and the system phases. For example, aA quantifies the strength of the hydrogen bond between the solute hydrogen bond donation, A, and system pair hydrogen bond acceptance, a. Both sets of upper case and lower case parameters are required to make the estimate.
Quantum chemical methods have been developed to estimate the upper case Abraham solute parameters. They rely solely on the molecular structure of the solute. The lower case Abraham system parameters are estimated by fitting the above equation to data sets of experimentally determined partition coefficients for many solutes. This also requires the Abraham solute parameters.
The available databases for the bioconcentration factor (BCF = concentration of chemical in organism/concentration of chemical in water) in fish, plants, and soil invertebrates have been used to build the models. In addition the BCFs for a soil invertebrate (oligochaete Eisenia andrei) and a plant (barley Hordeum vulgare L) for munitions components and analog chemicals have been measured and added to the databases. This was necessary in order to expand the chemical space to include munitions compounds that are not represented in the extent data bases.
A methodology was developed using quantum chemical computations to estimate Abraham solute parameters for two types of applications: for use with presently existing Abraham models, or for use in building new Abraham models. In both cases the solute parameters have a smaller error than all other available estimation methods. This was the case for non-munitions component and for munitions components for which parameters were experimentally determined for the comparisons.
Bioaccumulation models for plants (grasses) and soil invertebrates (oligochaetes) that were exposed from compounds in soils have been developed as well as a bioconcentration model for fish exposed from compounds in water. The three models require partition coefficients for which models have also been developed: partitioning between water and organism lipid, organism protein, and plant cuticle, and a previously developed model for water-soil organic carbon. The number of observations in the data sets were: oligochaetes (57), protein (69) grasses (191), lipids (248), soil organic carbon (444) and fish (601).
The performance of the models was gauged using the root mean square error of the residuals: the difference between log10 modeled and log10 observed of either partition coefficients or concentration, which was equal to log10 of the ratio: model/observed. It is approximately equal to 0.40 for all the models, corresponding to approximately 80% of the residuals between 1/3 and 3 and approximately 90% of the residuals between 1/5 and 5.
The models can be used in a number of ways such as to assess the environmental impact of new munitions compounds. All that is required to make estimates of the extent of bioaccumulation for grasses, oligochaetes and fish is the molecular structure of the compound. In addition the physical chemical parameters: octanol-water and air-water (Henry’s Law constant) partition coefficients and aqueous solubility can also be estimated. Both the physical chemical and environmental partition coefficients are not as accurate as experimental determinations. However they are sufficiently quantitative for a number of tasks, for example to rank a sequence of proposed new munitions compounds to bioaccumulate in organisms, or to establish that the risk is either low enough so that no further experimental information is necessary, or, conversely that the risk is estimated to be high enough so that further investigation is required if the compound is to move forward. This ability to determine the potential extent of environmental risk for new munitions compounds is the most immediate benefit from using the models developed in this project. (Project Completion - 2018).
Kuo, D.T.F and D.M. Di Toro. 2013. A Reductionist Mechanistic Model for Bioconcentration of Neutral and Weakly Polar Organic Compounds in Fish. Environmental Toxicology and Chemistry, 32(9):2089-2099.
Liang, Y., D.T.F. Kuo, H.E. Allen, and D.M. Di Toro. 2016. Experimental Determination of Solvent-Water Partition Coefficients and Abraham Parameters for Munition Constituents. Chemosphere, 161:429-437.
Liang, Y., T.L. Torralba-Sanchez, and D.M. Di Toro. 2018. Estimating System Parameters for Solvent–Water and Plant Cuticle–Water using Quantum Chemically Estimated Abraham Solute Parameters. Environmental Science: Process & Impacts, 20:813-821.
Liang, Y., R. Xiong, S.I. Sandler, and D.M. Di Toro. 2017. Quantum Chemically Estimated Abraham Solute Parameters Using Multiple Solvent–Water Partition Coefficients and Molecular Polarizability. Environmental Science & Technology, 51(17):9887-9898.
Torralba-Sanchez, T.L., Y. Liang, and D.M. Di Toro. 2017. Estimating Grass–Soil Bioconcentration of Munitions Compounds from Molecular Structure. Environmental Science & Technology, 51(19):11205-11214.