We report an empirical poly-parameter linear free energy relationship (LFER) for estimating the mass-normalized rate constants for the abiotic reduction of munition compounds (MC) in soil. A total of 131 kinetic experiments were performed, using three classes of MC (nitroaromatic [TNT, DNAN], nitramine [RDX], and azole [NTO]) and 11 soils having highly varied organic carbon and iron contents and reduced with dithionite to different electron contents. The LFER has the same form as that for MC reduction by FeIII (oxyhydr)oxide-FeaqII redox couples and predicts MC reduction rate constants to within an order of magnitude, using only the aqueous-phase one electron reduction potential (EH1) of the MC and the pe and pH of the soil. As previously shown for azoles, which exhibited markedly higher reactivity toward iron than toward carbon reductants relative to all neutral MC, NTO reduction rate depended on soil composition and hence a correction to model prediction was necessary at soil iron-to-carbon mass ratios ≲1. This is the first successful attempt to predict the reduction kinetics of structurally diverse nitro compounds in compositionally complex soils based on their thermodynamic properties. The LFER would be useful in the management/restoration (e.g., natural or enhanced attenuation) of soils impacted by MC or other nitro pollutants.

The electron accepting capacity (EAC) of soil plays a pivotal role in the biogeochemical cycling of nutrients and transformation of redox-labile contaminants. Prior EAC studies of soils and soil constituents utilized different methods, reductants, and mediators, making cross-study comparison difficult. This study was conducted to quantify and compare the EACs of two soil constituents (hematite and Leonardite humic acid) and 12 soils of diverse composition, using chemical redox titration (CRT) with dithionite as the reductant and mediated electrochemical reduction (MER) with diquat as the mediator. The EACs of hematite and humic acid measured by CRT (EACCRT) and MER (EACMER) are similar and close to the theoretical/reported values. For soils, EACCRT and EACMER increased with iron and organic carbon (TOC) contents, suggesting iron and carbon were the main contributors to soil EAC. EACCRT > EACMER for all soils, and their difference (Delta EAC = EACCRT- EACMER) increased with TOC, presumably due to the longer contact time in CRT and thus more complete reduction of carbonaceous redox moieties. We propose an equation that relates EACCRT to EACMER (Delta EAC = 1796fTOC + 32) and another that predicts EACCRT from dithionite-reducible Fe and TOC (EACCRT = 2705 mu mol e-/g C x fTOC + 17907 mu mol e-/g Fe x fFedithionite-reducible). Our results suggest that at least 10-15% of soil organic carbon contributed to EACCRT.

Abiotic reduction by iron minerals is arguably the most important fate process for munition compounds (MCs) in subsurface environments. No model currently exists that can predict the abiotic reduction rates of structurally diverse MCs by iron (oxyhydr)oxides. We performed batch experiments to measure the rate constants for the reduction of three classes of MCs (poly-nitroaromatics, nitramines, and azoles) by hematite or goethite in the presence of aqueous Fe2+. The surface area-normalized reduction rate constant (kSA) depended on the aqueous-phase one-electron reduction potential (EH1) of the MC and the thermodynamic state (i.e., pe and pH) of the iron oxide-Feaq2+ system. A linear free energy relationship (LFER), similar to that reported previously for nitrobenzene, successfully captures all MC reduction rate constants that span 6 orders of magnitude: log(kSA)=(1.12±0.04)[0.53EH159mV-(pH+pe)]+(5.52±0.23). The finding that the rate constants of all the different classes of MCs can be described by a single LFER suggests that these structurally diverse nitro compounds are reduced by iron oxide-Feaq2+ couples through a common mechanism up to the rate-limiting step. Multiple mechanistic implications of the results are discussed. This study expands the applicability of the LFER model for predicting the reduction rates of legacy and emerging MCs and potentially other nitro compounds.

No single linear free energy relationship (LFER) exists that can predict reduction rate constants of all munition constituents (MCs). To address this knowledge gap, we measured the reduction rates of MCs and their surrogates including nitroaromatics [NACs; 2,4,6-trinitrotoluene (TNT), 2,4-dinitroanisole (DNAN), 2-amino-4,6-dinitrotoluene (2-A-DNT), 4-amino-2,6-dinitrotoluene (4-A-DNT), and 2,4-dinitrotoluene (DNT)], nitramines [hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) and nitroguanidine (NQ)], and azoles [3-nitro-1,2,4-triazol-5-one (NTO) and 3,4-dinitropyrazole (DNP)] by three dithionite-reduced quinones (lawsone, AQDS, and AQS). All MCs/NACs were reduced by the hydroquinones except NQ. Hydroquinone and MC speciations were varied by controlling pH, permitting the application of a speciation model to determine second-order rate constants (k) from observed pseudo-first-order rate constants. The intrinsic reactivity of MCs (oxidants) decreased upon deprotonation, while the opposite was true for hydroquinones (reductants). The rate constants spanned ∼6 orders of magnitude in the order NTO ≈ TNT > DNP > DNT ≈ DNAN ≈ 2-A-DNT > DNP- > 4-A-DNT > NTO- > RDX. LFERs developed using density functional theory-calculated electron transfer and hydrogen atom transfer energies and reported one-electron reduction potentials successfully predicted k, suggesting that these structurally diverse MCs/NACs are all reduced by hydroquinones through the same mechanism and rate-limiting step. These results increase the applicability of LFER models for predicting the fate and half-lives of MCs and related nitro compounds in reducing environments.

Iron (oxyhydr)oxides comprise a significant portion of the redox-active fraction of soils and are key reductants for remediation of sites contaminated with munition constituents (MCs). Previous studies of MC reduction kinetics with iron oxides have focused on the concentration of sorbed Fe(II) as a key parameter. To build a reaction kinetic model, it is necessary to predict the concentration of sorbed Fe(II) as a function of system conditions and the redox state. A thermodynamic framework is formulated that includes a generalized double-layer model that utilizes surface acidity and surface complexation reactions to predict sorbed Fe(II) concentrations that are used for fitting MC reduction kinetics. Monodentate- and bidentate Fe(II)-binding sites are used with individual oxide sorption characteristics determined through data fitting. Results with four oxides (goethite, hematite, lepidocrocite, and ferrihydrite) and four nitro compounds (NB, CN-NB, Cl-NB, and NTO) from six separate studies have shown good agreement when comparing observed and predicted surface area-normalized rate constants. While both site types are required to reproduce the experimental redox titration, only the monodentate site concentration controls the MC reaction kinetics. This model represents a significant step toward predicting the timescales of MC degradation in the subsurface.

Dissolved organic matter (DOM) comprises a sizeable portion of the redox-active constituents in the environment and is an important reductant for the abiotic transformation of nitroaromatic compounds and munition constituents (NACs/MCs). Building a predictive kinetic model for these reactions would require the energies associated with both the reduction of the NACs/MCs and the oxidation of the DOM. The heterogeneous and unknown structure of DOM, however, has prohibited reliable determination of its oxidation energies. To overcome this limitation, humic acids (HAs) were used as model DOM, and their redox moieties were modeled as a collection of quinones of different redox potentials. The reduction and oxidation energies of the NACs/MCs and hydroquinones, respectively, via hydrogen atom transfer (HAT) reactions were then calculated quantum chemically. HAT energies have been used successfully in a linear free energy relationship (LFER) to predict second-order rate constants for NAC reduction by hydroquinones. Furthermore, a linear relationship between the HAT energies and the reduction potentials of quinones was established, which allows estimation of hydroquinone reactivity (i.e., rate constants) from HA redox titration data. A training set of three HAs and two NACs/MCs was used to generate a mean HA redox profile that successfully predicted reduction kinetics in multiple HA/MC systems.

3-Nitro-1,2,4-triazol-5-one (NTO) is a major and the most water-soluble constituent in the insensitive munition formulations IMX-101 and IMX-104. While NTO is known to undergo redox reactions in soils, its reaction with soil humic acid has not been evaluated. We studied NTO reduction by anthraquinone-2,6-disulfonate (AQDS) and Leonardite humic acid (LHA) reduced with dithionite. Both LHA and AQDS reduced NTO to 3-amino-1,2,4-triazol-5-one (ATO), stoichiometrically at alkaline pH and partially (50-60%) at pH ≤ 6.5. Due to NTO and hydroquinone speciation, the pseudo-first-order rate constants (kObs) varied by 3 orders of magnitude from pH 1.5 to 12.5 but remained constant from pH 4 to 10. This distinct pH dependency of kObs suggests that NTO reactivity decreases upon deprotonation and offsets the increasing AQDS reactivity with pH. The reduction of NTO by LHA deviated continuously from first-order behavior for >600 h. The extent of reduction increased with pH and LHA electron content, likely due to greater reactivity of and/or accessibility to hydroquinone groups. Only a fraction of the electrons stored in LHA was utilized for NTO reduction. Electron balance analysis and LHA redox potential profile suggest that the physical conformation of LHA kinetically limited NTO access to hydroquinone groups. This study demonstrates the importance of carbonaceous materials in controlling the environmental fate of NTO.

A linear free energy model is presented that predicts the second order rate constant for the abiotic reduction of nitroaromatic compounds (NACs). For this situation previously presented models use the one electron reduction potential of the NAC reaction. If such value is not available, it has been has been proposed that it could be computed directly or estimated from the electron affinity (EA). The model proposed herein uses the Gibbs free energy of the hydrogen atom transfer (HAT) as the parameter in the linear free energy model. Both models employ quantum chemical computations for the required thermodynamic parameters. The available and proposed models are compared using second order rate constants obtained from five investigations reported in the literature in which a variety of NACs were exposed to a variety of reductants. A comprehensive analysis utilizing all the NACs and reductants demonstrate that the computed hydrogen atom transfer model and the experimental one electron reduction potential model have similar root mean square errors and residual error probability distributions. In contrast, the model using the computed electron affinity has a more variable residual error distribution with a significant number of outliers. The results suggest that a linear free energy model utilizing computed hydrogen transfer reaction free energy produces a more reliable prediction of the NAC abiotic reduction second order rate constant than previously available methods. The advantages of the proposed hydrogen atom transfer model and its mechanistic implications are discussed as well.

3-Nitro-1,2,4-triazol-5-one (NTO) is an insensitive munition compound (MC) that has replaced legacy MC. NTO can be highly mobile in soil and groundwater due to its high solubility and anionic nature, yet little is known about the processes that control its environmental fate. We studied NTO reduction by the hematite-Fe2+ redox couple to assess the importance of this process for the attenuation and remediation of NTO. Fe-(aq)(2+) was either added (type I) or formed through hematite reduction by dithionite (type II). In the presence of both hematite and Fe-(aq)(2+), NTO was quantitatively reduced to 3-amino-1,2,4-triazol-5-one following first-order kinetics. The surface area-normalized rate constant (k(SA)) showed a strong pH dependency between 5.5 and 7.0 and followed a linear free energy relationship (LFER) proposed in a previous study for nitrobenzene reduction by iron oxide-Fe2+ couples, i.e., log k(SA) = -(pe + pH) + constant. Sulfite, a major dithionite oxidation product, lowered k(SA) in type II system by similar to 10-fold via at least two mechanisms: by complexing Fe2+ and thereby raising pe, and by making hematite more negatively charged and hence impeding NTO adsorption. This study demonstrates the importance of iron oxide-Fe2+ in controlling NTO transformation, presents an LFER for predicting NTO reduction rate, and illustrates how solutes can shift the LFER by interacting with either iron species.

Determining the fate of nitroaromatic compounds (NACs) in the environment requires the use of predictive models for compounds and conditions for which experimental data are insufficient. Previous studies have developed linear free energy relationships (LFERs) that relate the thermodynamic energy of NAC reduction to its corresponding rate constant. We present a comprehensive LFER that incorporates both the reduction and oxidation half‐reactions through quantum chemically calculated energies. Environ Toxicol Chem 2020;39:2389–2395. © 2020 SETAC

Cr(vi) has posed a serious risk for the environment and human beings because of its pollution and toxicity. It is essential to understand the equilibrium behavior of Cr(vi) in soils. In this study, the adsorption of Cr(vi) on fourteen soils was studied with batch experiments and quantitative modeling. The batch experiments included the adsorption edge and adsorption isotherm experiments, investigating the adsorption of Cr(vi) with varying soil properties, solution pH, and initial Cr(vi) concentrations. The experimental data were then modeled using the surface complexation models in Visual MINTEQ of CD-MUSIC by considering the adsorption of Cr(vi) and ions onto Fe (hydr)oxides and Al (hydr)oxides, and the Stockholm Humic Model and the fixed charge site model by accounting for the adsorption of the cations to soil organic matter and clay, respectively. Particularly, the modeling method of this study introduced an important parameter RO- to account for the amount of soil organic matter irreversibly adsorbed on soil minerals. Overall, the model predicted reasonably well for the equilibrium partition of Cr(vi) under various conditions with a root-mean-square-error of 0.35 for the adsorption edge data and 0.19 for the adsorption isotherm data. According to the model calculations, ferrihydrite dominated the binding of Cr(vi) at pH of 3.0-7.0. The content of ferrihydrite and reactive soil organic matter was found to be the main factor influencing RO-. The modeling results help to understand and predict Cr(vi) adsorption on different soils and are beneficial to environmental risk assessment and pollution remediation.

Nitroaromatic compounds (NACs) are a class of prevalent contaminants. Abiotic reduction is an important fate process that initiates NAC degradation in the environment. Many linear free energy relationship (LFER) models have been developed to predict NAC reduction rates. Almost all LFERs to date utilize experimental aqueous-phase one-electron reduction potential ( EH1) of NAC as a predictor, and thus, their utility is limited by the availability of EH1 data. A promising new approach that utilizes computed hydrogen atom transfer (HAT) Gibbs free energy instead of EH1 as a predictor was recently proposed. In this study, we evaluated the feasibility of HAT energy for predicting NAC reduction rate constants. Using dithionite-reduced quinones, we measured the second-order rate constants for the reduction of seven NACs by three hydroquinones of different protonation states. We computed the gas-phase energies for HAT and electron affinity (EA) of NACs and established HAT- and EA-based LFERs for six hydroquinone species. The results suggest that HAT energy is a reliable predictor of NAC reduction rate constants and is superior to EA. This is the first independent, experimental validation of HAT-based LFER, a new approach that enables rate prediction for a broad range of structurally diverse NACs based solely on molecular structures.

Energetic materials (EMs) bound to propellant residues can contribute to environmental risk and public health concerns. This work investigated how nitrocellulose, a common binding material in propellants, may control the release dynamics of nitroglycerin (NG) and 2,4-dinitrotoluene (2,4-DNT) from propellant residues. Batch adsorption/desorption experiments on nitrocellulose and re-interpretation on results from past leaching studies involving propellant-bound EMs were conducted. Mechanistic modeling of adsorption/desorption kinetics based on intra-particle diffusion (IPD) predicted aqueous intrinsic diffusivities (Diw) to within a factor of 2 of expected values. Furthermore, the IPD model was able to predict effective diffusivities (Deff) during the early leaching of NG from propellant residues to within a factor of 2 over a 3-log unit range. Prediction of leaching Deff's associated with fired residues was less successful probably due to the neglect of compositional and morphological heterogeneity within the residues. Close correlations were found between the early and late Deff's of residue-bound NG and between the fast- and slow-domain rate constants for both EMs, suggesting that the late leaching kinetics of bound-EMs may be empirically assessed from the early kinetics. This work illustrates that, in addition to dissolution, retarded diffusion through nitrocellulose matrix may also limit the overall release and transformation of residue-bound EMs in the field. Implications and limitations of the current study, and the steps forward are also presented.

Constants for a Terrestrial Biotic Ligand Model (TBLM) to predict the Cu toxicity to barley root elongation (RE) were developed from controlled sand culture experiments. These constants were used to predict RE in soil culture. The competition of H+, Ca2+, and Mg2+ to Cu2+ toxicity were studied individually and independently, and linear relationships between EC50 free Cu2+ and H+, Ca2+, and Mg2+ activities were found, meaning that the cations H+, Ca2+, and Mg2+ will alleviate the toxicity of Cu2+ in solutions. Toxicity accompanying increasing concentration of solution ions other than Cu2+ was observed and modeled as an osmotic effect which improved soil culture toxicity prediction. The Root Mean Square Error (RMSE) of %RE and EC50 (50% effective concentration) for soil toxicity prediction using TBLM parameters developed from sand culture are 13.0 and 0.23 respectively, which are as good as that of 14.0 and 0.24 using parameters that developed from soil culture itself. A model including the activity at the root plasma membrane surface was tested and found not to provide improvement over the use of bulk solution activity to predict metal toxicity. TBLM parameters obtained from water solution culture were unable to accurately predict the EC50s in soils whereas the parameters obtained from sand culture were able to predict the toxicity in soils. Including the toxicity of CuOH+ was found to improve the toxicity prediction slightly.

Sustainable management of military ranges requires effective assessment of surface mobility and leaching potential of propellant compounds (PCs). Previous studies have focused mostly on PCs' dissolution from fired residues and their sorption to soil components. This work investigated the potential role of nitrocellulose, a major component in propellants, in the binding of PCs to propellant residues. Sorption isotherms of military grade nitrocellulose for dissolved nitroglycerine (NG) or 2,4-dinitrotoluene (2,4-DNT) was measured in batch experiments and were determined to be SNG=102.39(±0.05)CNG0.916(±0.032) and S2,4-DNT=103.08(±0.01)C2,4-DNT0.668(±0.010) (S and C in mg/kgnitrocellulose and mg/Lwat, respectively). Solid-to-water partitioning for NG and 2,4-DNT was 100 times greater in propellant residues than in typical military ranges soils. Since nitrocellulose can sorb NG and 2,4-DNT up to 23 and 5% of its mass, respectively, it can slow down, through retarded diffusion, the leaching of PCs from fired residues over the typical composition ranges of common propellants. The slow leaching of PCs from propellant grains in column studies can be better interpreted by considering their sorptive interaction with nitrocellulose in addition to dissolution kinetics. With nitrocellulose as the carrying matrix, residue-bound PCs may migrate farther and persist longer in subsurface environment.

Plants growing in the soils at military ranges and surrounding locations are exposed, and potentially able to uptake, munitions compounds (MCs). The extent to which a compound is transferred from the environment into organisms such as plants, referred to as bioconcentration, is conventionally measured through uptake experiments with field/synthetic soils. Multiple components/phases that vary among different soil types and affect the bioavailability of the MC, however, hinder the ability to separate the effects of soil characteristics from the MC chemical properties on the resulting plant bioconcentration. To circumvent the problem, this work presents a protocol to measure steady state bioconcentration factors (BCFs) for MCs in barley (Hordeum vulgare L.) using inert laboratory sand rather than field/synthetic soils. Three MCs: 2,4,6 trinitrotoluene (TNT), 2,4-dinitrotoluene (2,4-DNT), and 2,4-dinitroanisole (2,4-DNAN), and two munition like compounds (MLCs): 4-nitroanisole (4-NAN) and 2-methoxy-5-nitropyridine (2-M-5-NPYNE) were evaluated. Approximately constant plant biomass and exposure concentrations were achieved within a one month period that produced steady state log BCF values: 0.62 +/- 0.02, 0.70 +/- 0.03,1.30 +/- 0.06, 0.52 +/- 0.03, and 0.40 +/- 0.05 L kg(plant)(-1) dwt for TNT, 2,4-DNT, 2,4-DNAN, 4-NAN, and 2-M-5-NPYNE, respectively. Furthermore, results suggest that the upper bounds of the BCFs can be estimated within an order of magnitude by measuring the partitioning of the compounds between barley biomass and water. This highlights the importance of partition equilibrium as a mechanism for the uptake of MCs and MLCs by barley from interstitial water. The results from this work provide chemically meaningful data for prediction models able to estimate the bioconcentration of these contaminants in plants. (C) 2017 Elsevier Ltd. All rights reserved.

We investigated the bioconcentration potential of nitrogen-based compounds 4-nitroanisole (4-NAN), 3,5-dinitro-o-toluamide (3,5-DNoTAME), and 2-methoxy-5 nitropyridine (2-M-5-NPYNE) using earthworm Eisenia andrei exposures in aqueous exposure media in sand. Separate toxicity studies were conducted prior to bioconcentration studies using a range of chemical concentrations to establish the sublethal exposure conditions for the earthworms. The objectives of the present studies were to: (1) develop an experimental test system for estimating bioconcentration potentials of new and emerging munition compounds that partition into earthworms, using aqueous exposure media; and (2) apply this experimental model to establish original bioconcentration data for 4-NAN, 3,5-DNoTAME, and 2-M-5-NPYNE. Experimental design includes earthworm exposures to chemicals for up to 14 days in aqueous media (Rombke medium; 0.08 mM KCl, 2 mM CaCl2, 0.5 mM MgSO4, and 0.8 mM NaHCO3) in the presence of water-washed coarse sand (0.5-1.0 mm) substrate. Concentrations of test chemicals in respective exposure media remained relatively stable during these independent studies. Tissue analyses revealed a rapid uptake of each chemical by the earthworms; a steady state was attained within 24 h from commencement of these exposures. Estimated steady-state bioconcentration factors (BCFss; mL/g dry tissue) were 47, 6, and 11 for 4-NAN, 3,5-DNoTAME, and 2-M-5-NPYNE, respectively. These results will contribute to the BCF database being developed for use in models aimed at predicting environmentally significant parameters for new munition compounds in soil. Published by Elsevier B.V.

There is concern about the environmental fate and effects of munition constituents (MCs). Polyparameter linear free energy relationships (pp-LFERs) that employ Abraham solute parameters can aid in evaluating the risk of MCs to the environment. However, poor predictions using pp-LFERs and ABSOLV estimated Abraham solute parameters are found for some key physico-chemical properties. In this work, the Abraham solute parameters are determined using experimental partition coefficients in various solvent-water systems. The compounds investigated include hexahydro-1,3,5-trinitro-1,3,5-triazacyclohexane (RDX), octahydro-1,3,5,7-tetranitro-1,3,5,7-tetraazacyclooctane (HMX), hexahydro-1-nitroso-3,5-dinitro-1,3,5-triazine (MNX), hexahydro-1,3,5-trinitroso-1,3,5-triazine (TNX), hexahydro-1,3-dinitroso-5- nitro-1,3,5-triazine (DNX), 2,4,6-trinitrotoluene (TNT), 1,3,5-trinitrobenzene (TNB), and 4-nitroanisole. The solvents in the solvent-water systems are hexane, dichloromethane, trichloromethane, octanol, and toluene. The only available reported solvent-water partition coefficients are for octanol-water for some of the investigated compounds and they are in good agreement with the experimental measurements from this study. Solvent-water partition coefficients fitted using experimentally derived solute parameters from this study have significantly smaller root mean square errors (RMSE = 0.38) than predictions using ABSOLV estimated solute parameters (RMSE = 3.56) for the investigated compounds. Additionally, the predictions for various physico-chemical properties using the experimentally derived solute parameters agree with available literature reported values with prediction errors within 0.79 log units except for water solubility of RDX and HMX with errors of 1.48 and 2.16 log units respectively. However, predictions using ABSOLV estimated solute parameters have larger prediction errors of up to 7.68 log units. This large discrepancy is probably due to the missing R2NNO2 and R2NNO2 functional groups in the ABSOLV fragment database.