(82a) A Quantum Mechanical Approach to Predict One-Electron Reduction Potentials for Nitroaromatic Compounds

Nitroaromatic compounds (NACs) are widespread environmental contaminants due to their use as pesticides, explosives, and in a variety of other applications. The toxicity of many NACs and/or their environmental transformation products has prompted considerable interest in the fate of NACs in the environment. In particular, an understanding of reduction - the key transformation pathway - is critical for evaluating the risks associated with NAC-contamination and developing remediation strategies. The rate-determining step in the overall reduction of NACs is often considered to be the transfer of the first electron from a reductant to the NAC. The tendency for this first electron transfer to take place depends on the one-electron reduction potential, E^o_H. A thermodynamically favorable reaction requires that the E^o_H value of the NAC be less negative than that of the oxidized reductant under the environmental conditions. While E^o_H is a thermodynamic parameter, it can also reveal kinetic information, since E^o_H is found to be linearly correlated with the rate constant for the reduction of NACs under certain conditions. Therefore, knowledge of the E^o_H value for an NAC is important for modeling its environmental fate. Unfortunately, measurements are not available in many cases; for instance, the highly energetic nature of some explosive NACs makes experimental work difficult. In addition, some of the available data are subject to large uncertainties. Furthermore, a priori determination of E^o_H values is important for environmental impact assessment in the design of new chemicals. Therefore, a method is needed for reliably predicting E^o_H values for NACs that does not rely on any experimental data. In this work we have examined various quantum mechanical (QM) approaches for predicting E^o_H values for substituted nitrobenzenes in water (pH 7) at 25^oC. In the absence of empirical corrections, large deviations (in excess of 100 mV, compared with the reported typical measurement accuracy of 10 to 20 mV) were found between several of our predicted E^o_H values and the measured values reported in the literature. These errors can be attributed principally to inaccuracies in the solvation calculations, which is not surprising since such calculations are known to be challenging. Based on this finding, we propose an alternative method for predicting the E^o_H values that uses only gas-phase QM calculations. This approach relies on the strong correlation found between E^o_H values for NACs and their gas-phase analog, adiabatic electron affinities (EA). In the first step, EA values are calculated using density functional theory (DFT; B98/MG3S), and then scaled by a factor of 0.802 (empirical) to account for systematic errors in the DFT calculations. In this way, EA values can be predicted with both high accuracy and efficiency. (Alternatively, measured EA values may be used in place of the QM predictions, where they are available.) In the final step, E^o_H values are predicted from the empirical relationship E^o_H [mV] = 260.3 EA [eV] ? 740.2, which accounts for differences between the energetics of the gas-phase and aqueous-phase one-electron reduction processes. Using this method, E^o_H predictions for 25 substituted nitrobenzene compounds were all within 75 mV of the measured or estimated values reported in the literature, with a mean absolute deviation of 25 mV. No decrease in accuracy was observed for compounds outside of the training sets used to establish the empirical relationships. These findings support the general applicability of this method for accurate and efficient predictions of E^o_H values for substituted nitrobenzenes.