Introduction
Nitroaromatics comprise an important class of potential environmental contaminants because of their wide use as agrochemicals, textile dyes, munitions and other classes of industrial chemicals. As a result, there has been considerable interest in determining the fate pathways of nitroaromatics in soil and aquatic environments. Nitroaromatics are known to be susciptible to reductive transformations to aromatic amines in anoxic environments (1,2). Because of the general concern for the ecotoxicity exhibited by aromatic amines, there is a strong impetus for developing models for the reduction of nitroaromatics.
A number of environmentally significant nitroaromatics contain two or more nitro groups. Examples of such chemicals are illustrated here. For environmental assessment needs, the ability to predict the initial site of reduction is important because the regiomeric reaction products can exhibit very different chemical (3) and toxicity (4) characteristics. For polynitroaromatics, the rate of reduction for the first nitro group is generally much more rapid than the rate of reduction of the remaining nitro groups. Understanding the specifity with which the nitro groups are reduced allows prediction of what longer-lived nitroaromatic amines should be expected to contaminate soils and groundwater.
Pictured are two schemes illustrating the mechanism of nitro group reduction and its selectivity in polynitroaromatic cases (5).
Laboratory studies have demonstrated that the reduction of polynitroaromatics can be selective and thus result in the regiospecific formation of nitroaromatic amines that are resistant to further reduction. For example, Weber and Adams (6) observed that the chemical and sediment-mediated reduction af 2-bromo-4,6-dinitroaniline, an important textile dye intermediate, was almost entirely selective for the nitro group in the 4-position. The resulting reaction product, 3-bromo-5-nitro-1,2-diaminobenzene, did not undergo further reduction. Only a trace amount of the substrate resulting from reduction of the nitro group in the 6-position was observed in anoxic sediment-water systems. Similar results have been observed for the chemical (7,8) and enzymatic (9) reduction of 2,4-dinitroaniline and 2,4-dinitrophenol. By contrast, the chemical (7), sediment-mediated (3) and microbial reduction (10) of 2,4,6-trinitrotoluene occurs at the nitro group in the 4-position.
We report here experimentally measured reduction selectivities for a number of polynitroaromatics. Two sets of experimental reaction conditions were assayed. All compounds discussed below were reduced in homogeneous aqueous solutions containing sulfide (which has been proposed to be an important reductant in anoxic environments (2)) and micromolar amounts of jugalone. Jugalone is a quinone that is commonly used as a model for naturally occuring dissolved organic matter; laboratory studies have demonstrated that jugalone acts as an electron-mediator to shuttle electrons between sulfide and nitroaromatics (11). An electron shuttle system involving dissolved organic matter has been postulated to account for the facile reduction of nitroaromatics in natural systems. In addition, data are presented for reductions in reducing Cherokee Park sediments.
Theoretical work is presented that rationalizes the observed selectivities. Based on the mechanism of nitro group reduction, the step that establishes the regioselectivity is the first proton transfer to the radical anion that is created after the initial one-electron reduction of the starting material. The nitro group that is selectively protonated then goes on to be completely reduced. Given this analysis, localization of charge in the initial radical anion, assuming it occurs, will influence the site of protonation. A key point to be made is that, in the gas phase, charge would not necessarily be expected to localize because of the significant cost of Coulomb repulsion between the electrons. However, in aqueous solution, the favorable effects of solvation can overcome this cost, so charge localization at one nitro group more than at the other seems reasonable. Thus, we present calculations using the AM1-SM2 solvation model to determine preferred geometries and electronic structures of the polynitroaromatic radical anions in aqueous solution. Visualization of the electrostatic potential on the van der Waals surface provides considerable insight into charge localization, which correlates extremely well with the experimental results.
We begin by presenting the various electrostatic potential surfaces. Following this, we discuss several details of the theoretical work, including (i) the utility of semiempirical calculations in these systems by comparison to ab initio, (ii) the effects of solvation on geometries and electronic structures, and (iii) the situation observed when experiment indicates little selectivity in the reduction process. Theoretical and experimental methodologies are detailed following the Discussion section.
Return to Table of Contents
Results
A number of polysubstituted aromatics, including either two or three nitro
groups as substituents, were studied in order to assess computational methods for predicting
regioselectivity in nitro group reduction. Experimental selectivities were
taken from the literature and/or separately measured as described in the
Methods section.
Presented below are figures for ten of the eleven polynitroaromatics
studied (the remaining one is found in the Discussion section). Each contains a line structure to the left, and a ball-and-stick model
illustrating the stereochemistry of the AM1-SM2-optimized structures in aqueous
solution to the right. In the center of each figure the electrostatic potential
is color-mapped onto the 0.002 au
isodensity surface of the optimized structure. This isodensity surface roughly
corresponds to a van der Waals surface, i.e., the depiction is a space-filling
one. The electrostatic potential, calculated as described in the Methods
section, is mapped onto the spectrum so that red represents regions having
maximal negative potential for a positive point charge, while blue regions are
least attractive to a positive point charge. The maximum and minimum magnitude
values, corresponding to the red and blue extremes of the spectrum,
respectively, are listed beneath each figure, together with particular details
associated with the individual molecule (e.g., changes in geometry on going
from the gas-phase to aqueous solution, multiple low-energy geometries, etc.),
2,4-Dinitroaniline (Solvation effects on the geometry are detailed in the Discussion section.)
2-Bromo-4,6-dinitroaniline
2,4-Dinitroanisole (Solvation
effects on the geometry are detailed in the Discussion section)
2,4-Dinitrobenzaldehyde (Solvation effects
on the geometry are detailed in the Discussion section.)
Visualization of the electrostatic potential is critical to accurately predict the selectivity of reduction in this molecule, since a simple analysis of the total charge on each individual nitro group would lead one to predict the para group to be more likely to be reduced--see Discussion section.
2,4-Dinitrobromobenzene As illustrated in this figure, there are two low-energy conformations for this radical anion. Each predicts a different selectivity for reduction. Although the structure localizing charge on the ortho nitro group lies only 2.9 kcal/mol higher in energy than the para alternative, this is inconsistent with the experimental observation of complete selectivity for the ortho position. (see Discussion section for details.)
2,4-Dinitrochlorobenzene Again, two low-energy structures are predicted which would predict different regioselectivities. In this case, however, the experimental data also indicate a mixture of reduction products. We note that for this system, the para-localized radical anion is 1.0 kcal/mol lower in energy, while the dominant experimentally observed product is from ortho reduction. However, these methods should not in any case be expected to be quantitative. Rather, visualization of the electrostatic potential provides a qualitative estimate of reduction selectivity (see Discussion section for details.)
2,4-Dinitrophenol (Solvation
effects on the geometry are detailed in the Discussion section.)
2,4-Dinitrotoluene
2,4,6-Trinitrotoluene
2-Amino-4,6-dinitrotoluene This molecule is the result of a (deliberate) ortho-reduction of 2,4,6-trinitrotoluene.
Return to Table of Contents
A number of environmentally significant nitroaromatics contain two or more nitro groups. Examples of such chemicals are illustrated here. For environmental assessment needs, the ability to predict the initial site of reduction is important because the regiomeric reaction products can exhibit very different chemical (3) and toxicity (4) characteristics. For polynitroaromatics, the rate of reduction for the first nitro group is generally much more rapid than the rate of reduction of the remaining nitro groups. Understanding the specifity with which the nitro groups are reduced allows prediction of what longer-lived nitroaromatic amines should be expected to contaminate soils and groundwater.
