Redox and Nitrogen Speciation in Subsurface Waters
We expend considerable effort generating original data in the lab and field , but, to make our efforts useful to modelers and regulators, we strive to express our findings as mathematical relationships so that they can be incorporated into models.
I. Kinetic Control of Oxidation State at Thermodynamically Buffered Potentials in Subsurface Waters:
Together with temperature, pressure and pH, oxidation state is one of the chief state variables for defining the chemistry of a system. Unlike these other state variables, in complex settings such as the environment, definition of oxidation state often is ambiguous. Presently, the most widely accepted conceptual model for characterizing oxidation state is the terminal electron accepting process (TEAP). According to TEAP, microbes reduce oxidants in the environment sequentially from highest to lowest energy producing reductions; oxygen then nitrate then manganese (IV), and so on, ferric iron, sulfate and carbon dioxide. To evaluate the general applicability of TEAP, we collected samples from several environments and analyzed them for multiple redox-active species. To examine variation of redox couples through time, we also monitored the quality of a spring for two years.
| We subjected these data to thermodynamic analysis. The data for the monitored spring (Figure 1), as well as those of the other study sites showed that redox potentials tended to form two clusters. While this observation of clustering for thermodynamically calculated potentials is suggestive of a redox control, the occurrence of two potentials, as opposed to a single potential expected for equilibrium, is suggestive of kinetic constraints as well. |
|
| Figure 1: Redox potential vs. time in SpW2. |
| Slow reaction rates can be caused by either small rate constants or low reactant concentrations. To evaluate which of these variables is the dominant in constraining redox reactions we reasoned that if the reactions are reactant-rate limited, deviation from equilibrium should be a function of reactant concentration. Alternatively, if rates are constrained dominantly by small rate constants, a plot of reactant concentration vs deviation from equilibrium should yield a random shotgun-blast pattern. Further, if reactant concentrations constrain reaction rate, the potential of the most concentrated complementary reactant should approximate the equilibrium endpoint toward which a couple is drawn. Using this approximation, Figure 2 depicts deviation from equilibrium as a function of reactant concentration and clearly supports that concentration plays a large part in rate limitation for redox reactions.
Figure 2 also shows that reactions having reactant concentrations in excess of 1 mM are near equilibrium with their dominant complementary reactant. This observation suggests that redox reactions can proceed in parallel as opposed to sequentially as often conceived in the TEAP conceptual model. These efforts are reported in a manuscript submitted to Geochimica et Cosmochimica Acta. |
Figure 2: Deviation from equilibrium vs. reactant concentration. |
II. Forms of Nitrogen in Subsurface Waters:
Based on data we have collected for several sample locations, dissolved organic nitrogen that does not react with the urease enzyme represents a significant fraction of the total nitrogen pool in groundwater. The persistence of organic nitrogen in aquatic ecosystems has been a subject of considerable conjecture in recent years. Some suggest that large fractions of organic nitrogen are as recalcitrant heterocyclic functional groups such as pyridines, making the organic-nitrogen fraction recalcitrant. However, recent experimental Xanes work has shown that soil organic nitrogen is present dominantly as amide functional groups and 15N NMR has shown that organic nitrogen in seawater is, likewise, dominantly present as amide groups.
Though the free-energy of formation of these complex nitrogen-bearing organic compounds is highly variable and unknown, the free energy of reaction for breaking the amide bond likely is relatively invariant (Figure 3). Consequently, based on this recent experimental work showing organic nitrogen dominantly is as amides, we have modeled our dissolved organic nitrogen as the simple amide, urea, normalized to the free-energy of reaction for a single amide bond.
Figure 3: Amide group hydrolysis
(after Larson and Weber. 1994. Reaction Mechanisms in Environmental Organic Chemistry. Lewis Publishers.)
Using this approximation, organic nitrogen appears to be nearly in equilibrium with a cluster of oxidation potentials comprised of other nitrogen redox couples, oxygen-hydrogen peroxide and ferrous-ferric iron (Figure 4). This approximation suggests that organic nitrogen might be persistent because it nearly has equilibrated with dominant environmental oxidants rather than because it is functionally recalcitrant. A manuscript reporting this work is in progress.
Figure 4: Redox potential vs. time with Norg oxidation modeled as an amide group (see bold line).
Questions? Contact John Washington.