Ocean Optics NeoFox oxygen sensors can be calibrated and used even at extreme partial and total pressures, such as those found in sedimentary formations in the upper few hundred meters of the Earth’s crust. This application note describes studies on pyrite reaction kinetics conducted by the Department of Applied Geology at Kiel University, Germany, using a T1000-TS-NEO oxygen probe calibrated for use up to 11 bars partial pressure of oxygen. The research was carried out within the framework of the ANGUS+ joint research project on consequences of subsurface use, financed by the German Federal Ministry of Education and Research.
It is well known that the introduction of oxygen-rich water into aquifers containing pyrite (FeS2) leads to oxidation of the mineral, potentially increasing the acidity and the sulfate content of the water present. The reaction kinetics of pyrite oxidation are not well understood, however, for aquifers with high partial pressures of oxygen and near-neutral pH levels. These conditions may exist in sites under remediation using in situ chemical oxidation, in aquifers being used for temporary water storage, or in the case of leakage from a compressed air energy storage facility. Measurements under such conditions represented by lab experiments are the key to developing models of the impact and effects of high partial pressures of oxygen in these environments to understand potential side effects on groundwater resources.
Hydrostatic pressure can result in much higher partial pressures of oxygen than are seen at the surface. For example, at a depth of 50 m, the total air pressure can be ~5 bars, of which roughly 1 bar can be attributed to oxygen. Even higher dissolved concentrations of oxygen, in equilibrium with even higher partial pressures (p(O2)), can be seen in the case of leakage from Compressed Air Energy Storage (CAES) facilities, where compressed air is pumped into a subsurface reservoir at pressures up to 50 bars during periods of excess energy production, then released during energy lulls to drive turbines. A leakage of high pressure air from a CAES facility into a groundwater aquifer can quickly cause oxidation of the pyrite commonly present, deteriorating groundwater composition, and potentially influencing the quality of the drinking water produced from that aquifer.
Flow-through column experiments provide a good laboratory-based simulation of the conditions present in aquifers at different depths with corresponding hydrostatical pressures, enabling research into the effect of introducing high pressure air on the resulting water composition. In this study, the conditions studied simulated air intrusion in aquifers at up to 500 m depth, providing new information about pyrite oxidation kinetics at oxygen partial pressures from 0-11 bars.
Low and high pressure flow-through columns were created using a sediment composed of quartz sand and crushed pyrite, percolated by tap water from a local aquifer. The water was equilibrated with varying partial pressures (p(O2)) of oxygen, from 0-1 bar for the low pressure experiments, and from 1-11 bars for the high pressure experiments.
Water properties were measured throughout the studies, which lasted 164 days for the low pressure experiment, and 91 days for the high pressure experiments. The parameters measured continuously included dissolved oxygen concentration (partial pressure), redox potential, pH level, and specific electrical conductivity. Dissolved concentrations of anions, cations, and organic and inorganic carbon species were measured in samples typically taken daily. Since conventional Clark electrodes cannot accurately measure above ~1 bar, a different type of oxygen sensor was needed for the high pressure experiments. Fluorescence-based sensors like the NeoFox from Ocean Optics overcome this limit, looking instead at changes in the lifetime of the sensor fluorophore in the presence of oxygen to determine partial pressure.
Given the pressures involved, we recommended the rugged T1000-TS-NEO oxygen probe, which is ¼” in diameter and rated for pressures up to 3000 psi (207 bars) in process and high-pressure applications. It also contains an embedded thermistor for convenient temperature compensation in a single assembly. A BIFBORO-1000-2 bifurcated fiber assembly was used to connect the probe to the NeoFox phase fluorimeter for detection.
A special calibration function was also needed, developed in cooperation between the researchers at the Applied Geology group in Kiel and Ocean Optics technical staff in Germany and the U.S. The new function for this broad pressure range relates fluorescence lifetime (τ) to p(O2) for the range 0-11 bars (see calibration curve). This calibration function was tested at three known partial pressures to confirm accuracy.
Water samples taken were analyzed for concentration of anions using an ion chromatograph, and for concentrations of cations with an ICP-AES. The total dissolved concentrations of organic and inorganic carbon species were measured using a multipurpose TIC/TOC analyzer. Sediment samples were also taken for solid phase analysis at the conclusion of each experiment, and the microbial community of the water was also studied.
Low Pressure Experimental Results
At low oxygen partial pressures (0.21 and 1 bars), pyrite oxidation was evident immediately, as observed by a steady decrease in oxygen and increase in sulfate concentration along the flow path. All of the dissolved iron produced remained in the solid phase in the column, and no sulfur species except sulfate was found. As dissolved oxygen decreased, sulfate concentration increased nearly stoichiometrically. The pH remained constant after an initial slight decrease for all starting p(O2) values, including 0 bar, indicating no significant change in pH due to pyrite oxidation. Sulfate concentration increased linearly along the column, indicating that the reaction rate in the column is constant; this allowed the pyrite oxidation rate to be easily calculated. It was found that increasing the partial pressure of oxygen from 0.21 bar to 1 bar roughly doubled the rate of pyrite oxidation, acknowledging earlier suggestions from the literature.
High Pressure Experimental Results
At high oxygen partial pressures (1, 5 and 11 bars), pyrite oxidation was similarly indicated by both a decrease in oxygen and increase in sulfate concentration, varying linearly along the flow path. No dissolved iron species or sulfur species other than sulfate were observed. Unlike at low pressures, however, acidification declined, returning to near-neutral pH conditions. Each time water with a higher partial pressure of oxygen was introduced, the pyrite oxidation rate jumped suddenly, then decreased over time. The rate of decline of oxidation appeared higher for higher p(O2) values.
Scanning Electron Microscope observations of the pyrite grains acquired after ending the experiment showed the presence of a ~3 µm thick layer covering the pyrite grains, as well as 0.2-0.5 µm large pockmarks on the surface where this layer was absent, created most probably by microbial activity. This was observed for all experiments involving oxygen, prompting further analysis using a Raman microspectrometer. While the material could not be identified, it was determined not to be pyrite, nor any of the other minerals hypothesized in the literature.
Having observed both the development of a layer on the pyrite grains, and seeing a decreasing pyrite oxidation rate compared to the traditional models at elevated pressures included in the experiment, the model was modified. A new passivation term was added to the model, dependent on the mass of pyrite oxidized, the partial pressure of oxygen and three ﬁtting parameters. This passivation term has its basis in the fact that pyrite oxidation at near neutral pH conditions should result in the formation of Fe(III)-oxyhydroxide coatings on the surface of pyrite, significantly reducing the reaction rate for oxidation. While the suggested specific passivation term differs from those proposed previously, it was able to very accurately model the production of sulfate at the various oxygen partial pressures tested.
As seen in the figure above, the rate of pyrite oxidation decreases more rapidly at higher p(O2) values. The passivation layer forms completely within just a few days at 11 bars p(O2), after which the rate of sulfur production remains roughly constant. Though this behavior appears quite unusual, it is accurately predicted by the new model developed.
Unlike the established rate laws, which are dependent only on the partial pressure of oxygen and pH as variables, the model developed as a result of these measurements was extended by a passivation term. In doing so, the new model was able to reproduce independent experimental results acquired using different experimental setups. While traditional models theorize that an increase in oxygen partial pressure will accelerate pyrite oxidation rates, the new model found passivation to be the dominating factor under these conditions. The new model also resolves previously existing discrepancies between low-pressure and high-pressure pyrite oxidation, allowing results from low pressure experiments representing a specific CAES site to be extrapolated to higher pressure conditions accurately, thus improving the risk assessments needed to ensure the safety of groundwater supplies.
Without the accurate measurements of p(O2) at values from 1-11 bars provided by the NeoFox oxygen measurement sensor, it would not have been possible to verify a model accurately predicting the unusual behavior of pyrite under high pressure conditions such as those in aquifers and CAES facilities. Ocean Optics oxygen sensors are used in a wide range of applications, from the monitoring of biological samples, headspace gases and industrial slurries to cosmetics, foods and liquids in natural environments.