We have used solid-state Au/Hg voltammetric electrodes to understand redox and biogeochemical processes in hot spring and deep sea hydrothermal environments. These electrodes are non-specific and have the capability of measuring simultaneously a suite of chemical species including several of the principal redox species involved in early diagenesis (O-2, Mn2+, Fe2+, H2S/HS-, and I-) as well as some Fe species (FeS and Fe3+) and sulfur species (S-x(2-) and S2O32-). Here we demonstrate how in situ data obtained in complex environments can be used to study specific iron and sulfur reactions and processes at (sub)millimeter to centimeter resolution and over short time scales. Examples include the oxidation of Fe2+ by O-2 produced by cyanobacterial mats in Yellowstone National Park hot springs and the formation of S2O32- in diffuse flow waters from the hydrothermal vents at Lau Basin. In one example, profiles of redox species in cyanobacterial mats from Yellowstone National Park hot springs show that in the light dissolved Fe2+ is completely removed from the source waters as cyanobacterial mats produce O-2 and oxidize the Fe2+. Performing kinetic experiments in the dark and light at the depth of maximum O-2 production indicates that the decay of Fe2+ follows a zero order rate law consistent with photosynthesis as the source of 0, These dynamic environments show how kinetic data can be obtained in situ and be used to understand the interactions between biology and chemistry. We know of no other analytical technique that can provide this information in both clear and turbid waters on the time scales (seconds) observed.

We studied the role of microbial photosynthesis in the oxidation of Fe(II) to Fe(III) in a high Fe(II) and high Mn(II) hot spring devoid of sulfide and atmospheric oxygen in the source waters. In situ light and dark microelectrode measurements of Fe(II), Mn(II) and O2 were made in the microbial mat consisting of cyanobacteria and anoxygenic photosynthetic Chloroflexus sp. We show that Fe(II) oxidation occurred when the mat was exposed to varying intensities of sunlight but not near infrared light. We did not observe any Mn(II) oxidation under any light or dark condition over the pH range 5–7. We observed the impact of oxygenic photosynthesis on Fe(II) oxidation, distinct from the influence of atmospheric O2 and anoxygenic photosynthesis. In situ Fe(II) oxidation rates in the mats and cell suspensions exposed to light are consistent with abiotic oxidation by O2. The oxidation of Fe(II) to form primary Fe(III) phases contributed to banded iron-formations (BIFs) during the Precambrian. Both oxygenic photosynthesis, which produces O2 as an oxidizing waste product, and anoxygenic photosynthesis in which Fe(II) is used to fix CO2 have been proposed as Fe(II) oxidation mechanisms. Although we do not know the specific mechanisms responsible for all Precambrian Fe(II) oxidation, we assessed the relative importance of both mechanisms in this modern hot spring environment. In this environment, cyanobacterial oxygen production accounted for all the observed Fe(II) oxidation. The rate data indicate that a modest population of cyanobacteria could have mediated sufficient Fe(II) oxidation for some BIFs.

We have used solid‐state Au/Hg voltammetric electrodes to understand redox and biogeochemical processes in hot spring and deep sea hydrothermal environments. These electrodes are non‐specific and have the capability of measuring simultaneously a suite of chemical species including several of the principal redox species involved in early diagenesis (O 2 , Mn 2+ , Fe 2+ , H 2 S/HS − , and I − ) as well as some Fe species (FeS and Fe 3+ ) and sulfur species (S x 2− and S 2 O $\rm{ _3^{2 - } )}$ . Here we demonstrate how in situ data obtained in complex environments can be used to study specific iron and sulfur reactions and processes at (sub)millimeter to centimeter resolution and over short time scales. Examples include the oxidation of Fe 2+ by O 2 produced by cyanobacterial mats in Yellowstone National Park hot springs and the formation of S 2 O $\rm{ _3^{2 - } }$ in diffuse flow waters from the hydrothermal vents at Lau Basin. In one example, profiles of redox species in cyanobacterial mats from Yellowstone National Park hot springs show that in the light dissolved Fe 2+ is completely removed from the source waters as cyanobacterial mats produce O 2 and oxidize the Fe 2+ . Performing kinetic experiments in the dark and light at the depth of maximum O 2 production indicates that the decay of Fe 2+ follows a zero order rate law consistent with photosynthesis as the source of O 2 . These dynamic environments show how kinetic data can be obtained in situ and be used to understand the interactions between biology and chemistry. We know of no other analytical technique that can provide this information in both clear and turbid waters on the time scales (seconds) observed.

Solid-state voltammetric (micro)electrodes have been used in a variety of environments to study biogeochemical processes. Here we show the wealth of information that has been obtained in the study of sediments, microbial mats, cultures and the water column including hydrothermal vents. Voltammetric analyzers have been developed to function with operator guidance and in unattended mode for temporal studies with an in situ electrochemical analyzer (ISEA). The electrodes can detect the presence (or absence) of a host of redox species and trace metals simultaneously. The multi-species capacity of the voltammetric electrode can be used to examine complex heterogeneous environments such as the root zone of salt marsh sediments. The data obtained with these systems clearly show that O2 and Mn2+ profiles in marine sedimentary porewaters and in microbial biofilms on metal surfaces rarely overlap indicating that O2 is not a direct oxidant for Mn2+. This lack of overlap was suggested originally by Joris Gieskes' group. In waters emanating from hydrothermal vents, Fe2+, H2S and soluble molecular FeS clusters (FeSaq) are detected indicating that the reactants for the pyrite formation reaction are H2S and soluble molecular FeS clusters. Using the ISEA with electrodes at fixed positions, data collected continuously over three days near a Riftia pachyptila tubeworm field generally show that O2 and H2S anti-correlate and that H2S and temperature generally correlate. Unlike sedimentary environments, the data clearly show that Riftia live in areas where both O2 and H2S co-exist so that its endosymbiont bacteria can perform chemosynthesis. However, physical mixing of diffuse flow vent waters with oceanic bottom waters above or to the side of the tubeworm field can dampen these correlations or even reverse them. Voltammetry is a powerful technique because it provides chemical speciation data (e.g.; oxidation state and different elemental compounds/ions) as well as quantitative data. Because (micro)organisms occupy environmental niches due to the system's chemistry, it is necessary to know chemical speciation. Voltammetric methods allow us to study how chemistry drives biology and how biology can affect chemistry for its own benefit.

The oxidation and precipitation of H2S were investigated in Torquay Canal and Bald Eagle Creek, two tributaries of northern Rehoboth Bay, one of the Delaware Inland Bays. These man-made dead end canals develop seasonal anoxia and have been the site of past fish kills and harmful algal blooms. The canals have multiple holes over 5.5m deep compared to an average low tide depth of 2m. In situ determination for dissolved O2, H2S and other Fe and S redox species were conducted with a solid-state Au/Hg microelectrode in 2003 and 2004. Laboratory analyses of discrete samples were also performed to measure dissolved and particulate Fe, Mn, and S8 to follow the seasonal dynamics of O, S, Fe and Mn redox species. Our results indicate that the water in the holes becomes stratified with O2 decreasing with depth and H2S increasing with depth. Dissolved Fe was as high as 30μM whereas dissolved Mn was only 0.2μM in the water column, indicating that Fe is the dominant metal involved in S redox cycling and precipitation. In surface oxic waters, the dominant form of Fe was particulate Fe(III) (oxy)hydroxides. When seasonal anoxia developed, Fe(III) (oxy)hydroxides were reduced by H2S to Fe(II) at the oxic–anoxic interface. The Fe(II) reduced from particulate Fe can be re-oxidized to Fe(III) by O2 above and at the interface to form a catalytic cycle to oxidize H2S. Elemental S is the predominant oxidation product and was as high as 30μM level (as S0) at the interface. When the system was stable, the Fe catalytic cycle prevented H2S from being released into surface waters during seasonal anoxia. However, when storms came, the water column was overturned and H2S was released to the surface water. The reaction rates for the Fe catalytic cycle are not fast enough and the concentration of Fe was not high enough to regulate the high concentration of H2S in surface waters during storm and mixing events.

The suitability of the bismuth film working electrode was investigated as an alternative to the gold-amalgam electrode used in solid state microelectrodes for in situ voltammetric analysis of redox chemistry in the natural environment (e.g., lakes, oceans, sedimentary pore waters). Chemical redox species measurable with the Au-amalgam include O-2, H2S, S-x(2-), S2O32-, Fe2+, Fe3+, Mn2+, I-. Bismuth was electrochemically deposited to form a solid film analogous to the Au-amalgam on a polished gold disk electrode. The useable potential window of the Bi-film was found to be narrower than that of the Au-amalgam, precluding the detection of dissolved O-2, I- and S2O32-, whose redox potentials fall outside the Bi-film's range. The Bi-film was able to detect free H2S/HS- and total sulfide (AVS), but not Fe2+ or Mn2+. The Bi-film was less sensitive to low levels of total sulfide (< 10 mu M) than the Au-amalgam; however, the Bi-film was able to accurately quantify very high concentrations of sulfide (at least 15 mM), with a linear response up to an order of magnitude higher than that of the Au-amalgam. Thus the Bi-film appears to have limited application as an alternative to the Au-amalgam microelectrode for in situ analysis of redox species in natural waters.

Soluble manganese(III) [Mn(III)] has been thought to disproportionate to soluble Mn(II) and particulate Mn(IV)O2 in natural waters, although it persists as complexes in laboratory solutions. We report that, in the Black Sea, soluble Mn(III) concentrations were as high as 5 micromolar and constituted up to 100% of the total dissolved Mn pool. Depth profiles indicated that soluble Mn(III) was produced at the top of the suboxic zone by Mn(II) oxidation and at the bottom of the suboxic zone by Mn(IV)O2 reduction, then stabilized in each case by unknown natural ligands. We also found micromolar concentrations of dissolved Mn(III) in the Chesapeake Bay. Dissolved Mn(III) can maintain the existence of suboxic zones because it can act as either an electron acceptor or donor. Our data indicate that Mn(III) should be ubiquitous at all water column and sediment oxic/anoxic interfaces in the environment.

We coupled an in situ electrochemical analyzer to a CTD pump profiler system to measure redox species across the oxic–anoxic interface of the Black Sea water column. Voltammetry was performed using gold–amalgam working electrodes to measure simultaneously oxygen and dissolved sulfur species (S8, Sx2-, HS−/H2S) both in situ (at <1 m vertical intervals) and in an on-deck flow cell attached to the outflow of a pump profiler (vertical resolution of about 1.5 m). In situ data agreed with measurements made in the flow cell and with measurements made from samples collected by rosette bottle casts. In situ voltammetry provided undisturbed, high-resolution measurements, and revealed significant yet subtle features not seen by traditional methods because of small spatial separation between the features and the measurements. Layers of oxygen intrusion (<5 m thick, from 10 to 150 μM O2) were present within the suboxic zone of the southwest Black Sea that are not present in the west-central and northeast Black Sea. Oxygen injection also occurs at other depths throughout the southwest and corresponds with small temperature anomalies, suggesting influence by Bosporus inflow up to 150 km from its entrance to the Black Sea. Such an inflow of oxygen, as well as spatial variations of the halocline, affect both manganese and, subsequently, sulfide oxidation for a large portion of Black Sea intermediate water (H2S onset occured ∼60 m deeper in the southwest as compared to the west-central). In situ voltammetric analyses provided rapid redox information, thus enabling more accurate targeting of specific geochemical features by the CTD rosette package.

We coupled an in situ electrochemical analyzer to a CTD to conduct high-resolution, real-time profiling of redox species across the oxic–anoxic transition zone of the Black Sea water column. Voltammetry was performed using gold–amalgam working electrodes to measure simultaneously soluble oxygen and sulfur species (H2S/HS−, Sx2-, S8) at a resolution of greater than one measurement per meter. In situ data agreed with measurements made in an on-deck voltammetry flow cell coupled to a pump profiling system, and from water samples collected with conventional CTD rosette bottle casts. In situ voltammetric analyses provided rapid redox information, thus enabling more accurate targeting of specific geochemical features by the CTD rosette package. We observed much less lateral oxygen injection from the Bosphorus in 2003 (less than 95km from Bosphorus) than in 2001 (up to 150km). This difference can be attributed to variability in physical processes including seasonal temperature and wind variations between winter conditions (2003) and early summer conditions (2001). Furthermore, suboxic zone thickness varied basin-wide, exhibiting changes in the depth of oxygen extinction (minimum detection limit=3μM) and sulfide onset (minimum detection limit=30nM). The density surface for oxygen extinction was more variable than the density for the onset of sulfide. Vertical shifts in oxygen extinction and sulfide onset also were observed at the western central gyre station for seven profiles measured over 21 days in 2003.

In 2001, the development of seasonal anoxia was studied in two waterways located at the head of Delaware’s northern inland bay, Rehoboth Bay. Bald Eagle Creek is a northern tributary of the bay, which has tidal exchange with Torquay Canal (a dead-end canal) via a short channel with a 1.4 m sill. Mean low water depth in Torquay Canal is about 2 m, but dredging produced over a dozen depressions with a total water depth of 5.5 m. During the summer of 2000, four major fish kills were reported in Torquay Canal and Bald Eagle Creek with more than 2.5 million juvenile menhaden ( Brevoortia tyrannus ) killed. Low O 2 concentration was assumed to be the problem but production of toxic H 2 S is more likely. From late spring 2001, we conducted in situ determination of temperature, salinity, pH, dissolved O 2 , and H 2 S in Torquay Canal and Bald Eagle Creek. During spring, water column stratification began in the depressions with warmer and less salty water observed in the upper layer, and cooler, saltier water below 2 m. O 2 was at saturation levels in the surface waters but was not detectable below 2 m by the end of May. The depressions were anoxic with H 2 S accumulating to mM concentrations in June. A storm event prior to July 12 mixed these two layers with a subsequent loss of H 2 S. The H 2 S levels again increased in the deep water due to stratification and reached another maximum in late August. Another storm event occurred at this time resulting in no detectable O 2 and up to 400 μM H 2 S in surface waters. H 2 S appears to be the primary reason for fish kills in these tributaries. Aerators installed in Torquay Canal on June 21 had no significant effect on abating stratification and anoxic conditions beyond their immediate area.

A four month study of a man-made lake used for hydroelectric power generation in northeastern Pennsylvania USA was conducted to investigate seasonal anoxia and the effects of sulfide species being transported downstream of the power generation equipment. Water column analyses show that the system is iron-rich compared to sulfide. Total Fe(II) concentrations in the hypolimnion are typically at least twice the total sulfide levels. In situ voltammetric analyses show that free Fe(II) as [Fe(H 2 O) 6 ] 2+ or free H 2 S as H 2 S/HS - are either not present or at trace levels and that iron-rich sulfide complexes are present. From the in situ data and total Fe(II) and H 2 S measurements, we infer that these iron-rich sulfide complexes may have stoichiometries such as Fe 2 SH 3+ (or polymeric forms of this and other stoichiometries). These iron-rich sulfide complexes appear related to dissolution of the iron-rich FeS mineral, mackinawite, because IAP calculations on data from discrete bottle samples obtained from bottom waters are similar to the pK sp of mackinawite. Soluble iron-sulfide species are stable in the absence of O 2 (both in lake waters and the pipeline) and transported several miles during power generation. However, iron-sulfide complexes can react with O 2 to oxidize sulfide and can also dissociate releasing volatile H 2 S when the waters containing them are exposed to the atmosphere downstream of the powerplant. Sediment analyses show that the lake is rich in oxidized iron solids (both crystalline and amorphous). Fe concentrations in FeS solids are low (<5 μmole/gr dry wt ) and the pyrite concentration ranges from about equal to the solid FeS to 30 times the solid FeS concentration. The degree of pyritization is below 0.12 indicating that pyrite formation is limited by free sulfide, which can react with the iron-rich sulfide complexes.