tubes (e.g., Pyrex #9826-13), which were chosen because they can be inserted directly into the sample chamber of a fixed-wavelength spectrophotometer (Spectronic 20, Bausch & Lomb). A volume of 8.5 mL of slurry nearly filled these tubes, thereby minimizing the potential for indicator reoxidation by O₂ from the headspace. To initiate a reaction, 75 mL of 1-mM indicator stock solution was added to each tube, and then the tubes were capped and thoroughly mixed by hand. For contact times of more than a few minutes, tubes were mixed continuously and gently by rotation (at 7.5 rpm) on a mechanical shaker.

In order to obtain quantitative concentration versus time data, it is desirable to have a way to periodically measure absorbance of the indicator without opening the culture tubes and risking exposure to atmospheric O₂. To do this, each tube was centrifuged to produce a clear supernatant, and absorbance of the supernatant was measured by inserting the whole tube directly into the spectrophotometer. Small amounts of sediment formed a small pellet at the bottom of the tube, which did not interfere with the light path of the spectrophotometer; however, with cohesive sediments, large amounts of sediment could be accommodated by centrifuging the pellet into the top of the tube. After the measurement, the pellet was resuspended by vortexing. This procedure complicates the definition of reaction time because the dye has little contact with the sediment while the sediment is compressed in a pellet. This effect can be corrected by counting only time when the sediment is suspended.

Indicator Reduction Kinetics with Method II. Quantitative determinations of indicator reduction kinetics were done with a new method that was developed to give more accurate results with a wider range of reducing sediments. In an anaerobic chamber, a wide-mouth pipette was used to dispense 10 mL of sediment slurry into large culture tubes. The reaction was initiated by adding 75 mL of the dye stock solution to a tube and mixing (briefly by hand). The mixture was immediately poured into the open end of a 10-mL Luer-Lok disposable syringe that contained a plug of glass wool and was fitted with a 0.45-mm syringe filter (25-mm Whatman Puradisc). This arrangement minimized plugging of the filter by the sediment. The syringe was used to force 5 mL of filtrate directly into a 10-mL screw-cap cuvette (10-mm-path-length optical glass) containing 75 mL of dilute 1% H₂O₂ (to quench dissolved reductants). The procedure was repeated for all of the desired contact times (measured from when the indicator and sediment were mixed together to when the filtrate was separated from the sediment into the cuvette). Finally, the sealed cuvettes were removed from the anaerobic chamber and the absorbance of the filtrates was measured from 340 to 700 nm with a scanning spectrophotometer. Justification for the details of this method is provided in the discussion of Figure 7.

Experiments & Results

Redox Indicators. A large number of redox indicators have been developed for various applications [20], but few of them are suitable for use with environmental materials due mainly to sorption and interfering background color. The color changes shown in Figures 2 and 3 illustrate some of the various effects that can be obtained when redox indicators are exposed to materials of environmental interest. Further interpretation of these effects is provided in the discussion that follows.

Indigo and its mono-, di-, tri-, and tetra-sulfonates, eq 3, constitute a homologous series of indicators, all of which provide strong contrast between oxidized (blue) and reduced (yellow) forms. Indigo is the original dye used to make blue denim, and indigo 5,5’-disulfonate (I2S) is the indicator commonly known as indigo carmine.

(3)

The sulfonated forms of indigo have high water solubility and low tendency to adsorb to solids, so color changes in the solution phase give a visible indication of redox conditions even in heterogeneous systems [13]. We prefer indigo 5,5’,7,7’-tetrasulfonate (I4S) because it has the least tendency to sorb to sediments, but I3S and I2S generally give satisfactory results.

Figure 2 illustrates the behavior of I4S in the presence of a reducing sediment. All tubes were prepared identically by mixing and then centrifuging briefly to form a sharp sediment–water interface. However, tubes A through F were prepared at time intervals such that A represents the shortest contact time (< 1 day) and F represents the longest contact time (ca. 1 week). The effect of this procedure is to simulate the gradual progression of a front of reduced dye (yellow) up the tube starting from the sediment water interface. Reoxidation was

Figure 2. Reduction of indigo tetrasulfonate (I4S) by an anaerobic sediment. The oxidized form of I4S is blue; the reduced form is yellow. Tubes A through F show progression of the reduced front up from the sediment–water interface with increasing contact time. The indicators and solids were mixed briefly, allowed to settle, and then not disturbed until the photograph was taken. Tube G was fully reduced, but the I4S began to reoxidize from above after the cap was loosened.

Figure 3. Color changes in two indicators (resazurin in tubes A thorugh C and phenosafaranine in tubes E thorugh G) in the presence of two reductants (iron metal in tubes B and F and anaerobic sediment in tubes C and E). Controls are A (resazurin only), D (sediment only), and G (phenosafranine only). The indicators and solids were mixed briefly, allowed to settle, and then not disturbed for several days until the photograph was taken.

initiated by loosening the cap of tube G, thereby illustrating the reversibility of the color change corresponding to eq 3.

Phenosafranine, by contrast, is a red dye that bears a positive charge, as does its colorless reduced form, eq 4. The redox couple is reversible below pH 3 and above pH 7.5, but between pH 3 and 7.5 the indicator undergoes an irreversible, secondary reaction that leads to drift in the measured reduction potential of the couple [20].

(4)

Because most minerals have a negative surface charge under environmental conditions, both forms of phenosafranine tend to adsorb to sediments by cation exchange. This makes loss of color from the solution phase difficult to interpret because it may be due to reduction or to sorption (or both). The challenge is illustrated by the tubes shown in the right half of Figure 3. In tube E, phenosafranine and sediment were mixed and allowed to settle, resulting in immediate decolorization of the water column. Tube F was mixed with iron metal, but this tube shows no sign of decolorization (c.f. the control, tube G). The combination of these results suggests that decolorization of phenosafranine in the presence of sediment is mostly due to adsorption, not reduction.

Adsorbed dyes may still be redox active, however, and the expected color change might be observed if enough indicator is provided to produce visible staining of the particles. The effect of sorption on the lability of real environmental contaminants to transformation reactions is an important practical issue that has received only a modest amount of study [21, 22]. Although adsorption is a complication when it occurs on sediments, the process can be turned to advantage when binding can be used to immobilize the indicator onto a physical probe, such as a fiber optic sensor [23].

Resazurin reduces in two steps, irreversibly to resorufin and then reversibly to dihydroresorufin, providing color changes from purple to pink to colorless, eq. 5. Reduction of resazurin (disappearance of purple) has been used to characterize bacterial contamination in milk [24], quality of semen [25], and to quantify microbial activity in sediments [26]. The reoxidation of dihydroresorufin (appearance of pink) is widely used in anaerobic microbiology to indicate contamination with oxygen [27]. In addition, dealkylation of resorufin ethers is used as an assay for enzyme activity [28, 29]

(5)

These uncharged molecules are moderately hydrophobic and therefore sorb to sediments by hydrophobic partitioning to organic matter. Changes from strongly colored to weakly or uncolored (leuco) forms can provide ambiguous results in heterogeneous media because loss of color may also be due to adsorption. Figure 3 illustrates the interpretation of color changes involving resazurin in the presence of reducing sediment and granular iron metal. In both cases, resazurin (purple) was reduced to resorufin (pink) before the particles settled (c.f., tubes B and C with the control, tube A). The intermediate color in the upper part of tube B probably reflects partial reduction due to more rapid settling and/or lower reactive surface area of the iron metal. The red band at the bottom of tube B was observed on several occasions and may be due to partially overlapping gradients of pH and redox potential

Figure 4. Stability (or Pourbaix) diagram showing equilibria involving water, iron, and indigo tetrasulfonate at conditions of natural waters (total [Fe] = 10^–6 M, 25°C). Hematite (a-Fe₂O₃) and magnetite (Fe₃O₄) are assumed to be the controlling solid phase for iron speciation, but amorphous Fe(OH)₃ would give qualitatively similar implications with respect to reaction with the indicators. Data were taken from [31, 32] for the inorganic species, and [20] for the indicators.

(recall that the corrosion of Fe⁰ favors increased pH [30]). Clearly, even simple, qualitative experiments such as those illustrated by Figures 2 and 3, can pose challenging and edifying problems of interpretation for students.

Thermodynamics of Reduction. The speciation of most redox indicators is complicated by the presence of ionogenic functional groups in the oxidized and/or reduced forms. For example, the reduced form of I4S has two hydroxyl groups that are subject to deprotonation, resulting in three reduced species: the hydroquinone, the monophenoxide, and the diphenoxide. The relative stabilities of these species are represented as a function of potential (E) and pH in Figure 4.

Each line in Figure 4 represents a transition from predominance of one species to predominance of another. For example, the oxidized (blue) form of I4S is stable above line A, and the various reduced (yellow) forms are stable below line A. The lines dividing stability fields are drawn from the Nernst and Henderson-Hasselbach equations. For example, line A derives from the Nernst equation for the indigo dyes, eq 6 [20],

      (6)

using the formal potential  and dissociation constants of the reduced form (K_r1, K_r2) of I4S (Table 1), and assuming that the total concentration of all oxidized forms (C_o) equals the total concentration of all reduced forms (C_r). Line B corresponds to the first pK_a for reduced I4S (pK_r1) and is derived from the Henderson–Hasselbach equation using K_r1 for

                                                               (7)

I4S from Table 1 and assuming equal concentrations of the protonated (C_a) and deprotonated (C_b) forms. Line C reflects the second dissociation of the reduced form of I4S and is drawn from eq 7 using the corresponding value of pK_r2(Table 1).

Superimposed on the stability fields for I4S, we have drawn the stability limits for various forms of iron using data from standard aquatic chemistry texts [e.g., 1, 31]. Comparison of the two reveals that the iron couple has a lower formal potential than the I4S couple at pH values greater than 6.1 (where line A crosses the line dividing Fe²⁺ from Fe₂O₃). Therefore, reduction of I4S by Fe(II) is thermodynamically favorable in this region, and it may be contributing to the color change observed in the presence of anaerobic sediment (recall Figure 2). The stability region for iron metal lies below –0.61 V (assuming the total [Fe] = 10^–6 M) at all pHs, so Fe⁰ is a possible reductant of I4S under all experimental conditions. The resazurin/resorufin couple could not be added to Figure 4 because the half reaction is not reversible and a standard potential is not available; however, the rapid reduction of resazurin by anaerobic sediments and iron metal suggests that the half reaction has a potential that is higher than that of I4S. In fact, we found that resazurin is reduced by the reduced form of I4S but not the reduced form of methylene blue indicating that the reduction potential for resazurin falls between that of I4S and methylene blue.

Diagrams of E versus pH, such as Figure 4, are used quite commonly to summarize the thermodynamic properties of complex chemical systems [32–34]; however, they can contribute to misconceptions, particularly with respect to interpretation of the boundaries between stability fields. Sharp lines suggest sharp boundaries to many students, but these lines should be interpreted as continuous transitions from predominance of one species to predominance of another. This subtlety can be visualized with redox indicators such as I4S because the transition from the oxidized (blue) to the reduced (yellow) forms results in a transition zone that is green (this can be seen in Figure 2). Of course, the green transition zone also reflects a degree of disequilibrium due to kinetic controls on reactions in heterogeneous systems.

Kinetics of Reduction. The kinetics of reduction vary widely with the indicator, reducing medium, and other experimental variables. In fact, qualitative (visual) characterizations of the rate of indicator color change can be sufficient to provide students with evidence for the effects of reactant concentrations, mixing, etc. The importance of mixing is very apparent just by comparison of the color in sealed tubes set up with the same indicator and Fe⁰ but subject to various methods of shaking. More subtle manifestations of mixing effects can be seen in the zones of color in Figure 2 and Figure 3. For example, progression of the color transition up tubes A through F in Figure 2 could be due to diffusion of reduced I4S or some soluble reductant released from the sediment. The red band in tube B of Figure 3 is an even more challenging manifestation of diffusion effects on redox chemistry.

Diffusion-controlled effects such as these take days to develop, but they can be destroyed in minutes by moving the tubes from a laboratory bench to a sunny window. The resulting convection currents due to solar heating are particularly dramatic with the color change afforded by the indigo sulfonates.

Figure 5. Kinetics of indigo tetrasulfonate reduction by Fluka Fe⁰ (143 g/L) in a sealed, quartz cuvette with hand mixing between scans. Inset shows time courses for absorbance at the l_max for I4S in its oxidized (593 nm) and reduced (383 nm) forms. The exact l_max for each form is slightly affected by the presence of Fe²⁺ from corrosion of Fe⁰, and interactions between the dye and Fe²⁺ are also responsible for the peak at 700 nm. Fitting initial time course data to a first-order kinetic model (eqs 8 and 9) gives the smooth curves shown in the inset and rate constants given in the text.

Quantitative treatment of the reaction kinetics is possible by measuring absorbance as a function of time, either at one of the l_max values given in Table 1 or by performing full scans of absorbance from ca. 300 to 800 nm. In Figure 5, we illustrate how the former method is related to the latter, using data for the reduction of I4S by Fe⁰. The inset shows the resulting disappearance plot for the blue color of the oxidized form of I4S (l_max = 593 nm). The curve shows agreement between the data and a first-order kinetic model for decrease in absorbance, A,

                                                                         (8)

and a disappearance rate constant (k_obs) equal to 0.041 ± 0.002 min^–1 (t_1/2 = 17 min). Essentially identical values were obtained by non-linear regression using eq 8 and by linear regression using ln-transformed absorbance data (r² = 0.995). The rate constant could also have been evaluated without recourse to regression [35].

Note that we have only fit data for the first 20 minutes (n = 5), because substantial tailing is apparent after one half-life. The major reason for this tailing is probably that the tube was mixed intermittently and only when it was not in the spectrophotometer. For advanced students, it may be informative to contrast the kinetics of indicator reduction in this system (where the reaction is far from equilibrium) to the tailing observed with acid–base indicators in buffered solutions (where reversibility of the reaction is important) [36]. The validity (both fundamental and pedagogical) of analogizing between redox and acid–base processes has been discussed in detail previously [37–39]. Another extension of the treatment given here would be to develop a kinetic model that includes terms for sorption, as well as reduction, of the dye.

Appearance of the yellow color associated with reduced I4S (at l_max = 383 nm) is also shown in the inset to Figure 5. The initial data show good agreement to first-order appearance kinetics (eq 9)

                                                                (9)

using nonlinear regression to obtain k_obs equals 0.038±0.001 min^–1. The approximate agreement between the rate constants for appearance and disappearance of the two forms of I4S suggest that eq 3 accounts for most of the observed color change. Additional evidence that eq 3 alone is responsible for the color changes can be found in the consistent isosbestic points (where all spectra have the same absorptivity). Sharp isosbestic points are expected in simple systems where all absorbing components are linked by a single reaction [40]. To demonstrate material balance directly, absorbance data would have to be converted to molar concentrations using appropriate extinction coefficients or calibration curves.

Because values of k_obs can be determined quickly and routinely, teams of students can design their own experiments to compare the effects of various factors on k_obs. Some experimental variables that will influence k_obs include mixing rate and grain size of the solids in the medium. The net effect of pH may be challenging for students to interpret because pH can influence the indicator (e.g., most indicators have one or more pK_as, which makes their color and absorbance spectra pH-dependent), the reductant (e.g., corrosion of Fe⁰ is greatly accelerated as pH decreases), or the kinetics of reaction between the two. Figure 6 illustrates the effect of two other variables (degree of sulfonation and mass of reductant) on k_obs for reaction between derivatives of indigo and Fe⁰. For all three indicators, k_obs appears to vary linearly with the concentration of iron (mass of Fe⁰ per volume of solution), as has been observed previously for the reduction of pollutants such as carbon tetrachloride and vinyl chloride [41, 42].

Regression on the three sets of data in Figure 6 gives estimates of the slope for each line, and these slopes represent mass-normalized rate constants, k_m, for each indicator. The results, summarized in Table 2, indicate that the order of reactivity with Fe⁰ is I3S > I2S > I4S. The differences in reactivity are small, however, and this is consistent with previously reported results obtained with anaerobic sediment as the reducing medium [13]. Dividing each value of k_m by the specific surface area of the Fe⁰ used in this study gives surface-area-normalized rate constants, k_SA. The resulting values (Table 2) are comparable to those for dehalogenation of many chlorinated hydrocarbons [c.f., 19, 43].

The intercepts in Figure 6 are not significantly different from zero (statistical data not shown), suggesting that there is no reduction of the indicators in the absence of Fe⁰. Corrosion of Fe⁰ involves dissolution of Fe²⁺, and this species could serve as a dissolved phase reductant of other solutes [41]; however, studies with immobilized substrates suggest that contact with the metal surface is the dominant reduction pathway under conditions that are environmentally relevant [44]. With anaerobic sediments, results similar to Figure 6 have been obtained by varying the ratio of solids to water and measuring k_obs for the reduction of various contaminants [21, 45]. The conclusion from such studies is that most contaminant reduction reactions in sediments are predominantly surface

Table 2. Rate Data for Reduction of Indigo Dyes by Zero-Valent Iron

Oxidized form	kma	kSAb	r2	n	(V)
Disulfonate	0.059 ± 0.007	0.29 ± 0.04	0.957	5	–0.125
Trisulfonate	0.074 ± 0.003	0.37 ± 0.01	0.996	5	–0.081
Tetrasulfonate	0.047 ± 0.011	0.23 ± 0.06	0.769	7	–0.046

^aIn hr^–1 g^–1 L. ^bIn hr^–1 m^–2 L assuming specific surface area of the iron was 0.2 m²/g.

 

Figure 6. The effect of iron loading on the pseudo-first-order rate constants (k_obs) for reduction of indigo sulfonates. Each data point represents an independent experiment equivalent to the one shown in Figure 5. Linear regression on the k_obs versus [Fe] data gives the lines shown and mass-normalized rate constants (k_m) given in Table 2.

Figure 7 Kinetics of resazurin reduction by unfiltered pore water in a sealed, quartz cuvette with hand mixing between scans. Inset shows time courses for absorbance at the l_max of resazurin (600 nm) and resorufin (570 nm). Fitting initial time-course data to the first-order kinetic model (eqs 8 and 9) gives k_obs = 0.172 ± 0.002 min^–1 for resazurin disappearance and 0.19 ± 0.01 min^–1 for resorufin appearance.

reactions, and that there is little if any reaction occurring in the pore water [22]. Leading students to think about the chemistry of heterogenous systems is one of the major benefits of the experiments described in this manuscript.

Some redox indicators appear to be more labile to reduction than most contaminants, however, so they might provide a more sensitive probe for dissolved-phase reducing species in environmental systems. In fact, while testing the syringe filtration method (method II) on indicator/sediment combinations that gave complete reaction in as little as 1 min, we noted that the color of resazurin was not stable in 0.45-mm filtered pore water. The absorbance spectra in Figure 7 were taken on one of these samples, and they confirm that the observed color change was due to reduction of the indicator. Subsequent control experiments (not shown) proved that addition of H₂O₂ was effective at quenching the reduction of resazurin, and that H₂O₂ does not oxidize resorufin back to resazurin (recall from eq 5 that this reaction is generally not reversible). Therefore, we adopted a protocol that involved filtering roughly 5 mL of sediment pore water directly into a cuvette containing 75 mL of dilute H₂O₂. The method proved to be robust as long as the amount of dilution was controlled by taking care to add consistent amounts of filtrate to the cuvette.

Conclusions

Anaerobic sediments and permeable reactive barriers of granular iron metal are two examples of environmental systems where the characteristic pathway for contaminant transformation is reduction. Redox indicator dyes can be used to probe the physico-chemical properties of these reducing systems in research on the environmental fate of contaminants, but also to form educational exercises that illustrate some of the many facets of environmental chemistry. Key concepts that can be highlighted with the methods described in this manuscript include (i) thermodynamic versus kinetic control on reactivity, (ii) mass transport versus chemical control on reaction kinetics, (iii) initial rates and pseudo-first-order kinetics, (iv) the relationship between chemical structure and reactivity, (v) reversibility, (vi) stability analysis with E and pH as master variables, (vii) adsorption versus reaction at surfaces, (viii) chemical versus biologically mediated reactions in natural systems, (ix) spatial and temporal variability in environmental samples, and (x) the “chemical” nature of both pristine and contaminated environments. The methods can be adapted to provide lessons at levels that range from high school to graduate school, as reflected in the diversity of people that have contributed to this work.

Acknowledgments. Primary support for this work came from the Murdock Trust through the Research Corporation’s Partners in Science program (award numbers HS0446 and HS0520). The work also benefited from related projects funded by the Dreyfus Foundation (SG-98-101), National Science Foundation (CHE-9708554), and Office of Naval Research (N00014-96-1-0983). We would also like to thank Theron Morgan-Brown (Lewis & Clark College) for his help with the graphics in Figure 1.

References and Notes

1.       Stumm, W.; Morgan, J. J. Aquatic Chemistry, 3rd ed.; Wiley: New York, 1996; p. 1022.

2.       Tratnyek, P. G.; Macalady, D. L. Oxidation–Reduction Reactions in the Aquatic Environment. In Handbook of Property Estimation Methods for Chemicals: Environmental and Health Sciences; Mackay, D.; Boethling, R. S., Eds.; Lewis: Boca Raton, FL, 2000; pp 383–415.

3.       Bedard, D. L.; Quensen, J. F., III Microbial Reductive Dedechlorination of Polychlorinated Biphenyls. In Microbial Transformation and Degradation of Toxic Organic Chemicals; Young, L. Y.; Cerniglia, C. E., Eds.; Wiley-Liss: New York, 1995; pp 127–216.

4.       Sanders, J. E. PCB Pollution in the Upper Hudson River. In Contaminated Marine Sediments-Assessment and Remediation; Committee on Contaminated Marine Sediments, Ed.; National Academy Press: Washington, DC, 1989; pp 365–400.

5.       Ikalainen, A. J.; Allen, D. C. New Bedford Harbor Superfund Project. In Contaminated Marine Sediments-Assessment and Remediation; Committee on Contaminated Marine Sediments, Ed.; National Academy Press: Washington, DC, 1989; pp 312–350.

6.       Tratnyek, P. G. Putting Corrosion to Use: Remediation of Contaminated Groundwater with Zero-Valent Metals. Chem. Ind. (London) 1996, 13, 499–503.

7.       Roberts, A. L.; Totten, L. A.; Arnold, W. A.; Burris, D. R.; Campbell, T. J. Reductive Elimination of Chlorinated Ethylenes by Zero-Valent Metals. Environ. Sci. Technol. 1996, 30 (8), 2654–2659.

8.       Campbell, T. J.; Burris, D. R.; Roberts, A. L.; Wells, J. R. Trichloroethylene and Tetrachloroethylene Reduction in a Metallic Iron–Water–Vapor Batch System. Environ. Toxicol. Chem. 1997, 16 (4), 625–630.

9.       Totten, L. A.; Roberts, A. L. Investigating Electron-Transfer Pathways During Reductive Dehalogenation Reactions Promoted by Zero-Valent Metals. Abstracts of Papers, 209th National Meeting of the American Chemical Society, Anaheim, CA, April 2–6, 1995; American Chemical Society: Washington, DC, 1995; pp 706–709.

10.     Castro, C. E.; Oshea, S. K.; Bartnicki, E. W. Site Reactivity Probes: [1,2-¹³C]Chloroacetic Acid, a Reactivity Probe for Soil. Environ. Sci. Technol. 1995, 29 (8), 2154–2156.

11.     Smolen, J. M.; Weber, E. J.; Tratnyek, P. G. Molecular Probe Techniques for Reductant Identification in Reducing Sediments. Environ. Sci. Technol. 1999, 33 (3), 440–445.

12.     Kieber, D. J.; Blough, N. V. Determination of Carbon-Centered Radicals in Aqueous Solution by Liquid Chromatography with Fluorescence Detection. Anal. Chem. 1990, 62 (21), 2275–2283.

13.     Tratnyek, P. G.; Wolfe, N. L. Characterization of the Reducing Properties of Anaerobic Sediment Slurries Using Redox Indicators. Environ. Toxicol. Chem. 1990, 9 (3), 289–295.

14.     Iyere, P. A. Reduction of Viologen Biphosphate Dihalide with Aluminum Foil. J. Chem. Educ. 1996, 73 (5), 455–456.

15.     Hile, L. Letter on The Blue Bottle Revisited. J. Chem. Educ. 1998, 75 (9), 1067.

16.     Vandaveer, W. R., IV; Mosher, M. M. The Blue Bottle Revisited. J. Chem. Educ. 1997, 74 (4), 402.

17.     Vandaveer, W. R., IV; Mosher, M. M. Author's Reply: The Blue Bottle Revisited. J. Chem. Educ. 1998, 75 (9), 1067.

18.     Cook, A. G.; Tolliver, R. M.; Williams, J. E. The Blue Bottle Experiment Revisited: How Blue? How Sweet? J. Chem. Educ. 1994, 71 (2), 160–161.

19.     Johnson, T. L.; Scherer, M. M.; Tratnyek, P. G. Kinetics of Halogenated Organic Compound Degradation by Iron Metal. Environ. Sci. Technol. 1996, 30 (8), 2634–2640.

20.     Bishop, E. Indicators; Pergamon: Oxford, 1972, p 746.

21.     Weber, E. J.; Wolfe, N. L. Kinetic Studies of the Reduction of Aromatic Azo Compounds in Anaerobic Sediment/Water Systems. Environ. Toxicol. Chem. 1987, 6, 911–919.

22.     Zepp, R. G.; Wolfe, N. L. Abiotic Transformation of Organic Chemicals at the Particle–Water Interface. In Aquatic Surface Chemistry: Chemical Processes at the Particle–Water Interface; Stumm, W., Ed.; Wiley: New York, 1987; pp 423–455.

23.     Lemmon, T. L.; Westall, J. C.; Ingle, J. D., Jr. Development of Redox Sensors for Environmental Applications Based on Immobilized Redox Indicators. Anal. Chem. 1996, 68 (6), 947–953.

24.     Foster, E. M.; Nelson, F. E.; Speck, M. L.; Doetsch, R. N.; Olson, J. C., Jr. Dairy Microbiology; Prentice-Hall: Englewood Cliffs, NJ, 1957; p 492.

25.     Reddy, K. V. R.; Meherji, P. K.; Shahani, S. K. Resazurin Reduction Test to Evaluate Semen Quality. Indian J. Exp. Biol. 1997, 35 (4), 369–373.

26.     Liu, D.; Strachan, W. M. J. Characterization of Microbial Activity in Sediment by Resazurin Reduction. Ergebnisse der Limnologie, Suppl. to Archiv für Hydrobiologie 1979, 12, 24–31.

27.     Manual of Methods for General Bacteriology; Gerhardt, P., Ed.; American Society for Microbiology: Washington, DC, 1981; p 525.

28.     Vindimain, E.; Namour, P.; Munoz, J.-F.; Gril, J.-J.; Migeon, B.; Garric, J. Ethoxyresorufin-O-deethylase Induction in Fish from a Watershed Exposed to Nonpoint-Source Pollution of Agricultural Origin. Wat. Res. 1993, 27 (3), 449–455.

29.     Elangbam, C. S.; Quall, C. W., Jr.; Lochmiller, R. L. O-Dealkylation of Resorufin Ethers as an Indicator of Hepatic Cytochrome P-450 Isoenzyme Induction in the Cotton Rat (Sigmodon hispidus): A Method for Monitoring Environmental Contamination. Bull. Environ. Contam. Toxicol. 1991, 47, 23–28.

30.     Jones, D. A. Principles and Prevention of Corrosion; Macmillan: New York, 1992; p 568.

31.     Pankow, J. F. Aquatic Chemical Concepts; Lewis: Chelsea, MI, 1991, p 673.

32.     Pourbaix, M. Atlas of Electrochemical Equilibria in Aqueous Solutions; Pergamon: Oxford, 1996; p 644.

33.     Bailey, S. I.; Ritchie, I. M.; Hewgill, F. R. The Construction and Use of Potential–pH Diagrams in Organic Oxidation–Reduction Reactions. J. Chem. Soc. Perkin Trans. II 1983, 5, 645–652.

34.     Brookins, D. G. E_h-pH Diagrams for Geochemistry; Springer: Berlin, 1988, p 176.

35.     Borderie, B.; Lavabre, D.; Levy, G.; Micheau, J. C. A Simple Method for Analyzing First-Order Kinetics. J. Chem. Educ. 1990, 67 (6), 459–460.

36.     Hile, L. Reaction of Phenolphthalein. J. Chem. Educ. 1991, 68 (5), 446.

37.     Hostettler, J. D. On the Importance of Distinguishing E_h from pH. Am. J. Sci. 1985, 285 (9), 859–863.

38.     Hostettler, J. D. Electrode Electrons, Aqueous Electrons, and Redox Potentials in Natural Waters. Am. J. Sci. 1984, 284, 734–759.

39.     Stumm, W.; Morgan, J. J. On the Conceptual Significance of pH. Am. J. Sci. 1985, 285 (9), 856–859.

40.     Chylewski, C. Generalization of the Isosbestic Point. Angew. Chem. Inter. Ed. 1971, 10 (3), 195–196.

41.     Matheson, L. J.; Tratnyek, P. G. Reductive Dehalogenation of Chlorinated Methanes by Iron Metal. Environ. Sci. Technol. 1994, 28 (12), 2045–2053.

42.     Deng, B.; Campbell, T. J.; Burris, D. R. Kinetics of Vinyl Chloride Reduction by Metallic Iron in Zero-Headspace Systems. Abstracts of Papers, 213th National Meeting of the American Chemical Society, San Francisco, CA, April 13–17, 1997; American Chemical Society: Washington, DC, 1997; Vol. 37, No. 1, pp 81–83.

43.     Scherer, M. M.; Balko, B. A.; Gallagher, D. A.; Tratnyek, P. G. Correlation Analysis of Rate Constants for Dechlorination by Zero-Valent Iron. Environ. Sci. Technol. 1998, 32 (19), 3026–3033.

44.     Weber, E. J. Iron-Mediated Reductive Transformations: Investigation of Reaction Mechanism. Environ. Sci. Technol. 1996, 30 (2), 716–719.

45.     Jafvert, C. T.; Wolfe, N. L. Degradation of Selected Halogenated Ethanes in Anoxic Sediment–Water Systems. Environ. Toxicol. Chem. 1987, 6, 827–837.

46.     Clark, W. M. Oxidation-Reduction Potentials of Organic Systems; Williams & Wilkins: Baltimore, 1960; p 584.