Electron Pulse Radiolysis Determination of Hydroxyl Radical Rate

L/mg-m) and C/N ratio (90 mg/mg). Saguaro Lake DOM isolates (hydrophobic acid, hydrophobic neutral, transphilic acid) had intermediate values SUVA val...
0 downloads 0 Views 265KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 4640-4646

Electron Pulse Radiolysis Determination of Hydroxyl Radical Rate Constants with Suwannee River Fulvic Acid and Other Dissolved Organic Matter Isolates P A U L W E S T E R H O F F , * ,† STEPHEN P. MEZYK,‡ WILLIAM J. COOPER,§ AND DAISUKE MINAKATA† Department of Civil and Environmental Engineering, Arizona State University, Box 5306, Tempe, Arizona 85287-5306, Department of Chemistry and Biochemistry, California State University at Long Beach, Long Beach, California 90840, and Urban Water Research Center and Department of Civil and Environmental Engineering, University of California, Irvine, Irvine, California 92697-2175

Pulse radiolysis experiments were conducted on dissolved organic matter (DOM) samples isolated as hydrophobic and hydrophilic acids and neutrals from different sources (i.e., stream, lake, wastewater treatment plant). Absolute bimolecular reaction rate constants for the reaction of hydroxyl radicals (•OH) with DOM (k•OH, DOM) were determined. k•OH, DOM values are expressed as moles of carbon. Based on direct measurement of transient DOM radicals (DOM•) and competition kinetic techniques, both using pulse radiolysis, the k•OH, DOM value for a standard fulvic acid from the Suwannee River purchased from the International Humic Substances Society was (1.60 ( 0.24) × 108 M-1 s-1. Both pulse radiolysis methods yielded comparable k•OH, DOM values. The k•OH, DOM values for the seven DOM isolates from different sources ranged from 1.39 × 108 M-1 s-1 to 4.53 × 108 M-1 s-1, and averaged 2.23 × 108 M-1 s-1 (equivalent to 1.9 × 104 (mgC/L)-1 s-1). These values represent the first direct measurements of k•OH, DOM, and they compare well with literature values obtained via competition kinetic techniques during ozone or ultraviolet irradiation experiments. More polar, lower-molecular-weight DOM isolates from wastewater have higher k•OH, DOM values. In addition, the formation (microsecond time scale) and decay (millisecond time scale) of DOM• transients were observed for the first time. DOM• from hydrophobic acids exhibited broader absorbance spectra than transphilic acids, while wastewater DOM isolates had narrower DOM• spectra more skewed toward shorter wavelengths than did DOM• spectra for hydrophobic acids.

Introduction Dissolved organic matter (DOM) is ubiquitous in surface, ground, drinking, and waste waters, but its exact character * Corresponding author phone: 480-965-2885; fax: 480-965-0557; e-mail: [email protected]. † Arizona State University. ‡ California State University at Long Beach. § University of California, Irvine. 4640

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

varies depending on its origin (1-4). DOM affects many biogeochemical processes (e.g., metal complexation, redox conditions), photochemical processes (e.g., sunlight irradiation), oxidation of micropollutants (e.g., ozone, ultraviolet irradiation, sonication), and formation of halogenated disinfection byproducts during water treatment. DOM reacts with the hydroxyl radical (•OH) and thus can be a major scavenger in advanced oxidation processes during water treatment or natural biogeochemical processes (5-10). The only rate constants for the reaction of •OH with DOM are based upon indirect measurement techniques using ozone or UV irradiation (11-18). But by using well-established radiation chemistry methods, namely pulse radiolysis, it is possible to directly and accurately quantify these rate constants and formation of transient DOM radical species (DOM•), as has been done for numerous low-molecularweight organic chemicals (19). The goal of this study was to determine absolute rate constants for •OH reactions with DOM. Electron pulse radiolysis and transient adsorption spectroscopy experiments were performed on well-characterized DOM isolates purchased from the International Humic Substances Society (Suwannee River Fulvic Acid) as well as fractionated DOM (hydrophobic and hydrophilic acids and neutrals) from a surface water reservoir and a wastewater treatment plant. These isolates represent a broad spectrum of allochthonous and autochthonous DOM (1, 3, 20). The determined rate constants were related to the spectroscopic characteristics of DOM and DOM moieties, represented by model organic compounds. In addition to determining rate constants between •OH and DOM, we also report absorption spectra for transient DOM radicals (DOM•) for the first time.

Materials and Methods DOM Isolates. Seven DOM isolates from three different sources were used for pulse radiolysis experiments. A fulvic acid (a subfraction of hydrophobic acids) isolated from the Suwannee River was purchased from the International Humic Substances Society (cat. no. 1S101H). DOM isolates were fractionated from a surface water reservoir (Saguaro Lake, AZ) that serves as a drinking water supply, (20-22), or from the effluent of a wastewater treatment plant using aeration basins (Nogales, AZ) (23). Isolates were fractionated by resin chromatography using XAD8 (retains hydrophobic organics) and XAD4 (retains transphilic organics) resins and eluted with strong base (acid fractions) and acetonitrile solvents (neutral fractions) (see details in Supporting Information) (20, 24). The DOM isolates exhibited a range of specific UV absorbance at 254 nm (SUVA) and elemental atomic carbon to nitrogen ratios (C/N). Suwannee River fulvic acid (a subfraction of hydrophobic acids) has the highest SUVA (3.98 L/mg-m) and C/N ratio (90 mg/mg). Saguaro Lake DOM isolates (hydrophobic acid, hydrophobic neutral, transphilic acid) had intermediate values SUVA values (2.75, 1.35, 2.03 L/mg-m) and C/N ratios (41, 34, 17), respectively. The WWTP DOMisolates(hydrophobicneutral,transphilicacid,transphilic acid) had the lowest SUVA values (1.08, 1.22, 1.00 L/mg-m) and lower C/N ratios (29, 14, 10), respectively, compared against the other DOM isolates. Solutions containing DOM were prepared by adding lyophilized DOM to pH 7.0 buffered (10 mM phosphate) nanopure water. Pulse Radiolysis Techniques. The linear accelerator (LINAC) electron pulse radiolysis system at the Department of Energy Radiation Laboratory, University of Notre Dame, was used for all bimolecular rate constants and transient spectral measurements in this study (see Supporting Infor10.1021/es062529n CCC: $37.00

 2007 American Chemical Society Published on Web 05/22/2007

mation for details and ref 26). Irradiation of water produces •OH, hydrated electrons, and hydrogen atoms. To measure only the reactions of the •OH, all solutions were presaturated with nitrous oxide (N2O) to remove dissolved oxygen gas and to quantitatively convert the hydrated electrons and hydrogen atoms to •OH. Dosimetry for the electron pulse radiolysis was based on the transient absorbance produced in N2O-saturated 1.0 × 10-2 M KSCN solution at λ ) 475 nm (using G ) 5.2 × 10-4 m2 J-1) with doses of 3-5 Gy per 45 ns pulse (25). All experiments were conducted at pH 7.0 ( 0.1 in the presence of 10 mM phosphate buffer, and at room temperature (22 ( 1 °C). Two approaches were utilized to quantify bimolecular reaction rates in this study. All experiments were replicated at least three times. First, reaction rate constants were determined using SCN- competition kinetics; 100 µM KSCN in N2O-saturated water was used as a standard. As •OH is a strongly oxidizing species, it was expected to react with DOM mostly by abstracting a hydrogen atom from a C-H bond of one of the DOM’s constituent moieties. Addition to an aromatic ring may also occur. For either reaction, the overall competition is based on these two reactions: •

k1

OH + DOM 98 H2O + DOM•

k1 ) k•OH,DOM

(1)

k2

OH + SCN - (+SCN - ) 98 OH - + (SCN)2-•

•

k2 ) 1.05 × 1010 M-1 s-1 (2)

This competition can be analyzed to give the expression

[(SCN)2•-]o [(SCN)2•-]

)1+

k1[DOM] k2[SCN-]

(4)

where At is the absorbance at time t, Ao is the maximum absorbance, and B allows for baseline changes (such as when fitting bleaching curves, see below). For two DOM samples, the WWTP transphilic neutral and acid fractions, a small degree of absorption decay was also observed on the timescales of measurement, especially at higher DOM concentrations. This decay followed second-order kinetics, and is attributed to reactions of DOM• with other radicals present (R•)

DOM• + R• f products

k6

DOM 98 Products

(6)

where the observed transient absorbance data is for the intermediate DOM•. For the intermediate species, this consecutive reaction mechanism has an analytical solution of the form

[DOM•] )

[•OH]ok1 -k1t {e - e-k6t} + B k6 - k1

(7)

where the maximum absorbance of DOM• is related to the initial hydroxyl radical concentration [•OH]o. Analytical Methods. The concentration of DOM in all samples was quantified as dissolved organic carbon (DOC), mg C L-1, which was determined using a high-temperature combustion analysis on a Shimadzu TOC-V instrument. Rate constants for hydroxyl radical reactions with DOM (k•OH,DOM) are reported as the molar concentration of DOC, assuming 12 g C per mole C. This nomenclature for DOM rate constants is consistent with previously published values (11, 14, 16, 18, 27, 28). UV/vis absorbance spectra of non-irradiated samples were measured with a diode array spectrophotometer (Shimadzu UV1601). To characterize the structure of the DOM isolates, elemental analysis and 13C NMR analysis were conducted following methods described elsewhere (4, 20).

Results and Discussion (3)

where [(SCN)2•-]o is the absorbance of this transient at 475 nm when only SCN- is present. The absorbance at 475 nm decreases with added DOM, so a plot of [(SCN)2•-]o/[(SCN)2•-] against the concentration ratio [DOM]/[SCN-] gives a straight line of slope k1/k2. Using the rate constant for the reaction of •OH with SCN-, k2 ) 1.05 × 1010 M-1 s-1 (19), allows the k•OH, DOM (k1) rate constant to be readily determined (26). The second method used to quantify bimolecular reaction rates in this study involved directly monitoring the change in the growth and/or decay of the absorbance of the DOM transient species (DOM•) produced in N2O-saturated aqueous solutions. Acknowledging that DOM is not a unique compound, but rather a heterogeneous mixture of compounds, transient DOM• absorbance spectra were obtained, and the absorbance kinetics at selected wavelengths were determined for several DOM concentrations. As the initial hydroxyl radical concentration was much less than that of the DOM, pseudofirst-order kinetics were always obtained, and the absorbance growths were fitted by the standard expression

At ) Ao(1 - e-k1t) + B

To account for this slight decay, thus improving the fitting statistics for these two systems, the initial, small portion of this second-order decay was modeled by first-order kinetics. This gave the reaction sequence of eq 1 followed by

(5)

Absorbance Spectra of DOM. Figure 1A presents absorbance spectra for the DOM isolates. These indicate that Suwannee River fulvic acid has the highest molar absorbance and Nogales WWTP transphilic neutral has the lowest. Molar absorbance in the region of 250-280 nm is an indicator for the presence of sp2 hybridized carbon (i.e., carbon-carbon double bonds). It has previously been shown that increasing molar absorptivity in this region indicates DOM with higher reactivity toward certain oxidants (18, 29, 30). The Suwannee River and Saguaro Lake isolates yielded spectra with exponentially decreasing absorbance at longer wavelengths. The WWTP isolates also had decreasing absorbance with longer wavelengths, but exhibited a secondary absorbance peak around 280 nm. This may be due to the presence of surfactants in the wastewater (4). Transient absorbance spectra for DOM radicals (DOM•) produced upon reaction of •OH with five DOM isolates are presented in Figure 1B. The Suwannee River and Saguaro Lake isolates have broad transient absorptions, with peaks near 400 and 350 nm, respectively. In contrast, the WWTP isolates have sharper peaks with maximum values near 325 nm. This could imply that the DOM components that are most reactive with the hydroxyl radical are fundamentally different between the WWTP and two surface water isolates. The transient DOM• spectra gives additional insight into the major radical species produced in the oxidation of these isolates. Because the transient spectra differ, it is possible that the oxidation products differ among the DOM isolates. Suwannee River Fulvic Acid Reaction Kinetics with Hydroxyl Radicals. Rate constants for the •OH reaction with Suwannee River fulvic acid DOM were determined by both competition kinetics and direct monitoring of transient absorption kinetics to cross-validate these two methodologies. As expected, DOM addition decreased the maximum intensity of (SCN)2•- absorption (Figure 2). Although the direct reaction of the •OH with this DOM fraction produced an absorbance peak at 475 nm, the transient intensity of DOM• VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4641

FIGURE 2. Thiocyanate (100.0 µM) competition kinetics for determination of hydroxyl radical reaction rate constant with Suwannee River Fulvic Acid at pH 7.0 and room temperature. All curves are an average of 15 pulses, and contain zero (0), 0.52 (O), 1.07 (4), or 2.105 mM ()) DOM.

FIGURE 1. Molar absorbance spectra for DOM isolates (a). Transient absorption spectra (b) for hydroxyl radical reaction with DOM isolates in 0.10 M phosphate buffered, N2O-saturated, water at pH 7.0. All spectra taken ∼50 µs after electron pulse and normalized to dose of 5.0 Gy. (9) Suwannee River Fulvic Acid, (red filled circle) Saguaro Lake Hydrophilic Neutral, (green open square) WWTP Hydrophilic Neutral, (blue open circle) WWTP Transphilic Neutral, and (fuschia diamond) WWTP Transphilic Acid isolates. was very small compared to (SCN)2•-, and hence was ignored in data analysis. The maximum intensities of (SCN)2-• can be transformed to the form of eq 3, to give the linear plot shown in Figure 3. The slope of this line yields a rate constant (k•OH, DOM) of (1.55 ( 0.04) × 108 M-1 s-1 for the reaction of the hydroxyl radical with Suwannee River fulvic acid DOM. Direct measurement of k•OH, DOM for Suwannee River fulvic acid DOM was conducted at two wavelengths, 272 and 400 nm. The 400 nm wavelength corresponds to the maximum absorbance of the transient DOM• (Figure 1b). The 272 nm wavelength was selected to monitor a secondary transient DOM• for the kinetics across this broad transient absorbance. Significantly different kinetic behavior was observed for the two transient species (Figure 4). At 272 nm, an absorbance bleach effect occurred: at low DOM concentrations the absorbance decreased and became negative over time (Figure 4a). This indicated a net loss in the DOM material that originally absorbed at this wavelength. In contrast, at 400 nm a more absorbing transient species formed (Figure 4b). At both wavelengths DOM• was relatively long4642

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

FIGURE 3. Transformed thiocyanate competition kinetics plots for three different DOM isolates. Error limits as shown are one standard deviation. (0) Suwannee River Fulvic Acid, slope ) (1.371 ( 0.036) × 10-2, R 2 ) 0.998; (O) Saguaro Lake Hydrophobic Acid, slope ) (1.527 ( 0.035) × 10-2, R 2 ) 0.999; (4) Saguaro Lake Transphilic Acid, slope ) (1.276 ( 0.017) × 10-2, R 2 ) 1.000. lived, slowly decaying on the scale of milliseconds (not shown) which is slower than decay of many organic radicals produced from model organic compounds. Each set of transient kinetic data was fit by exponential decay or growth functions corresponding to the pseudo-first-order kinetics obtained. From plots of these fitted values against DOM concentrations, second-order reaction rate constants for •OH with Suwannee River fulvic acid were obtained (Figure SI.1 in the Supporting Information). The k•OH,DOM value based on the bleach kinetics at 272 nm was (1.87 ( 0.07) × 108 M-1 s-1, slightly faster than the rate constant of (1.39 ( 0.16) × 108 M-1 s-1) determined at 400 nm. Overall, k•OH, DOM values with Suwannee River fulvic acid DOM were remarkably similar for all experimental techniques. Based on competition kinetics and the direct measurement of the two transient absorption spectra, the average rate constant from these three measurement techniques is (1.60 ( 0.24) × 108 M-1 s-1. Other DOM Isolates. Because of the close agreement between the second-order rate constants determined by competition kinetics and direct analysis of transient species, only one of these techniques was used for each of the other DOM isolates. The technique chosen depended on the available mass of each isolate, as competition kinetic

FIGURE 4. (a) Transient absorbance kinetics observed at 272 nm for 121 µM (4), 342 µM (O), and 458 µM (0) of Suwannee River fulvic acid DOM. Solid lines are single-exponential decays, with pseudo-first-order values of 4.63 × 104, 8.85 × 104, and 1.14 × 105 s-1, respectively. (b) Equivalent kinetic behavior at 400 nm for these three concentrations of this DOM, with solid lines showing single growth exponential fits with first-order rate constants of 5.20 × 104, 8.18 × 104, and 1.04 × 105 s-1, respectively. experiments required more mass than transient DOM• experiments. Data analysis proceeded along nearly identical lines as that for Suwannee River fulvic acid. Table 1 summarizes the kinetic results of this study. Competition kinetics measurements for Saguaro Lake hydrophobic acid and transphilic acid isolates yielded k•OH, 8 -1 s-1 and (1.45 ( 0.02) DOM values of (1.73 ( 0.04) × 10 M × 108 M-1 s-1, respectively (e.g., Figure 3). These values are very similar to our measured rate constant for Suwannee River fulvic acid. The values are approximately half the values (average value of (3.6 ( 0.5) × 108 M-1 s-1) previously determined for seventeen hydrophobic and hydrophilic acids using the indirect method of a probe compound (parachlorobenzoic acid) and competition kinetic analysis during ozonation experiments (18). The ozonation experiments used different DOM isolates, and inherently may have greater uncertainty for obtaining k•OH, DOM values than pulse radiolysis because ozone itself reacts rapidly with some DOM moieties. Hydroxyl radical reaction rate constants for the other four DOM isolates were obtained using transient kinetic analysis (see values in Table 1). No bleaching occurred in these DOM isolates. Instead, growth of a new transient absorption was observed (Figure SI.2). For the WWTP transphilic acid and neutral fractions, the transient growth kinetics also included a small amount of decay on the timescales of measurement. Therefore, these kinetic data were fitted by eq 7 rather than eq 4, although the second-order reaction rate constant differed by less than 5% between the two calculations. Relationship Among DOM Isolates. The k•OH, DOM values for Saguaro Lake ((2.18 ( 0.13) × 108 M-1 s-1) and WWTP ((1.72 ( 0.13) × 108 M-1 s-1) hydrophobic neutral DOM isolates were only slightly higher than those of the other Saguaro Lake and Suwannee River DOM isolates (see Table 1). The WWTP transphilic acid and neutral fractions had the highest k•OH, DOM values, (3.63 ( 0.31) × 108 M-1 s-1 and (4.53 ( 0.53) × 108 M-1 s-1, respectively. Overall, our directly measured k•OH, DOM values averaged 2.23 × 108 M-1 s-1 (Table 1), which is similar to previously reported values ranging from 1.9 × 108 M-1 s-1 to 4.4 × 108 M-1 s-1 (18).

There was no relationship (R < 0.5) between the k•OH, DOM value and the DOM isolate’s specific ultraviolet absorbance (SUVA ) UVA254/DOC); SUVA is correlated with aromatic carbon content (1). Previous research had weakly implied this correlation (18), although those data were biased by one humic acid isolate with a very large SUVA. This lack of correlation with SUVA suggests that nonaromatic moieties present in the WWTP transphilic acid and neutral isolates may play an important role in controlling the range of hydroxyl radical reaction rate constants. The Saguaro Lake and wastewater hydrophobic acid and neutral fractions contained predominantly terpenoids, whereas the Suwannee River hydrophobic acid contained more phenolics. Furthermore, all the wastewater isolates contained some aromaticsulfonate moieties (4), as the 280 nm peaks may reflect. As such, focusing on UV absorbance at 254 nm alone may not be appropriate. Further attempts to relate k•OH, DOM values with SUVA or 13C NMR spectroscopy data did not reveal any statistically relevant trends. 13C NMR spectra are provided as Figures SI.3 and SI.4 in the Supporting Information. That the rate constants did not depend on the structural properties of DOM was initially surprising. Table 2 summarizes a range of model compounds that represent moieties that can potentially comprise DOM, and for which •OH rate constants are published (31-33). Other compounds were selected because they represent surfactants that could be in some samples (e.g., wastewaters). The •OH rate constants for the model compounds are from pulse radiolysis studies. These reported rate constants (M-1 s-1) were converted to k•OH, DOM values on a moles of carbon basis for comparison with DOM values from pulse radiolysis experiments. As compared to our k•OH, DOM values, hydroxyl radical rate constants are slightly larger for most benzene-based compounds and slightly lower for some carboxylic acids. However, some compounds containing thiols (e.g., cysteine) had the largest rate constants. Based on SUVA and 13C NMR data, the Suwannee River fulvic acid DOM should contain greater amounts of aromatic carbon than other isolates, but its k•OH, DOM rate constants were not the largest. Three hypotheses yet to be validated could explain these observations. First, molecular size and geometry of DOM could cause diffusion limitations for •OH, as the reactive moieties are “blocked” from reaction, which would imply that the upper limit for k•OH, DOM is near the observed highest calculated value of 5 × 108 M-1 s-1. Although weight- or number-averaged molecular weights are not available for the DOM isolates used in this study, transphilic acids generally have lower molecular weights than hydrophobic acids or neutrals (34, 35). The transphilic WWTP isolates had higher k•OH, DOM values than the presumably higher molecular weight hydrophobic neutral WWTP isolate. The Saguaro Lake DOM isolates, however, do not exhibit this trend. The two hydrophobic acid isolates from Saguaro Lake and Suwannee River had the lowest k•OH, DOM values but may have also had the highest molecular weights. Therefore, molecular size could be limiting the diffusive reaction of hydroxyl radicals in some cases, but this hypothesis requires additional research. One approach would be to conduct pulse radiolysis on size fractionated DOM isolates. The second hypothesis is that the k•OH, DOM value represents an averaged value for •OH reactions with an array of different moieties (variable reaction rates) and/or different reaction •OH mediated mechanisms (e.g., H-atom abstraction, •OH addition). Organic matter is comprised of a heterogeneous mixture of saturated and unsaturated carbon bonds, protonated and deprotonated carboxylic and hydroxyl acids, carbonyls, ketones, terpenoids, thiols, sulfonates, amides, and other moieties. Each of these moieties results in variable reaction rates with hydroxyl radicals. While the precise distribution of moieties may differ among DOM isolates, the VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4643

TABLE 1. Second-Order Rate Constants for •OH Reaction with DOM Isolates DOM isolate description

kinetic analysis method

λ monitored (nm)

k•OH,DOM (108 M-1 s-1)

Suwannee River Fulvic Acid Suwannee River Fulvic Acid Suwannee River Fulvic Acid Saguaro Lake Hydrophobic Acid Saguaro Lake Transphilic Acid Saguaro Lake Hydrophobic Neutral Nogales WWTP Hydrophobic Neutral Nogales WWTP Transphilic Neutral Nogales WWTP Transphilic Acid

direct transient growth direct transient growth SCN- competition SCN- competition SCN- competition direct transient growth direct transient growth direct growth and decay direct growth and decay

400 272 475 475 475 350 325 335 325

1.39 ( 0.16 1.87 ( 0.07 1.55 ( 0.04 1.73 ( 0.04 1.45 ( 0.02 2.18 ( 0.13 1.72 ( 0.13 4.53 ( 0.54 3.63 ( 0.31

TABLE 2. Summary of Hydroxyl Radical Reaction Rate Constants at near Neutral pH Levels (pH 7-9) for Model Compounds Representative of DOM and Surfactants

aRef

representative compound

k•OH (× 108 M-1 s-1)a

no. of carbon atoms

k•OH,DOM (108 Mcarbon s-1)b

salicylic acid citric acid tartaric acid catechol phthalic acid hydroquinone camphor EDTA at pH 4 (pH 9) oxalic acid nitrilotriacetate pH 4 (pH 9) benzaldehyde dodecylsulfate ion cysteine

120 3 14 110 59 52 41 4 (20) 1 8 (25) 44 80 190

7 6 4 6 8 6 10 10 5 6 7 12 3

17 1 4 18 7 9 4.1 0.4 (2) 0.2 1.3 6 7 63

19. b Rate constant expressed as moles carbon.

net distribution of faster and slower reacting moieties cam be represented by the following equation:

d[•OH] )-{ dt

∑ k [C ]}[ OH] ) - k •

i

i

T,c[CT,c][

•

OH] (8)

where ki and Ci are the second-order rate constant and concentration of a model organic compound “i”, respectively. kT,c and CT,c is the cumulative second-order rate constant and concentration (expressed as moles of carbon) for all the model compounds, respectively. Using equal concentrations of each compound in Table 2 results in 33% of the carbon being aromatic, a value which is higher than the aromatic carbon based upon 13C NMR of hydrophobic acids. The calculated kT,c value for this example using eq 8 is 8.7 × 108 M-1 s-1. Shifting the distributions to represent organic matter closer to DOM isolates considered here with aromatic carbon content of 11-17% results in lower predicted kT,c values of 3.4 × 108 M-1 s-1 and 4.9 × 108 M-1 s-1, respectively (see Table SI.1). These rate constant values are on the same relative order of magnitude as the observed k•OH, DOM values. If the aromatic content is maintained at 11% and the relative concentration of surfactants (EDTA, NTA) doubled, the resulting kT,c value is 3.6 × 108 M-1 s-1. This is only a 6% higher kT,c value calculated at half the surfactant concentration. Increasing the terpenoid fraction (represented by camphor) decreased slightly the value of kT,c for the same aromaticity, while adding even small relative amounts of thiol-containing moieties led to slightly higher kT,c. This example demonstrates that, within the percentage aromatic carbon range observed for the DOM isolates in this study, significant changes in the distribution of representative DOM compounds (i.e., Table 2) result in relatively minor differences in kT,c values, which is similar to the narrow range of k•OH, DOM values observed experimentally (i.e., Table 1). 4644

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

The third hypothesis is that upon radiolysis and oxidation by hydroxyl radicals, electrons are shuttled throughout the DOM molecules by quinone-like moieties. This could limit additional attack by hydroxyl radicals. The observation that DOM• transients are stable for milliseconds could suggest that this phenomena occurs. DOM molecules are known to contain quinones and upon oxidation by hydroxyl radicals can form semiquinones (36-38). Depending upon the availability of redox-active metals (e.g., iron), semiquinones could reform a quinone (36). While metal content is low in the DOM isolate solutions, further work (e.g., cobalt irradiation experiments) could be useful to assess the significance of quinone reformation. Future work should investigate these hypotheses. This could be accomplished by conducting hydroxyl radical rate constant experiments on a wide range of model biopolymers that more closely represent DOM moieties and biopolymers (see examples in refs 39 and 40) and span a range of molecular weights. Additionally, to fully incorporate DOM reactions into photolysis or advanced oxidation models, it will be necessary to obtain reaction rate constants for DOM with hydrated atoms (H•), aqueous electrons (eaq-), and superoxide (O2-). In systems with continuous HO• exposure (e.g., advanced oxidation processes) the role of organic matter may not be simply as a •OH scavenger (41, 42). The k•OH, DOM values represent “initial” reaction rates between DOM and •OH. Transient absorption spectra (Figure 1b) indicate different DOM• reaction products may form for different types of DOM isolates, as well as organic peroxides or other reactive byproducts (41-43). These byproducts may have lower rate constants with •OH than the initial DOM, and may vary depending on the source of the DOM (44). Future work is needed to elucidate the importance of these byproducts on subsequent, secondary, •OH consumption and reactions in

order to understand hydroxyl radical scavenging by DOM throughout continuous flow advanced oxidation processes.

Acknowledgments WateReuse Foundation provided funding to P.W. (WRF-05010) and W.J.C. (WRF 04-017). Jerry Leenheer (USGS retired) assisted in isolating DOM. Assistance is greatly appreciated for Department of Energy Radiation Laboratory at University of Notre Dame. The ASU Richard Snell Presidential fellowship, administered by Dr. John Crittenden, supported D.M. This is contribution 6 from the UCI Urban Water Research Center.

Supporting Information Available Methods for fractionating DOM isolates and conducting pulse radiolysis experiments; analysis for second-order rate constants at 272 nm and 400 nm with transient DOM• from Suwannee River fulvic acid; transient kinetic DOM• for four isolates at their peak transient absorbance and similar DOM concentrations; 13C NMR spectra for the DOM isolates as further means of differentiating the carbon bonds between the isolates; and tabulated data representing the model (eq 8). This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Aiken, G.; Cotsaris, E. Soil and hydrology - their effect on NOM. J. Am. Water Works Assoc. 1995, 87 (1), 36-45. (2) Thurman, E. M. Organic Geochemistry of Natural Waters; Martinus Nijhoff/Dr. Junk Publishers: Boston, MA, 1985. (3) Drewes, J. E.; Sprinzl, M.; Wendrock, A.; Williams, M. D.; Fox, P.; Westerhoff, P. Tracking residual DOC using XAD-fractionation and C13-NMR spectroscopy in indirect potable reuse. Vom Wasser 1999, 93, 95-107. (4) Chen, W.; Westerhoff, P.; Leenheer, J. A.; Booksh, K. Fluorescence excitation-emission matrix regional integration to quantify spectra for dissolved organic matter. Environ. Sci. Technol. 2003, 37 (24), 5701-5710. (5) Coale, K. H.; Chin, C. S.; Massoth, G. J.; Johnson, K. S.; Baker, E. T. In situ Chemical Mapping of Dissolved Iron and Manganese in Hydrothermal Plumes. Nature 1991, 352 (6333), 325-328. (6) Cooper, W. J.; Zika, R. G.; Petasne, R. G.; Fischer, A. M. Sunlight Induced Photochemistry of Humic Substances in Natural Waters: Major Reactive Species; American Chemical Society: Washington, DC, 1989; Vol. 219, pp 333-362. (7) Elovitz, M. S.; von Gunten, U. Hydroxyl radical ozone ratios during ozonation processes. I-The R-ct concept. Ozone: Sci. Eng. 1999, 21 (3), 239-260. (8) Elovitz, M. S.; von Gunten, U.; Kaiser, H. P. Hydroxyl radical/ ozone ratios during ozonation processes. II. The effect of temperature, pH, alkalinity, and DOM properties. Ozone: Sci. Eng. 2000, 22 (2), 123-150. (9) Yao, C. C. D.; Haag, W. R. Rate Constants For Direct Reactions of Ozone With Several Drinking-Water Contaminants. Water Res. 1991, 25 (7), 761-773. (10) Cooper, W. J.; Sawal, K. L.; Hoogland, Y. S.; Slifker, R. A.; Nickelsen, M. G.; Kurucz, C. N.; Waite, T. D. Disinfection byproduct precursor removal from natural waters using gamma radiation to simulate an innovative water treatment process. In Disinfection By-Products in Water Treatment; Amy, G. L., Minear, R. A., Eds.; Lewis Publishers: Boca Raton, FL, 1996; pp151-162. (11) Peyton, G. R.; Bell, O. In Practical modeling of contaminant destruction by HO radicals generated from ozone/peroxide/UV combinations; Proceedings of the 11th Ozone World Congress, San Fransisco, CA, 1993; p S-20-1-4. (12) Peyton, G. R.; Bell, O. J.; Girin, E.; LeFaivre, M. H.; Sanders, J. Effect of Bicarbonate Alkalinity on Performance of Advanced Oxidations Processes (Project #533); AWWA Research Foundation: Denver, CO, 1998. (13) Peyton, G. R.; Glaze, W. H. The Mechanism of Photolytic Ozonation. Abstr. Pap. Am. Chem. Soc. 1985, 189 (APR-), 5-GEOC. (14) Liao, C. H.; Gurol, M. D. Chemical oxidation by photolytic decomposition of hydrogen peroxide. Environ. Sci. Technol. 1995, 29, 3007-3014. (15) Hoigen, J.; Bader, H. Ozonation of water: oxidation-competition values of different types of waters used in Switzerland. Ozone Sci. Eng. 1979, 1, 357-372.

(16) Haag, W. R.; Yao, C. C. D. In Ozonation of US drinking water sources: HO concentrations and oxidation-competition values; Proceedings of the 11th Ozone World Congress, San Fransisco, CA, 1993; pp S-17-119-125. (17) Haag, W. R.; Yao, C. C. D. Rate Constants For Reaction of Hydroxyl Radicals With Several Drinking-Water Contaminants. Environ. Sci. Technol. 1992, 26, (5), 1005-1013. (18) Westerhoff, P.; Aiken, G.; Amy, G.; Debroux, J. Relationships between the structure of natural organic matter and its reactivity towards molecular ozone and hydroxyl radicals. Water Res. 1999, 33 (10), 2265-2276. (19) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. Critical review of rate constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals ( ×OH/×O-) in aqueous solution. J. Phys. Ref. Data 1988, 17 (2), 513-851. (20) Mash, H.; Westerhoff, P. K.; Baker, L. A.; Nieman, R. A.; Nguyen, M. L. Dissolved organic matter in Arizona reservoirs: assessment of carbonaceous sources. Org. Geochem. 2004, 35 (7), 831-843. (21) Nguyen, M. Sources and characteristics of dissolved organic carbon in arid-region water supplies. Ph.D. Dissertation, Arizona State University, Department of Civil and Environmental Engineering, Tempe, AZ, 2002. (22) Nguyen, M.; Baker, L.; Westerhoff, P. DOC and DBP precursors in western US watershed and reservoirs. J. Am. Water Works Assoc. 2002, 94 (5), 98-112. (23) Pinney, M. L.; Westerhoff, P. K.; Baker, L. Transformations in dissolved organic carbon through constructed wetlands. Water Res. 2000, 34 (6), 1897-1911. (24) Aiken, G.; McKnight, D.; Harnish, R.; Wershaw, R. Geochemistry of aquatic humic substances in the Lake Fryxell Basin, Antarctica. Biogeochemistry 1996, 34 (3), 157-188. (25) Mezyk, S. P.; Cooper, W. J.; Madden, K. P.; Bartels, D. M. Free radical destruction of N-nitrosodimethylamine in water. Environ. Sci. Technol. 2004, 38, 3161-3167. (26) Asmus, K. D. Pulse Radiolysis Methodology. Methods Enzymol. 1984, 105, 167-178. (27) Hoigne, J.; Bader, H. Ozonation of Water - Selectivity and Rate of Oxidation of Solutes. Ozone: Sci. Eng. 1979, 1 (1), 73-85. (28) Hoigne, J.; Bader, H. Ozonation of Water - Oxidation-Competition Values of Different Types of Waters Used in Switzerland. Ozone: Sci. Eng. 1979, 1 (4), 357-372. (29) Reckhow, D. A.; Singer, P. C. Chlorination by-products in drinking waters - from formation potentials to finished water concentrations. 1990, 82 (4), 173-180. (30) Westerhoff, P.; Chao, P.; Mash, H. Reactivity of natural organic matter with aquous chlorine and bromine. Water Res. 2004, 38, 1502-1513. (31) Radiation Chemistry Data Center of the Notre Dame Radiation Laboratory. Compilations of Chemical Property Data; http:// www.rcdc.nd.edu/index.html. (32) Bloom, P. R.; Brezonik, P. L.; Khwaja, A. R. Photochemical reactions and organic matter binding of mercuric ion and methylmercury in surface water; Water Research Center (University of Minnesota): St Paul, MN, 2001; p 7. (33) Leenheer, J. A.; Nanny, M. A.; McIntyre, C. Terpenoids as major precursors of dissolved organic matter in landfill leachates, surface water, and groundwater. Environ. Sci. Technol. 2003, 37 (11), 2323-2331. (34) Brown, A.; McKnight, D. M.; Chin, Y. P.; Roberts, E. C.; Uhle, M. Chemical characterization of dissolved organic material in Pony Lake, a saline coastal pond in Antarctica. Mar. Chem. 2004, 89 (1-4), 327-337. (35) Jarusutthirak, C.; Amy, G.; Croue, J. P. Fouling characteristics of wastewater effluent organic matter (EfOM) isolates on NF and UF membranes. Desalination 2002, 145 (1-3), 247-255. (36) Chen, R. Z.; Pignatello, J. Role of quinone intermediates as electron shuttles in Fenton and photoassisted Fenton oxidations of aromatic compounds. Environ. Sci. Technol. 1997, 31 (8), 2399-2406. (37) Cory, R. M.; McKnight, D. Fluorescence spectroscopy reveals ubiquitous presence of oxidized and reduced quinones in dissolved organic matter. Environ. Sci. Technol. 2005, 39 (21), 8142-8149. (38) Klapper, L.; McKnight, D.; Fulton, J. R.; Blunt-Harris, E. L.; Nevin, K. P.; Lovley, D. R.; Hatcher, P. G. Fulvic acid oxidation state detection using fluorescence spectroscopy. Environ. Sci. Technol. 2002, 36 (14), 3170-3175. (39) Leenheer, J. In Comprehensive Assessment of Precursors, Diagenesis, and Reactivity to Water Treatment of Dissolved and Colloidal Organic Matter; IHSS: Australia, 2005. VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4645

(40) Aluwihare, L. I.; Repeta, D. J.; Chen, R. F. A major biopolymeric component to dissolved organic carbon in surface sea water. Nature 1997, 387 (8), 167-169. (41) von Gunten, U. Ozonation of drinking water: Part I. Oxidation kinetics and product formation. Water Res. 2003, 37 (7), 14431467. (42) von Gunten, U. Ozonation of drinking water: Part II. Disinfection and by-product formation in presence of bromide, iodide or chlorine. Water Res. 2003, 37 (7), 1469-1487. (43) Buffle, M. O.; Schumacher, J.; Meylan, S.; Jekel, M.; von Gunten, U. Ozonation and advanced oxidation of wastewater: Effect of O-3 dose, pH, DOM and HO•-scavengers on ozone decomposi-

4646

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

tion and HO• generation. Ozone: Sci. Eng. 2006, 28 (4), 247259. (44) Goldstone, J. V.; Pullin, M. J.; Bertilsson, S.; Voelker, B. M. Reactions of hydroxyl radicals with humic substances: bleaching, mineralization, and production of bioavailable carbon substrates. Environ. Sci. Technol. 2002, 36 (3), 364-372.

Received for review October 20, 2006. Revised manuscript received April 1, 2007. Accepted April 13, 2007. ES062529N