Environ. Sci. Technol. 2007, 41, 2776-2781
Examination of NOM Chlorination Reactions by Conventional and Stop-Flow Differential Absorbance Spectroscopy G R E G O R Y V . K O R S H I N , * ,† MARK M. BENJAMIN,† HYUN-SHIK CHANG,† AND H E R V EÄ G A L L A R D † Department of Civil and Environmental Engineering, University of Washington, Box 352700 Seattle, Washington 98195-2700, and Laboratoire de Chimie de l’Eau et de l’Environnement, Universite´ de Poitiers, 40 Avenue du Recteur Pineau, Poitiers, France 86022
Mechanisms of chlorination of natural organic matter (NOM) in surface water (Lake Washington) were explored via differential spectroscopy. Two types of differential spectra (overall and incremental) were generated for this water chlorinated at pH 7 using varying chlorine doses and reaction times. The differential spectra contain two kinetically and spectroscopically distinct components. One of these components is attributable to functional groups that react rapidly with chlorine, while the other reflects transformations of slowly reacting chromophores that arise following the depletion of the fast chromophores. Small concentrations of disinfection byproducts (DBPs), exemplified in this study by dichloroacetic acid and chloral hydrate, were produced during the initial phase of chlorination, when the fast-reacting chromophores were being consumed. Rather, the release of those DBPs was correlated with the breakdown of the slowly reacting chromophores.
Introduction Since the recognition of chloroform as a ubiquitous disinfection byproduct (DBP) in the 1970s, numerous chlorineand bromine-containing DBP species have been found in drinking water (1-3). Trihalomethanes (THMs) and haloacetic acids (HAAs) normally account for ca. 50% of the total organic halogen (TOX) in drinking water, with the rest of the TOX being mainly incorporated into larger molecules of natural organic matter (NOM) (1, 4, 5). Chlorination sites in NOM are thought to be predominantly polyhydroxyaromatic (PHA) moieties (6-8) and, secondarily, esters and ketones (9, 10). Despite progress in exploring the reactivity of the PHA groups (8, 11, 12), mechanisms of their reactions with halogen species have not been ascertained. Even for model precursor compounds such as resorcinol and dihydroxybenzoic acid, the incorporation of chlorine involves multistep, branching reactions (13-16); the reactions of NOM and chlorine are undoubtedly more complex. These difficulties notwithstanding, numerous mechanistic and phenomenological models of DBP formation have been * Corrresponding author phone: (206) 543-2394; fax: (206) 6859185; e-mail:
[email protected]. † University of Washington. ‡ Universite ´ de Poitiers. 2776
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007
proposed. The mechanistic models assume that the chlorination of NOM is a multistep reaction that involves chlorine incorporation into the predominantly aromatic reactive centers of NOM, followed by breakdown of these groups to release individual DBPs (10, 17-20). Such models can be employed to develop a kinetic description of NOM halogenation, but the utility of this approach is limited by uncertainties regarding the chemical and kinetic properties of the reactive NOM functional groups. On the other hand, phenomenological models (e.g., based on regressions relating DBP formation to DOC, chlorine dose, pH, etc.) can achieve a tolerable goodness-of-fit, but only by inclusion of several, site-specific empirical fitting parameters (10, 19, 21, 22). The complexity of NOM properties and reactions has necessitated the use of surrogate parameters such as absorbance of light at 254 nm (A254) and the specific absorbance at that wavelength (SUVA254) to estimate its reactivity toward chlorination (8, 23-26). We previously proposed an alternative approach, in which the intensity of the differential absorbance at 272 nm (∆A272) is measured during the course of chlorination and correlated with the individual DBPs and/ or TOX concentrations generated in the solution (27-29). The value of ∆Aλ is computed as the change in absorbance induced by chlorination, measured at any time after chlorine has been dosed (eq 1)
∆Aλ(t) ) Aλ(t) - Aλ(0)
(1)
where Aλ(t) and Aλ(0) are the absorbances of NOM at wavelength λ after reaction time t and prior to chlorination, respectively. Because NOM chlorination leads to a decline in UV absorbance, ∆Aλ is negative. The differential spectra of chlorinated NOM from all surface waters investigated to date (27-29) exhibit a broad band with a maximum at a wavelength between 265 and 275 nm, and a monotonic increase in the magnitude of ∆Aλ with increasing chlorine dose or reaction time. The differential spectra undergo subtle changes in shape as a function of chlorine dose and reaction time (30), but these changes have not been systematically studied. Despite the strength of the correlations between ∆Aλ and the concentrations of individual DBPs, only a few attempts have been made to determine how the presence of kinetically distinct functionalities in NOM is manifested in the differential spectra, and how the corresponding features of the differential spectra relate to DBP release (30). In prior publications dealing with differential absorbance (27-30), we focused on the change in the absorbance spectrum from the initiation of chlorination. That spectrum is referred to here as the overall differential absorbance spectrum. In this work, we also consider incremental differential spectra; i.e., the value of ∆Aλ(t1,t2) in eq 2
∆Aλ(t1,t2) ) Aλ(t2) - Aλ(t1)
(2)
The overall differential absorbance spectrum is thus the incremental differential spectrum evaluated at t1 ) 0, i.e., it is ∆Aλ(0,t2). In this paper, we analyze both overall and incremental differential absorbance spectra of chlorinated NOM to gain insights into the mechanism of NOM chlorination and DBP release.
Experimental Section Water from Lake Washington (LW) in Seattle, WA, was used in all experiments. The DOC of the water was 3.0 mg/L. Chlorination was carried out at 20 °C. The pH was buffered at 7.0 with 0.03 M phosphate. All experimental procedures 10.1021/es062268h CCC: $37.00
2007 American Chemical Society Published on Web 03/17/2007
FIGURE 1. Differential absorbance spectra of chlorinated LW water. Cl2:DOC ) 1.5, pH 7.0, 20 °C. followed established practices that have been described in previous publications (e.g., 28). Chlorine was added as HClO and is reported here as the equivalent amount of Cl2. The chlorine dose is presented in terms of the weight ratio of Cl2 added to the concentration of dissolved organic carbon (Cl2: DOC). Experiments to examine short-term chlorination of NOM were carried out at pH 7.0 and 20 °C, utilizing an Applied Photophysics X.18MV stop-flow reactor. In this instrument, drive rams inject preset amounts of the reagents into the reaction chamber, where they are rapidly mixed. The reaction chamber also serves as an optical cell with a 1-cm path length, through which absorbance measurements can be taken at a very high frequency. The nominal dead time of the instrument is 1.1 ms, and the mixture aging time ranged from 0.015 to 100 s. In these experiments, the absorbance of the sample over time was recorded at one wavelength per experiment. The data for a given reaction time from all the experiments (i.e., for all wavelengths) were then combined to generate the absorbance spectrum for that time. The DOC concentration in the baseline water in the stopflow experiments had to be increased to improve the spectrophotometric precision and sensitivity of the tests. The preconcentration was accomplished using a Bu ¨ chi Labortechnik R-2 rotary evaporator equipped with a 2.5-L roundbottom flask. Following rotary evaporation, the sample was adjusted to pH 8-9, filtered through a 0.45-µm filter to remove calcium carbonate precipitate that formed during the preconcentration, and then acidified to pH 7. The final product water, with a DOC concentration ranging from 40 to 100 mg/L, was stored in a PTFE container at 4 °C. For longer term experiments, absorbance spectra were measured with a dual-beam Perkin-Elmer Lambda-18 spectrophotometer using 5-cm quartz cells. Deconvolution of the spectra was performed with Toolbox and MatLab software packages as described in the Supporting Information (SI).
Results and Discussion Overall differential spectra for LW water chlorinated at a Cl2:DOC ratio of 1.5 and varying reaction times are shown in Figure 1. All the spectra have a well-defined band with a maximum near 272 nm; such bands have appeared in all the differential spectra obtained in prior studies of NOM chlorination using several other natural water sources (2729). The intensity of the differential absorbance at λ > 250 nm increased with increasing reaction time and Cl2:DOC ratio, as shown in Figure 2. Although the general shape of the overall differential spectra is similar for all chlorination conditions, they undergo consistent (albeit subtle) changes in response to changes in reaction time and/or initial chlorine dose. For example, as is clear when each of the spectra in Figure 1 is normalized by dividing it by -∆A272 (Figure 3), the characteristic band
FIGURE 2. Changes of the intensity of differential absorbance at 272 nm at varying reaction times and Cl2:DOC ratios. Chlorinated LW water, pH 7.0, 20 °C. DOC concentration 3.0 mg/L, range of initial chlorine concentrations 0.3-4.5 mg/L.
FIGURE 3. Differential absorbance spectra of chlorinated LW water, normalized to the maximum value of -∆A in each spectrum. Cl2: DOC ) 1.5, pH 7.0, 20 °C. steadily broadens as the reaction progresses. This result suggests that the overall differential spectra comprise two or more distinct components, an inference that can be tested by exploring the incremental differential spectra. Incremental differential spectra quantify the changes in NOM chromophores between any two points in the reaction. If only one type of chromophore contributed to the spectroscopic response shown in Figure 1 and Figure 3, the shape of the incremental differential spectra would be the same for any t1 and t2, although the intensity of the spectra would depend on t1, t2, and the overall kinetics of the reaction. On the other hand, if chromophores with dissimilar spectroscopic and kinetic properties were involved, the different spectra of the more rapidly and more slowly reacting chromophores would cause the incremental spectra at shorter reaction times to differ from those at longer times. As a result, examination of the incremental differential spectra allows the emergence and/or decay of dissimilar chromophores to be discerned. A sequence of incremental differential spectra for the experimental system is shown in Figure 4, and the same spectra normalized to -∆A272 are shown in Figure S1 in the SI. The initial phase of reaction (0-5 min) is characterized by a well-defined band with a maximum at 272 nm and low absorbance at λ > 330 nm, but the intensity of that band is greatly diminished over subsequent time intervals. Figure 4 and Figure S1 establish that the differential spectra of chlorinated NOM contain at least two componentssone (which we refer to as Component A) associated primarily VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2777
FIGURE 4. Incremental differential absorbance spectra of chlorinated LW water. Cl2:DOC ratio 1.5, pH 7.0, 20 °C. FIGURE 6. Comparison of the shape of differential absorbance spectra of LW water generated in stop-flow and conventional experiments. pH 7.0.
FIGURE 5. Differential absorbance spectra of chlorinated, preconcentrated LW water (45 mg/L DOC) after various reaction times, from stop-flow experiments. Cl2:DOC ratio 0.22, pH 7.0. with the initial steps of chlorination and containing the welldefined band centered near 272 nm, and another (Component B) that is associated with chromophores that react more gradually and is characterized by a gradual decrease in absorbance at λ >272 nm. Thus, components A and B exhibit different spectroscopic characteristics and also appear at different phases of NOM chlorination. The significance of these findings for understanding NOM chlorination and formation of individual DBP species is addressed in the sections that follow. The characteristics of the rapidly reacting chromophores were explored in greater detail in stop-flow (SF) experiments in which the spectra of preconcentrated LW water were collected 0.2 to 100 s after chlorination. The resulting differential spectra are shown in Figure 5 and are compared with the differential spectra of unaltered LW water obtained with the conventional experimental system at the shortest reaction times and the lowest chlorine doses investigated (t ) 5 min and Cl2:DOC ) 0.1 and 0.2, respectively) in Figure 6. Although the absolute intensities of the spectra for the preconcentrated, stop-flow samples are an order of magnitude greater than those of the unaltered, conventional samples, the features of all the spectra are strikingly similar. Based on the similarity of the shapes of the spectra obtained in the SF and conventional modes during the initial phase of NOM chlorination, the normalized spectrum of Component A was assumed to be identical to that of the differential spectra at the shortest reaction times for chlorine: DOC ratios 0.1 and 0.2. The normalized spectrum of Component B and absolute values of the contributions of Components A and B were determined via numerical deconvolution of the overall differential spectra carried for all experimental conditions. The normalized spectra of 2778
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007
FIGURE 7. Normalized spectra of Components A and B of the differential spectra, as inferred by deconvolution of the overall differential spectra from conventional chlorination experiments with reaction times from 5 min to 7 days and Cl2:DOC ratios from 0.1 to 1.5.
FIGURE 8. Contributions of Components A and B to the overall differential spectra as a function of the progress of the reaction (as indicated by ∆A272). Components A and B (∆Aλ,A/∆Aλ,max,A and ∆Aλ,B/∆Aλ,max,B, respectively) are shown in Figure 7. The spectrum of Component B is considerably broader than that of Component A and has a peak at a slightly higher wavelength (281 nm). The relative contributions of these two components to the overall spectrum as a function of the progress of the overall reaction (as indicated by -∆A272) are shown in Figure 8, and the match between the observed and calculated spectra for a typical set of conditions is shown in
FIGURE 9. Change in ∆R as the chlorination reaction progresses (as indicated by the change in ∆A272). Chlorinated LW water. pH 7, 3 mg/L DOC, Cl2:DOC ratios from 0.1 to 1.5, reaction times from 5 to 10080 min. Figure S2 in the SI (the “calculated” spectrum is based on the deconvolution analysis). Figure 8 indicates that Component A accounts for all of the differential absorbance during the initial stages of the reaction (corresponding to -∆A272 < 0.008 cm-1), but Component B makes increasingly significant contributions as the reaction proceeds. Thus, it appears that only the fast chromophores are present in the initial NOM, and that the slow chromophores emerge only after a substantial fraction of the fast chromophores have been modified or consumed by reactions with chlorine. Because numerical deconvolution of absorbance data is not always convenient for practical applications, it is useful to introduce alternative parameters that reflect the engagement of the fast and slow steps in the chlorination reaction. We propose here that the contribution of Component B to the differential spectrum can be estimated based on the decrease in the ratio (R) of absorbances at 350 and 272 nm from the beginning of the reaction, i.e.:
∆R )
( ) ( ) A350 A272
-
0
A350 A272
(3)
t
The wavelength of 272 nm in the above expression corresponds to the maximum in the overall differential spectra. The choice of λ ) 350 nm is based on the fact that the highest relative change of the normalized absorbances of Components A and B shown in Figure 7 occurred at this wavelength. ∆R is expected to be negative at low chlorine doses and reaction times, because the decrease of absorbance at 272 nm during the initial phase of the reaction is greater than that at 350 nm. However, if the Cl2:DOC ratio and/or reaction time are large enough for the reaction to proceed further, loss of slowly reacting chromophores begins to exceed that of fast-reacting chromophores, so ∆R increases and can become positive. Thus, the behavior of ∆R is expected to reflect the emergence and predominance of the slow chromophores. As shown in Figure 9, computed values for ∆R from the experiments at all Cl2:DOC ratios collapsed to a single curve that followed these expectations. Specifically, ∆R was negative or close to zero at early stages of the reaction, passed through a minimum near -∆A272 ) 0.008 cm-1 (i.e., at the threshold for emergence of Component B), and then were positive and increased nearly linearly with -∆A272 at later stages. Moreover, ∆R was strongly and linearly correlated with -∆Aλ,B (Figure S3 in the SI), reinforcing that this easily measured parameter (∆R) can be employed to track the engagement of the slow chromophores.
FIGURE 10. Relationship between chlorine consumption and intensity of differential absorbance at 272 nm. LW water, pH 7, Cl2:DOC ratios from 0.1 to 1.5. Overall, the results presented above suggest that the full ∆Aλ spectrum (or, as a first approximation, simply -∆A272) is a quantitative indicator of the changes induced by chlorination in the entire system of NOM chromophores, while -∆Aλ,A and -∆Aλ,B (or, more simply, ∆R) indicate the relative involvement of kinetically distinct types of chromophores. Within this hypothesis, negative and decreasing ∆R values indicate that the rapidly reacting chromophores are affecting the absorbance spectrum more than the slowly reacting chromophores, and increasing, positive ∆R indicates the opposite. The consumption of chlorine in these systems is associated with oxidation and halogenation of NOM. The loss of absorbance is presumably a manifestation of these same reactions, reflecting changes in the chromophores that react with chlorine. Therefore, a correlation between chlorine consumption and -∆Aλ is expected. This expectation is confirmed by the data in Figure 10, where chlorine consumption is plotted against -∆A272. Similarly to ∆R (Figure 9), when plotted against ∆A272, the chlorine consumption data form a coherent dataset for all the Cl2:DOC ratios and reaction times investigated. Thus, even though chlorine does not react solely with NOM chromophores (it can also react with other functional groups that do not absorb light at λ > 250 nm, such as ketones or esters), the decrease of the UV absorbance of NOM upon chlorination can be used as a measure of the consumption of chlorine caused by reactions with NOM. The relationship in Figure 10 is approximately linear at -∆A272 > 0.011 cm-1, but is curvilinear and has a lower slope at lower values of -∆A272. The lower slope at values of -∆A272 that correspond to the predominance of Component A indicates that the initial reactions of the NOM with chlorine have a greater effect on A272 than the later reactions do, for a given amount of chlorine reacting. The region of Figure 10 with the lower slope corresponds closely to that of negative ∆R in Figure 9, reinforcing the idea that ∆R is a useful indicator of the relative involvement of different reactive groups in NOM with chlorine. As noted in the introduction, the major concern associated with chlorination of NOM is the formation of DBPs. Plots showing the concentrations of two representative DBP species (DCAA and chloral hydrate) as a function of -∆A272 are shown in Figures 11 and 12. These figures include all the data for the given DBP from experiments with Cl2:DOC ratios from 0.1 to 1.5. ∆R vs -∆A272 data are also included in the plots. The value of -∆A272 where Component B emerges is specifically indicated in these figures. Strong correlations exist between the concentrations of both of the DBPs and -∆A272. However, DCAA was generated VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2779
chloral hydrate, and other DBPs during chlorination of drinking water.
Acknowledgments This study was supported by American Water Works Association Research Foundation (Project 2597). The views represented in this publication do not necessarily represent those of the funding agency.
Supporting Information Available
FIGURE 11. Relationships between the intensity of differential absorbance at 272 nm, ∆R, and formation of DCAA. Chlorinated LW water, pH 7.0, Cl2:DOC ratios 0.1 to 1.5.
Brief description of numerical deconvolution and three figures showing the evolution of normalized incremental differential spectra, an example of the deconvolution of an overall differential spectrum into components corresponding to the fast and slow chromophores, and the correlation between ∆R and the differential absorbance associated with the slow chromophores. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited
FIGURE 12. Relationships between the intensity of differential absorbance at 272 nm, ∆R, and formation of chloral hydrate. Chlorinated LW water, pH 7.0, Cl2:DOC ratios 0.1 to 1.5.
approximately in proportion to -∆A272 (Figure 11), and it reached concentrations up to 10 µg/L under conditions where -∆Aλ,B ≈ 0 (-∆A272 < 0.008 cm-1). On the other hand, the chloral hydrate concentration follows ∆R (and, therefore, -∆Aλ,B) more closely than it follows -∆A272. Specifically, no chloral hydrate formed for -∆A272 values below the threshold of 0.008 cm-1, but above it the concentration of chloral hydrate increased rapidly. These results suggest that DCAA formation might be tied to the reactions of both the rapidly and slowly reacting chromophores in NOM, whereas chloral hydrate formation appears to be associated solely with the consumption of the slowly reacting chromophores. More broadly, the combination of these results with the observation that Component B is initially absent from the differential spectra and emerges as Component A disappears suggests that Component B is a spectroscopic signature of chlorinated intermediates formed during the conversion of the initial reactive sites in NOM to individual DBPs. Since DCAA is a dichlorinated species whereas chloral hydrate is trichlorinated, and since chloral hydrate formation is more closely correlated with the intensity of Component B, it may be that Component B is represents highly chlorinated intermediates from which chloral hydrate and other trichlorinated DBPs are cleaved. This hypothesis will be explored in a subsequent publication, in which the approach developed here for interpreting the differential spectra is used to examine the behavior of NOM in other surface waters and also as the basis for a mathematical model of the kinetics of the formation of DCAA, 2780
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007
(1) Krasner, S. W.; McQuire, M. J.; Jacangelo, J. C.; Patania, N. L.; Reagan, K. M.; Aieta, E. M. The occurrence of disinfection byproducts in US drinking water. J. Am. Water Works Assoc. 1989, 81 (8), 41-53. (2) Richardson, S. D.; Simmons, J. E.; Rice, G. Disinfection byproducts: The next generation. Environ. Sci. Technol. 2002, 36 (9), 198A-205A. (3) Roberts, M. G.; Singer, P. C.; Obolensky, A. Comparing total HAA and total THM concentrations using ICR Data. J. Am. Water Works Assoc. 2002, 94 (1), 103-116. (4) Singer, P. C.; Chang, S. D. Correlations between trihalomethanes and total organic halides during water treatment. J. Am. Water Works Assoc. 1989, 81 (8), 61-65. (5) Frimmel, F. H.; Hesse, S.; Kleiser, G. Technology-related characterization of hydrophilic disinfection by-products in aqueous samples. In Natural Organic Matter and Disinfection By-Products. Characterization and Control in Drinking Water; Barrett, S. E., Krasner, S. W., Amy, G. L., Eds.; American Chemical Society: Washington, DC, 2000. (6) Larson, R. A.; Weber, E. J. Reaction Mechanisms in Environmental Organic Chemistry; Lewis Publishers: Boca Raton, FL, 1994. (7) Martin-Mousset, B.; Croue´, J. P.; Lefebvre, E.; Legube, B. Distribution and characterization of dissolved organic matter of surface waters. Water Res. 1997, 31 (3), 541-553. (8) Croue´, J. P.; Korshin, G. V.; Benjamin, M. M. Characterization of Natural Organic Matter in Drinking Water; AWWARF: Denver, CO, 2000. (9) Leenheer, J. A.; Wilson, M. A.; Malcolm, R. L. Presence and potential significance of aromatic ketone groups in aquatic humic substances. Org. Geochem. 1987, 11 (4), 273-280. (10) McClellan, J. N.; Reckhow, D. A.; Tobiason, J. E.; Edzwald, J. K.; Smith, D. B. A comprehensive kinetic model for chlorine decay and chlorination by-products formation. In Natural Organic Matter and Disinfection By-Products. Characterization and Control in Drinking Water; Barrett, S. E., Krasner, S. W., Amy, G. L., Eds.; American Chemical Society: Washington, DC, 2000; Ch. 15. (11) Chin, Y. P.; Aiken, G.; O’Loughlin, E. Molecular weight, polydispersity and spectroscopic properties of aquatic humic substances. Environ. Sci. Technol. 1994, 28 (11), 1853-1858. (12) Leenheer, J. A.; Croue´, J. P. Characterizing aquatic dissolved organic matter. Environ. Sci. Technol. 2003, 37 (1), 18A-26A. (13) Rook, J. J. Chlorination reactions of fulvic acids in natural waters. Environ. Sci. Technol. 1977, 11 (5), 478-482. (14) Boyce, S. D.; Hornig, J. F. Reaction pathways of trihalomethane formation from the halogenation of dihydroxyaromatic model compounds for humic acid. Environ. Sci. Technol. 1983, 17 (4), 202-211. (15) Gonzalez, A. C.; Olson, T. M.; Rebenne, L. M. Aqueous chlorination kinetics and mechanisms of substituted dihydrobenzenes. In Water Disinfection and Natural Organic Matter; Minear, R. A., Amy, G., Eds.; ACS Symposium Series 649; American Chemical Society: Washington, DC, 1996, Ch. 4, pp 48-62. (16) Rios, R. V. R. A.; da Rocha, L. L.; Vieira, T. G.; Lago, R. M.; Augusti, R. On-line monitoring by membrane introduction mass spectrometry of chlorination of organics in water. Mechanistic and
(17) (18)
(19) (20) (21) (22) (23) (24) (25)
kinetic aspects of chloroform formation. J. Mass Spectrom. 2000, 35 (5), 618-624. Adin, A.; Katzhendler, J.; Alkaslassy, D.; Rav-Acha, C. THM formation in chlorinated drinking water: A kinetic model. Water Res. 1991, 25 (9), 797-805. Cowman, G. A.; Singer, P. C. Effect of bromide ion on haloacetic acid speciation resulting from chlorination and chloramination of aquatic humic substances. Environ. Sci. Technol. 1996, 30 (1), 16-24. Gallard, H.; Von Gunten, U. Chlorination of natural organic matter: Kinetics of chlorination and of THM formation. Water Res. 2002, 36 (1), 65-74. Nokes, C. J.; Fenton, E.; Randall, C. J. Modeling the formation of brominated trihalomethanes in chlorinated drinking waters. Water Res. 1999, 33 (17), 3557-3568. Gang, D. C.; Clevenger, T. E.; Banerji, S. K. Relationship of chlorine decay and THMs formation to NOM size. J. Hazard. Mater. 2003, 96 (1), 1-12. Clark, R. M.; Thurnau, R. C.; Sivaganesan, M.; Ringhand, P. Predicting the formation of chlorinated and brominated byproducts. J. Environ. Eng. - ASCE 2001, 127 (6), 493-501. Edzwald, J. K.; Becker, W. C.; Wattier, K. L. Surrogate parameters for monitoring organic matter and THM precursors. J. Am. Water Works Assoc. 1985, 77 (4), 122-132. Wang, Z. D.; Pant, B. C.; Langford, C. H. Spectroscopic and structural characterization of a laurentian fulvic acid: Notes on the origin of color. Anal. Chim. Acta 1990, 232 (1), 43-49. Kitis, M.; Karanfil, T.; Kilduff, J. E.; Wigton, A. The reactivity of natural organic matter to disinfection by-products formation
(26)
(27)
(28)
(29)
(30)
and its relation to specific ultraviolet absorbance. Water Sci. Technol. 2001, 43 (2), 9-16. Weishaar, J. L.; Aiken, G. R.; Bergamaschi, B. A.; Fram, M. S.; Fujii, R.; Mopper, K. Evaluation of specific ultraviolet absorbance as an indicator of the chemical composition and reactivity of dissolved organic carbon. Environ. Sci. Technol. 2003, 37 (20), 4702-4708. Korshin, G. V.; Li, C. W.; Benjamin, M. M. The decrease of UV absorbance as an indicator of TOX formation. Water Res. 1997, 31 (4), 946-949. Korshin, G. V.; Wu, W.; Benjamin, M. M.; Hemingway, O. Correlations between differential absorbance and the formation of individual DBP species. Water Res. 2002, 36 (13), 32733282. Korshin, G. V.; Benjamin; M. M.; Hemingway, O.; Wu, W. Development of Differential UV Spectroscopy for On-line DBP Monitoring; AWWA Research Foundation and American Water Works Association: Denver, CO, 2002. Korshin, G. V.; Benjamin, M. M.; Xiao, H. B. Interactions of chlorine with natural organic matter and formation of intermediates: Evidence by differential spectroscopy. Acta Hydrochim. Hydrobiol. 2001, 28 (7), 378-384.
Received for review September 22, 2006. Revised manuscript received January 30, 2007. Accepted February 16, 2007. ES062268H
VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2781