Reduction of Disinfection Byproduct Formation by Molecular

Reduction of Disinfection Byproduct. Formation by Molecular. Reconfiguration of the Fulvic. Constituents of Natural Background. Organic Matter†. WAL...
0 downloads 0 Views 378KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 6446-6452

Reduction of Disinfection Byproduct Formation by Molecular Reconfiguration of the Fulvic Constituents of Natural Background Organic Matter† WALTER J. WEBER, JR.,* QINGGUO HUANG, AND ROGER A. PINTO Energy and Environment Program, Department of Chemical Engineering, University of Michigan, Ann Arbor, Michigan 48109-2099

Experiments were performed to assess the effects of treating the fulvic acid fractions of background natural organic matter (NOM) by catalyst-induced oxidative coupling reactions. Changes in the molecular characteristics of the fulvic acids and related disinfection byproduct formation potentials of these important NOM constituents were investigated. The coupling reactions were induced by addition of horseradish peroxidase (HRP) and hydrogen peroxide to aqueous solutions of the fulvic acids (FAs) in semicontinuous flow reactors. Subsequent removal of organic matter by ultrafiltration was found to be markedly enhanced for FA solutions subjected to oxidative coupling treatment. Uniform formation condition tests further indicated that the disinfection byproducts formed upon chlorination of FAs treated via induced oxidative coupling were reduced significantly on a unit carbon basis relative to those formed upon chlorination of their untreated counterparts. Spectroscopic examinations revealed that the FA molecules were effectively reconfigured in the oxidative coupling reactions. Substantial conversion of aromatic hydroxyl groups into ether-bonded moieties is evident, and a loss of primary amine groups, probably through conversion into secondary or tertiary amines, was also observed. These conversions apparently result in cross-linking of the natural FA moieties to form stable species of larger sizes, thus rendering them more readily removable by ultrafiltration and less reactive with chlorine. The results of the study may be interpreted as indicating that catalytically induced oxidative coupling reactions of the type conducted in this work can be combined with ultrafiltration to provide an effective scheme for removal of disinfection byproduct precursors.

Introduction The natural organic matter (NOM) present in virtually all drinking water sources poses a variety of problems in treatment operations and distribution systems (1, 2), among which the formation of disinfection byproducts (DBPs) is perhaps of greatest concern (3, 4). Disinfection reagents, chlorine in particular, can react with NOM constituents via a combination of substitution and oxidation mechanisms to †

This paper is part of the Charles O’Melia tribute issue. * Corresponding author phone: (734)763-2274; fax: (734)936-4391; e-mail: [email protected]. 6446

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 17, 2005

yield such fragmented byproducts as halogenated C1-C3 aliphatics, acids, aldehydes, and ketones, many of which have been shown to be carcinogenic, mutagenic, hepatotoxic, and/or to cause adverse reproductive and developmental effects (2). DBP formation can be minimized by reducing overall NOM concentrations prior to points of disinfection, methods for which include enhanced coagulation, granular activated carbon (GAC) sorption, and membrane filtration (5). These methods have proven effective to varying degrees, depending on process conditions and NOM characteristics, but each invariably suffers one or more serious limitations. Coagulation, for example, selectively removes larger NOM species but is relatively ineffective for removal of smaller fulvic acid constituents (6, 7). Applications of GAC for background NOM removal are impaired by difficulties associated with the regeneration of NOM-preloaded carbons. GAC adsorption processes are, moreover, generally designed to remove specifically targeted organic contaminants, and NOM adsorption can result in decreased GAC performance in this regard by preloading active sites prior to their exposure to the contaminants targeted for removal (8-10). A number of studies have demonstrated that satisfactory background NOM removal can be achieved by nanofiltration (NF) and reverse osmosis (RO), but the pores of the membranes used for ultrafiltration (UF) and microfiltration (MF) are too large to reject many important NOM constituents (6, 11). The costs associated with NF and RO processes are generally much greater than those of UF and MF because of the higher driving forces and more demanding maintenance schedules required. The cost disadvantages of NF and RO are exacerbated by the fact that most NOMs fall in a size range that can readily cause membrane fouling (7). There is scientific evidence that NOM molecules may under appropriate conditions be effectively reconfigured by oxidative coupling, a class of reactions that facilitate the polymerization of molecules having phenolic or anilinic features (12, 13). These reactions can be catalyzed by a variety of naturally occurring extracellular enzymes and mineral oxides. For example, horseradish peroxidase (HRP) mediates the single-electron oxidation of a variety of phenolic and anilinic compounds, yielding their respective free radicals. The resultant radicals can then couple with each other by C-O-C, C-N-C, and/or C-C bonds to form polymeric products (14-18). Those products of the coupling reactions that are soluble can continue to serve as substrates for further oxidative coupling and form larger polymers. Polymerizations driven by oxidative coupling are in fact central to natural humification processes, leading to the formation and growth of soil organic matter from smaller building-block moieties (19-21). It may be hypothesized that oxidative coupling reactions that effectively increase certain NOM constituent sizes via sequential covalent linkages will render them more readily removed in such water treatment processes as coagulation, deep-bed filtration, or UF/MF. The reactivity of aquatic humic substances with halogens is generally thought to be associated with the “activated” aromatic content of these organic substances, e.g., aromatic rings bearing -OH, -NH2, and heterocyclic nitrogen atoms (22, 23). It may thus be further hypothesized that such NOM constituents when reconfigured through oxidative reactions may be much less reactive with disinfectants as a result of the conversion of phenolic and anilinic moieties into ether and secondary amine bonds and thus less able to function as DBP precursors. The work described here was designed to test the two foregoing hypotheses and to evaluate the 10.1021/es050220i CCC: $30.25

 2005 American Chemical Society Published on Web 07/09/2005

feasibility of using HRP-mediated oxidative coupling reactions in combination with ultrafiltration to mitigate DBP formation in potable water disinfection practice.

Experimental Section Materials. Three fulvic acids were tested as model NOM constituents. Suwannee River Fulvic Acid (SRFA) was obtained from the International Humic Substances Society (IHSS). Fulvic acids extracted from Canadian Peat (CPFA) and Chelsea Soil (CSFA) were prepared in our laboratory using the standard IHSS protocol (24). The fulvic acids were reconstituted in 1.0 M borate buffer (pH ) 8.0), filtered through 0.45 µm membranes, and stored in the dark at 4 °C prior to use. An analytical standard mixture of trihalomethanes (THMs) was provided by Supelco (Bellefonte, PA), and one mixture of haloacetic acids (HAAs) by Ultra Scientific (North Kingstown, RI). Sodium hypochlorite was obtained from Aldrich (St. Louis, MO). Extracellular horseradish peroxidase (HRP, type-1, RZ ) 1.3) and hydrogen peroxide (30.8%, ACS reagent) were provided by Sigma Chemical Co. (St. Louis, MO). HRP stock solutions were prepared freshly each day of an experiment, and activities were assessed by the ABTS method described elsewhere (25). One unit of HRP activity is defined as that amount catalyzing the oxidation of 1 µmol of ABTS per minute. Enzymatic Reactions. Enzymatic reactions were performed at room temperature using 50 mL flat-bottom volumetric flask reactors operated in semicontinuous flow mode. Each reactor initially contained 25 mL of a 1.0 M borate buffer (pH 8.0) solution of fulvic acid comprising 25 mg L-1 of total organic carbon (TOC), and the solution was continuously mixed by a glass-sealed magnetic stir bar. HRP and hydrogen peroxide stock solutions were continuously delivered to the reactor at a rate of 0.5 mL h-1 using two separate KD Scientific syringe pumps, so that the input rates for HRP and hydrogen peroxide were, respectively, 0.5 units mL-1 h-1 and 1 mM h-1. Therefore, over a typical treatment period of 8 h, a total of 4 units mL-1 of HRP and 8 mM H2O2 was added to the reactor. NOM solutions containing only hydrogen peroxide or HRP were used as controls. Ultrafiltration. Reagent delivery was terminated when a reaction had proceeded to a predetermined time, and the product solution then was subjected to ultrafiltration using a Minimate tangential flow ultrafiltration system equipped with a mini-ultrasette membrane cartridge (Pall Life Sciences, Ann Arbor, MI). This cartridge contains a low-protein-binding membrane comprising polyethersulfone on a polyethylene substrate, having an effective filtration surface of 50 cm2 and a nominal molecular weight cutoff of 3 kDa. The membrane was thoroughly backwashed with borate buffer after each filtration procedure, reconditioned with a pH 12 sodium hydroxide solution, and then rinsed with 50 mL of deionized water followed by the same volume of borate buffer (1 M, pH 8). Filtration of the reaction product solutions was conducted at a constant feed flow rate of 2 mL min-1, with the ratio of retentate/filtrate flows adjusted to 3:1. Filtration was stopped when 80% of the initial volume was collected as filtrate. Samples were taken from the filtrate for measurement of total organic carbon (TOC) content and assayed for DBP formation potential (DBP-FP) by the methods described below. TOC Analysis and DBP-FP Assay. Prior to TOC analysis and DBP-FP assay, 50 µL of catalase solution (10 units mL-1) was added to each mL of the sample, and these solutions then mixed for 10 min to remove any residual H2O2. We found that residual H2O2 can interfere with TOC measurements. It is also known that H2O2 interferes with the DBP-FP test by reducing the efficiency of chlorination (24). This amount of added catalase corresponds to 0.015 mg L-1 of lyophilized

catalase upon dilution in the sample solutions, and preliminary tests indicated that the organic content from this amount of catalase does not yield noticeable amounts of DBPs during the standard chlorination process used in the DBP-FP test. UV absorbance of the solution was measured using a Jenway 6405 UV-vis spectrophotometer. TOC concentrations in the sample solutions were measured using a Shimadzu 9000 TOC analyzer. Additional tests showed that HRP was completely retained upon filtration. This indicated that the TOC remaining in the filtrates was attributable solely to fulvic acids (FAs) not removed through the combined reaction and filtration processes. The ratio of the filtrate TOC to the TOC of the FA solution originally added to the reactor was then calculated. The filtrates were then diluted with borate buffer to normalize the TOC level to 4.0 mg L-1 and then assayed for DBP-FP as described below. DBP-FP was assessed using the uniform formation conditions (UFCs) described by Summers et al. (27). The test was performed in a headspace-free amber bottle containing 20 mL of the TOC-normalized sample solution. A 50 µL aliquot was withdrawn from this solution and replaced with the same volume of a pH 8 borate buffer solution of hypochlorite prepared by blending sodium hypochlorite with a pH 6.7 borate buffer. The chlorine demands of the FA solutions were pretested using a four-point dosage calibration. From the results of this test, the chlorine dose for the UFC was established as 2.25:1.0 (mg Cl2/mg TOC). This condition yielded a 24 h residual of 1.0 ( 0.4 mg L-1 free chlorine, which is representative of routine water treatment operating conditions. Blank tests were performed with background borate buffer and with the FA solutions. DBP-FP was characterized by the total yield of trihalomethanes (THMs) and haloacetic acids (HAAs) per unit mass of organic carbon after 24 h of chlorination as described above. Because no bromine or bromides were present in our samples, no brominated DBP species were formed, and the DBP-FPs measured were thus chloroform as representative of THMs and mono-, di-, and trichloroacetic acids as representative of HAAs. For the same reason, the DBP yields and speciations obtained in these tests may not necessarily be comparable to results obtained from tests with certain natural water samples. The 20 mL chlorinated samples were extracted using 3 mL of methyl tert-butyl ether (MTBE) prior to DBP analyses. THMs were measured using USEPA method 551.1, with certain adaptations to the gas chromatography (GC) procedure employed. The GC characterizations in this work were performed on a Hewlett-Packard 5890 Series II gas chromatograph equipped with a DB-5 column (0.25 mm in diameter, 0.25 µm film thickness, J & W Scientific/Agilent, Folsom, CA) and an electron capture detector. Nitrogen was used as the carrier gas at the flow rate of 6.4 mL min-1. A 2 µL MTBE extract of the sample solution was used for GC analysis in splitless mode, and temperatures for the injector, oven, and detector were maintained constant at 200, 60, and 300 °C, respectively. Tests showed that this GC program resulted in a complete separation of THM species within 8 min. HAAs were quantified by USEPA method 552.2, again with certain adaptations to the GC procedure. The acidic methanol esterifications of the MTBE extract were done as described in USEPA method 552.2 for sample preparation, but the same GC system and conditions described above were employed, except that the oven temperature was initially maintained at 36° C for 10 min, increased to 75° C at 5° C min-1, and maintained at this temperature for 3 min. The standard HAA-5 species was resolved within 15 min of GC time. FTIR Characterization. FTIR spectra of NOM were obtained using a Thermo Nicolet Nexus 670-ESP FTIR analyzer. A horizontal attenuated total reflectance (HATR) device containing a ZnSe prism was used as the sample VOL. 39, NO. 17, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6447

FIGURE 1. Fulvic TOC removal through combined oxidative coupling and ultrafiltration. Abscissa values indicate the amounts of HRP cumulatively added to the semicontinuous flow reactors in which the FA was treated. Reaction conditions and reactor operations are described in detail in the Experimental Section. Error bars indicate standard deviations of mean values of triplicate experiments. The dashed line indicates the average of TOC removal by ultrafiltration of control samples with 4 units mL-1 of HRP but no H2O2 added. interface. Aliquots of 200 µL in volume were spread on the surface of the ZnSe receptacle and evaporated in a freezedryer. IR signals were acquired by averaging 128 scans ranging from 600 to 4000 cm-1 at a resolution of 0.964 cm-1. FTIR absorbances were quantified by integration over selected wavenumber regions using T.Q. Analyst (version 6.1.1.356) for Thermo Nicolet.

Results and Discussion TOC Removal. Figure 1 illustrates fractional removals of TOC effected by ultrafiltration for the three different FA solutions treated to varying extents by HRP-induced oxidation reactions in the semicontinuous flow reactors. As described in the Experimental Section, HRP and hydrogen peroxide were delivered simultaneously to the reactor at rates of 0.5 units mL-1 h-1 and 1 mM h-1, respectively, but only cumulative HRP addition alone is represented on the abscissa of Figure 1 as the indication of reaction extent. As evident in Figure 1, fulvic TOC removal is enhanced significantly for FA solutions in which oxidative reactions have been induced. Near 60% removal of fulvic TOC was achieved for all three different FA mixtures after a total of 2 units mL-1 of HRP (and 4 mM H2O2) was added, while the removal for the untreated solutions was only about 5%. Ultrafiltration removal efficiencies for the control samples of each FA solution containing only 4 units mL-1 of HRP and no H2O2 were also tested. An average of 11 ( 1 % retention of fulvic TOC resulted, as indicated by the dashed line in Figure 1. This average level of removal is slightly greater than the removal levels of any of the three completely untreated FA solutions (i.e., no HRP), a result that might be attributable to interactions between the FAs and HRP. Additional tests showed that HRP, the molecular weight of which is reportedly 20 kDa, was completely retained by the membrane, which had a nominal molecular weight cutoff of 3 kDa. Reduction of DBP-FP. Figure 2 shows representative DBP formation per unit mass of organic carbon values obtained from UFC chlorination tests on the three FA solutions treated to different extents by induced oxidative coupling reactions and followed by ultrafiltration. Both chloroform (part A) and chloroacetic acid formation (part B) were reduced for all samples treated by oxidative coupling. Comparison tests in which each of the three FA samples was treated with H2O2 at a level of 8 mM but with no HRP added were then conducted. The product solutions were subjected to ultrafiltration and assessed for DBP-FP using 6448

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 17, 2005

FIGURE 2. (A) Chloroform and (B) chloroacetic acid formation in uniform condition tests for FA solutions treated to varying extents by HRP-mediated reactions followed by ultrafiltration. The term chloroacetic acid formation represents the total amount of mono-, di-, and trichloroacetic acids. Abscissa values indicate the amount of HRP cumulatively added to the semicontinuous flow reactors. Error bars indicate standard deviations of mean values of triplicate experiments. the UFC chlorination test. Changes in fulvic TOC levels following this treatment process are summarized in part A of Figure 3. For comparison, TOC changes in the samples treated with both HRP and H2O2 are shown in part B of Figure 3. The DBP-FPs of the fulvic solutions treated by these two different processes are compared in Figure 4. The differences in TOC levels before and after peroxide-only treatment shown in part A of Figure 3 indicate that the FAs were partially oxidized by the H2O2. One thing to note regarding the relatively high levels of oxidation observed is that catalase was added to the samples prior to TOC analysis and the DBP-FP assay to remove residual hydrogen peroxide, which can severely interfere with TOC measurements and DBP-FP tests. Catalase is a heme-containing enzyme whose primary function is to catalyze the conversion of H2O2 into O2 and H2O (29). However, the following possibilities cannot be ruled out: (i) Active oxygen species formed during the catalaseH2O2 reaction may play a role in FA oxidation, and/or (ii) H2O2 along with the denatured catalase can mediate Fentontype reactions of FA. Therefore, the FA degradation evident in Figure 3 may not necessarily be completely the result of straightforward reactions with H2O2 but may possibly involve catalase catalysis as well. In contrast to the decrease of fulvic content by reaction with H2O2 shown in Figure 3A, Figure 3B shows that fulvic TOC levels are maintained in solution at levels comparable to those before reaction for the solutions treated by HRPmediated oxidative coupling but that removal by subsequent ultrafiltration is increased significantly. These results, re-

FIGURE 3. Fulvic TOC changes at different steps of the treatment process. Part A shows samples that were treated with a total of 8 mM H2O2 alone, followed by ultrafiltration. Part B shows those treated with a total of 4 units mL-1 of HRP and 8 mM H2O2, followed by ultrafiltration. The reactions were performed in semicontinuous flow reactors, and the reaction conditions and the reactor operations are described in detail in the Experimental Section. Error bars indicate standard deviations of mean values of triplicate experiments. spectively, shown in Figures 3A and 3B are indicative of two distinctly different reaction phenomena and mechanisms, i.e., oxidative decomposition for treatments by H2O2 without HRP and oxidative coupling and polymerization when both H2O2 and HRP are present. Figure 4 shows that formation potentials for chloroform and chloroacetic acids for samples treated with H2O2 but without HRP were generally increased on a unit TOC basis compared to corresponding DBP-FP levels obtained for untreated control samples subjected to ultrafiltration. In other words, H2O2 oxidation was capable of lowering levels of fulvic TOC in solution, but the chemical and structural changes effected by oxidative decomposition of the components of that TOC contributed to increased levels of DBP-FPs. Conversely, the DBP-FPs of all samples treated by HRPfacilitated oxidative coupling were noticeably reduced with respect to formation potentials for both chloroform and chloroacetic acids. Molecular Basis. FTIR spectra were collected on SRFA samples subjected to the different reaction conditions followed by ultrafiltration, as illustrated in Figure 5. To facilitate comparisons of these spectra, FITR absorbances at specific wavenumber regions were quantified by integration over selected wavenumber regions using T.Q. Analyst (version 6.1.1.356) for Thermo Nicolet. The structural assignments of specific IR spectrum regions examined are given in Table 1. The integrated FTIR absorbances in the regions examined were all normalized to the absorbance in the aromatic carbon

FIGURE 4. (A) Chloroform and total (B) chloroacetic acid formation in uniform condition tests for FA solutions for samples treated with a total of 8 mM H2O2 alone and with 4 units mL-1 of HRP and 8 mM H2O2; all samples were subjected to ultrafiltration after these respective treatments. Chloroacetic acid formation represents the total amount of mono-, di-, and trichloroacetic acids. FA solutions that were not treated but were subjected to ultrafiltration were used as controls. Error bars indicate the standard deviations of mean values of triplicate experiments. region (1615-1580 cm-1), again to facilitate quantitative comparisons across different samples. As noted earlier, oxidative coupling serves principally to facilitate polymerization reactions by forming C-O-C, C-N-C, and/or C-C bonds (14-18), and the aromatic content of the molecules is conserved throughout these reactions. The specific absorbances obtained through normalization to the aromatic content are compared in Figure 6 for the different reaction conditions and associated controls. Figure 6 clearly shows a large reduction in aromatic hydroxyl (phenolic) responses and an increase in aromatic ether responses for solutions treated with HRP, strongly indicating cross-linking of NOM moieties by conversion of aromatic hydroxyl groups into ether bonds. We believe that the reductions in primary amine functionality evident in Figure 6 also result from the cross-coupling reactions of anilines, converting primary amines into secondary or tertiary amines, although we are not able to identify a band specific to secondary or tertiary amines to characterize such increases. In addition, because oxidative coupling serves mainly to facilitate polymerization reactions by forming C-O-C, C-N-C, and/or C-C bonds (14-18), it is not expected that these reactions would significantly alter the aliphatic or carbonyl contents of fulvic NOMs, and the FTIR absorbances shown in Figure 6 confirm this. For the sample treated with H2O2 but without HRP, the relative content of aliphatic and carbonyl moieties increases, but phenolic, amine, and aryl ether contents remain relatively unchanged. These structural changes support the concept of direct peroxide oxidation of FAs in the absence of HRP in that the aromatic components are probably more prone to such oxidation than the aliphatic components, with carbonyl groups likely forming as a result of direct oxidative reactions. VOL. 39, NO. 17, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6449

FIGURE 5. FTIR spectrum for the untreated SRFA solution subjected to ultrafiltration (as a reference control) and spectra for SRFA samples treated in the semicontinuous flow reactors with a total of 8 mM H2O2 alone and with 4 units mL-1 of HRP and 8 mM H2O2, then subjected to ultrafiltration. Reaction conditions and reactor operations are described in detail in the Experimental Section. It has been shown that oxygen-rich ketones and keto acids are factors that can enhance the reactivities of NOM species with chlorine to form DBPs (22, 30, 31). As such, the relative increases in carbonyl functionality of SRFA resulting from hydrogen peroxide oxidation shown in Figure 6 may have contributed to the generally increasing trend in its DBP-FP shown in Figure 4. In stark contrast, oxidative treatment involving HRP facilitates cross-linking of fulvic molecules by converting phenolic groups into aryl ethers and primary amines into secondary and tertiary amines. As noted earlier, phenolic and anilinic groups are also activating features of DBP precursors. Reductions in such functionalities through HRP-mediated oxidative coupling result in DBP-FP reduction, as shown in parts A and B of Figure 2. In addition, oxidative 6450

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 17, 2005

coupling of FAs forms species having larger molecular sizes through cross-linking of fulvic molecules. This apparently leads to enhanced retention of these NOM constituents by ultrafiltration, as evident in Figure 1. Recent studies by Piccolo and co-workers (12, 13, 32) indicate that humic substances in solution are self-associated supramolecules of relatively small molecules, and their sizes are increased through HRPmediated reactions at both molecular and supramolecular levels. UV absorbance was also measured on the samples for which FTIR results are shown in Figures 5 and 6. Molar absorptivity at 280 nm specific to molar concentrations of organic carbon (280) was calculated for those samples. The 280 values for the untreated SRFA control sample and those

TABLE 1. Structural Assignments of Specific FTIR Spectral Absorbance Wavenumber Regionsa region (cm-1)

assignments

3640-3530 2935-2915

phenolic O-H stretch C-H asymmetric/ symmetric stretch in -CH2carbonyl >CdO aromatic ring CdCsC stretch aryl-O stretch C-NH2 stretch

1725-1705 1615-1580 1270-1230 1090-1020 a

chemical functionalities aromatic hydroxyl saturated aliphatic

erization of NOM moieties to form larger and more stable species. These reaction products are thus more readily removed by ultrafiltration. Oxidative coupling reactions also effectively reduce active phenolic and anilinic moieties in the NOMs, thus mitigating their respective roles as DBP precursor agents.

Acknowledgments ketone, carboxylic acid aromatic unsaturation aromatic ether primary amine

Adapted from Coates (28).

This research was supported by Grant No. DE-FG0702ER63488 from the Environmental Management Science Program of the United States Department of Energy. The content of the paper does not necessarily represent the views of the funding agency. The authors thank the four anonymous reviewers of the paper for their careful readings and constructive comments.

Literature Cited

FIGURE 6. FTIR absorbances for SRFA samples at selected wavenumber regions normalized to aromatic carbon absorbance (1615-1580 cm-1). The samples were treated with 8 mM H2O2 alone or with 8 mM H2O2 and 4 units mL-1 of HRP and then subjected to ultrafiltration. An SRFA solution that was not treated but was subjected to ultrafiltration served as a control. FTIR absorbances were quantified by integration over selected wavenumber regions using T.Q. Analyst (version 6.1.1.356) for Thermo Nicolet. Error bars indicate standard deviations of mean values of triplicate experiments. for the reaction samples with hydrogen peroxide in the absence and presence of HRP are 508, 443, and 526 cm-1 M-1, respectively. It has been found that 280 correlates roughly to NOM aromaticity and molecular weight (33). The changes in the specific UV absorbances for the two reaction samples relative to the control sample, although relatively small, are in line with what we expected for the effects of different reactions, i.e., oxidative decomposition for treatments by H2O2 without HRP and oxidative coupling and polymerization when both H2O2 and HRP are present. It has also been generally thought that NOMs having higher specific UV absorbance tend to have greater DBP-FPs, but recent studies indicate that such correlations are not robust (34), because DBP-FPs can be affected by many factors other than aromaticity or molecular sizes of NOMs. Our data corroborate the point that specific UV absorbance may not necessarily be a good indicator of DBP-FP. Apparently, the structural changes reflected by the FTIR absorbance changes shown in Figure 6 are not readily reflected by specific UV absorbances. Preoxidation treatment of water involving the use of hydrogen peroxide and/or ozone has been proposed as a measure for DBP control (35). The data presented here however shows that hydrogen peroxide oxidation alone, although potentially reducing overall TOC levels, significantly increases DBP formation on a unit carbon basis, probably as a result of the oxidative breakdown of larger NOM molecules. NOM treatment with both hydrogen peroxide and oxidative coupling catalysts (e.g., HRP) is fundamentally different in terms of process mechanisms, with oxidative coupling being the primary reaction, one leading to polym-

(1) Karanfil, T.; Kilduff, J. E.; Schlautman, M. A.; Weber, W. J., Jr. The oxygen sensitivity of organic macromolecule sorption by activated carbon: Effects of solution chemistry. Water Res. 1998, 32, 154-164. (2) Singer, P. C.; Reckhow, D. A. Chemical oxidation. In Water Quality and Treatment: A Handbook of Community Water Supplies, 5th ed.; American Water Works Association, Ed.; McGraw-Hill: New York, 1999; Chapter 12. (3) Singer, P. C.; Reckhow D. A. Control of disinfection by-products in drinking-water. J. Environ. Eng. 1994, 120, 727-744. (4) Kitis, M.; Karanfil, T.; Kilduff, J. E.; Wigton, A. The reactivity of natural organic matter to disinfection byproducts formation and its relation to specific ultraviolet absorbance. Water Sci. Technol. 2001, 43, 9-16. (5) Richardson, S. D.; Simmons, J. E.; Rice, G. Disinfection byproducts: The next generation. Environ. Sci. Technol. 2002, 36, 198A205A. (6) Minear, R. A.; Amy, G. L. Water disinfection and natural organic matter: History and overview. In Water Disinfection and Natural Organic Matter; Minear, R. A., Amy, G. L., Eds.; ACS Symposium Series 649; American Chemical Society: Washington, DC, 1996; pp 1-9. (7) Water Quality and Treatment: A Handbook of Community Water Supplies, 5th ed.; American Water Works Association, Ed.; McGraw-Hill: New York, 1999. (8) Karanfil, T.; Kilduff, J. E.; Schlautman, M. A.; Weber, W. J., Jr. Adsorption of organic macromolecules by granular activated carbon. 1. Influence of molecular properties under anoxic solution conditions. Environ. Sci. Technol. 1996, 30, 2187-2194. (9) Karanfil, T.; Schlautman, M. A.; Kilduff, J. E.; Weber, W. J., Jr. Adsorption of organic macromolecules by granular activated carbon. 2. Influence of dissolved oxygen. Environ. Sci. Technol. 1996, 30, 2195-2201. (10) Weber, W. J., Jr. Preloading of GAC by natural organic matter in potable water treatment systems: Mechanisms, effects and design considerations. J. Water Supply Res. Technol. 2004, 53, 469-482. (11) Chellam, S. Effects of nanofiltration on trihalomethane and haloacetic acid precursor removal and speciation in waters containing low concentrations of bromide ion. Environ. Sci. Technol. 2000, 34, 1813-1820. (12) Piccolo, A.; Cozzolino, A., Conte P., Spaccini, R. Polymerization of humic substances by an enzyme-catalyzed oxidative coupling. Naturwissenschaften 2000, 87, 391-394. (13) Cozzolino, A.; Piccolo, A. Polymerization of dissolved humic substances catalyzed by peroxidase. Effects of pH and humic composition. Org. Geochem. 2002, 33, 281-294. (14) Simmons, K. E.; Minard, R. D.; Bollag, J. M. Oxidativeoligomerization of guaiacol and 4-chloroaniline. Environ. Sci. Technol. 1989, 23, 115-121. (15) Yu, J.; Taylor, K. E.; Zou, H.; Biswas, N.; Bewtra, J. K. Phenol conversion and dimeric intermediates in horseradish peroxidase-catalyzed phenol removal from water. Environ. Sci. Technol. 1994, 28, 2154-2160. (16) Park, J. W.; Dec, J.; Kim, J. E.; Bollag, J. M. Effect of humic constituents on the transformation of chlorinated phenols and anilines in the presence of oxidoreductive enzymes or birnessite. Environ. Sci. Technol. 1999, 33, 2028-2034. (17) Oguchi, T.; Tawaki, S.; Uyama, H.; Kobayashi, S. Soluble polyphenol. Macromol. Rapid Commun. 1999, 20, 401-403. (18) Zou, H.; Taylor, K. E. Products of oxidative coupling of phenol by horseradish peroxidase. Chemosphere 1994, 28, 1807-1817. VOL. 39, NO. 17, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6451

(19) Bollag, J. M. Decontaminating soil with enzymes. Environ. Sci. Technol. 1992, 26, 1876-1881. (20) Bollag, J. M. Enzymes catalyzing oxidative coupling reactions of pollutants. Met. Ions Biol. Syst. 1992, 28, 205-217. (21) Huang, Q. G.; Selig, H.; Weber, W. J., Jr. Peroxidase-catalyzed oxidative coupling of phenols in the presence of geosorbents: Rates of non-extractable product formation. Environ. Sci. Technol. 2002, 36, 596-602. (22) Norwood, D. L.; Johnson, J. D.; Christman, R. F.; Hass, J. R.; Bobenrieth, M. J. Reactions of chlorine with selected aromatic models of aquatic humic material. Environ. Sci. Technol. 1980, 14, 187-190. (23) Reckhow, D. A.; Singer P. C.; Malcom, R. L. Chlorination of humic materials: Byproduct formation and chemical interpretations. Environ. Sci. Technol. 1990, 24, 1655-1664. (24) Swift, R. S. Organic matter characterization. In Methods of Soil Analysis; Sparks, D. L., Ed.; SSSA: Madison, WI, 1996; Part 3, pp 1011-1069. (25) Putter, J.; Becker, R. Peroxidases. In Methods of Enzymatic Analysis; Bergmeyer, H. U., Bergmeyer, J., Grassl, M., Eds.; Verlag Chemie: Weinheim, Germany, 1983; pp 286-293. (26) Liu, W. J.; Andrews, S. A.; Stefan, M. I.; Bolton, J. R. Optimal methods for quenching H2O2 residuals prior to UFC testing. Water Res. 2003, 37, 3697-3703. (27) Summers, R. S.; Hooper, S. M.; Shukairy, H. M.; Solarik, G.; Owen D. Assessing DBP yield: Uniform formation conditions. J. Am. Water Works Assoc. 1996, 88, 80-93. (28) Coates, J. Interpretation of infrared spectra: A practical approach. In Encyclopedia of Analytical Chemistry; Meyers, R. A., Ed.; John Wiley & Sons: Chichester, U. K., 2000; pp 1081510837.

6452

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 17, 2005

(29) Jones, P. Roles of water in heme peroxidases and catalase mechanisms. J. Biol. Chem. 2001, 276, 13791-13796. (30) 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, 202-211. (31) Reckhow, D. A.; Singer P. C. Mechanisms of organic halide formation during fulvic acid chlorinations and implications with respect to preozonation. In Water Chlorination: Environmental Impacts and Health Effects; Jolley, R. L., Ed.; Ann Arbor Science Publishers: Ann Arbor, MI, 1985. (32) Conte, P.; Piccolo, A. Conformation arrangement of dissolved humic substances. Influence of solution composition on association of humic molecules. Environ. Sci. Technol. 1999, 33, 1682-1690. (33) Peuravuori, J.; Pihlaja, K. Molecular size distribution and spectroscopic properties of aquatic humic substances. Anal. Chim. Acta 1997, 337, 133-149. (34) 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, 47024708. (35) Kleiser, G.; Frimmel, F. H. Removal of precursors for disinfection by-products (DPBs)sDifferences between ozone- and OHradical-induced oxidation. Sci. Total Environ. 2000, 256, 1-9.

Received for review February 1, 2005. Revised manuscript received June 15, 2005. Accepted June 16, 2005. ES050220I