Chlorination Byproduct Formation in the Presence ... - ACS Publications

glycylglycine, diethylamine, and methylamine) were also evaluated for comparison. The formation of THMs and HAAs was found to decrease with increasing...
1 downloads 0 Views 216KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 4995-5001

Chlorination Byproduct Formation in the Presence of Humic Acid, Model Nitrogenous Organic Compounds, Ammonia, and Bromide XIN YANG AND CHII SHANG* Department of Civil Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong

The formation of trihalomethanes (THMs), haloacetic acids (HAAs), and cyanogen halides (CNXs) after chlorination of synthetic solutions containing humic acid, nitrogenous organic (N-organic) compounds, ammonia, and bromide ions was studied. Humic acid (from Aldrich) was used to provide the source of the precursors. Glycine was chosen as the primary model N-organic compound and other four model N-organic compounds (including glutamic acid, glycylglycine, diethylamine, and methylamine) were also evaluated for comparison. The formation of THMs and HAAs was found to decrease with increasing glycine and ammonia concentrations but to increase with increasing bromide ion concentration. CNX formation was found to be highly sensitive to free chlorine to glycine ratios, and its formation trends were significantly affected by the presence/ absence of ammonia. The incorporation of bromine changed the byproducts speciation toward brominated species and enhanced the yields of total THMs, HAAs, and CNXs. Different model N-organic compounds exerted different effects on the formation of THMs, HAAs, and CNXs. Their effects on the formation of THMs and HAAs were likely dependent on their reactivity to chlorine in competing with the humic acid chlorination reactions. The difference in the CNCl yields was attributable to the variations in the compound structures.

Introduction Chlorination is a well-developed and widely used process for water and wastewater disinfection because of its broadspectrum germicidal potency, low cost, and well-established practices. However, the reactions between chlorine and natural organic matter (NOM) produce a large range of disinfection byproducts (DBPs) (1), of which trihalomethanes (THMs) and haloacetic acids (HAAs) are the most prevalent ones by weight (2). The formation of THMs and HAAs has been reported to be sensitive to many factors such as pH, temperature, contact time, bromide ion concentrations, forms and concentrations of precursors, chlorine to ammonia-nitrogen ratios, and forms and concentrations of disinfectants and their application techniques (3). Because some of these DBPs have been identified as cancer-causing agents (4), the U.S. EPA set the current maximum contaminant levels (MCLs) of 80 and 60 µg/L for total THMs and HAA5 (the sum of five HAAs), respectively, in the Stage 1 Disinfectants/DBP Rule (5). * Corresponding author telephone: (852)2358-7885. fax: (852)2358-1534; e-mail: [email protected]. 10.1021/es049580g CCC: $27.50 Published on Web 08/20/2004

 2004 American Chemical Society

During chlorination, nitrogenous organic (N-organic) compounds can also react with chlorine to form organic chloramines (6). The interactions interfere in the formation of THMs and HAAs and produce specific DBPs. For example, a high correlation has been reported between the reactivity of N-organic compounds and chloroform formation (7). Compounds (e.g., glycine) with relatively high reactivity to chlorine suppressed chloroform formation but less-reactive compounds (e.g., urea and uracil) inhibited chloroform formation only at high chlorine concentrations. Decomposition of some organic chloramines produced dichloroacetonitrile (DCAN), which was reported to serve as an intermediate to yield dichloroacetic acid (DCAA), trichloroacetic acid (TCAA), and chloroform (8). In addition, some N-organic compounds were deemed as the precursors to form other DBPs such as cyanogen chloride (CNCl) (9). CNCl, a highly toxic compound with 24- and 48-h LC50 values, which are doses that kill 50% of a population after the specific exposure time, to Daphnia magna (5-day-old) of 86 and 65 µg/L, respectively (10), can be formed from water chlorination and chloramination (11). CNCl was included in the U.S. EPA Information Collection Rule (ICR) and is being considered for further regulatory consideration (12). After chlorinating bromide-rich water, cyanogen bromide (CNBr) is formed (13), and higher concentrations of bromide ions result in increases in the yields of CNBr and cyanogen halides (CNXs) (14). These previous findings indicate that the presence of N-organic compounds complicates the chlorination process and the chlorine chemistry involved in formation of THMs, HAAs, and CNXs. The important roles of N-organic compounds on the formation of these DBPs in the co-presence of DBP precursors, ammonia and/or bromide ions, are unknown and require detailed studies. Therefore, the objective of this study was to assess the effects of N-organic compounds on the formation of THMs, HAAs, and CNXs during chlorination of humic acid (from Aldrich) solutions in the presence of ammonia and/or bromide ions. Glycine [NH2-CH2-COOH] was chosen as the primary model Norganic compound, and other model N-organic compounds including glutamic acid [HOOC-(CH2)2-CH(NH2)-COOH], diethylamine [(CH3CH2)2-NH], methylamine [CH3-NH2], and glycylglycine [HOOC-CH2-NH-CO-CH2-NH2] were also used for comparison. Membrane introduction mass spectrometry (MIMS) methods (9, 15) were used to differentiate inorganic chloramines from organic interferences and to quantify the concentrations of CNCl and CNBr.

Experimental Section Solution Preparation. Solutions were prepared from reagentgrade chemicals or stock solutions. Dilution to target aqueous-phase concentrations was accomplished with double-distilled, deionized (DDDI) water. A stock of free chlorine (HOCl) solution was prepared from 5% sodium hypochlorite (NaOCl) (from Allied Signal), diluted to 20003000 mg/L (as Cl2) and stored in aluminum foil-covered glassstoppered flasks. The stock solution was periodically standardized by DPD/FAS titration (16). Glycine (Sigma), glutamic acid (Nacalai Tesque), methylamine (Sigma), diethylamine (Aldrich), glycylglycine (Sigma), and ammonium chloride (BDH) were dissolved into dilution water to make stock solutions of 100 mg/L as nitrogen. A humic acid stock solution was prepared by dissolving an aliquot of humic acid crystals (from Aldrich) into dilution water and filtering it through a 0.45-µm filter paper (Advantec MFS, Inc.), which produced VOL. 38, NO. 19, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4995

a humic acid solution of 250 mg/L as dissolved organic carbon (DOC). The standard solutions of monochloramine (NH2Cl), dichloramine (NHCl2), and trichloramine (NCl3) were freshly prepared and titrimetrically standardized (15). Due to the prohibition on importing CNCl to Hong Kong and the short shelf life of CNCl, a CNCl standard solution was freshly developed in situ by 30-min chlorination of a glycine solution (0.5 mg/L as N) at a free chlorine dosage of 5 mg/L (as Cl2) and phosphate-buffered pH of 7. This protocol had earlier demonstrated the yield of a consistent aqueous CNCl concentration of 900 µg/L while leaving no residual chlorine in the solution (9). A CNBr (Sigma) standard solution was prepared gravimetrically with the dilution water. Analytical Methods. Methods based on the application of MIMS were used for determining the concentrations of chlorine residuals, CNCl and CNBr (9, 15). The MIMS system was based on an HP 5892 benchtop GC/MS (Hewlett-Packard) containing an HP 5972A mass selective detector (MSD). Details of the configuration, setup, and operational conditions of the MIMS system can be obtained from the literature (15). It is known to differentiate unambiguously inorganic chloramines from organic chloramines. The selected ion monitoring (SIM) mode was used, and ions at 53, 89, 119, 61, and 105 amu were monitored for quantification of inorganic mono-, di-, and trichloramines, CNCl, and CNBr, respectively, by directly comparing ion abundance measurements with those developed from a series of standard solutions. The corresponding method detection limits (MDL) for these target compounds were 0.08, 0.10, and 0.06 mg/L (as Cl2) and 1 and 1 µg/L, respectively. Concentrations of free chlorine were estimated by the difference in inorganic monochloramine concentrations before and after sample ammonification (15). HAA and THM analyses were carried out on a gas chromatograph (Finnigan TRACE GC) with an electron capture detector (ECD) based on U.S. EPA Methods 552.2 (17) and 551.1 (18), respectively. The column used was a DB-5MS fused silica capillary column (30 m × 0.25 mm i.d. with 0.25 µm film thickness; J&W Scientific). Experimental Procedures. Chlorination experiments were, in all cases, carried out in aluminum-foil wrapped, well-mixed, glass-stoppered bottles using synthetic solution buffered at pH 7.0 with 10 mM phosphate buffer. Humic acid (from Aldrich) solutions (at 5 mg/L as DOC) were used as the basic water matrix in all cases. Glycine, ammonia, and bromide ions at various concentrations were added to the solutions to create various chemical schemes. Three types of synthetic solutions were used to study the general trends of DBP formation under these different schemes. These solutions included a solution containing only humic acid; a solution containing humic acid and glycine (0.1 mg/L as N); and a solution containing humic acid, glycine (0.1 mg/L as N), and ammonia (2 mg/L as N), for convenience, hereafter referred to as “HA” solution, “HA + Gly” solution, and “HA + Gly + NH3” solution, respectively. Due to the ubiquitous presence of bromide ions in water (with an average concentration of approximately 0.1 mg/L; 19), bromide concentrations of 0, 0.1, 0.2, 1, and 2 mg/L were added to these synthetic solutions. For the study comparing the formation of the DBPs during chlorination of different N-organic compounds including glycine, glutamic acid, glycylglycine, methylamine, and diethylamine, solutions were prepared in a similar manner except that the concentration of the organic compounds was kept at 0.5 mg/L as N and only two water matrixes, solutions containing humic acid and one of the N-organic compounds with or without adding ammonia (hereafter referred to as “HA + Org-N” solution, and “HA + Org - N + NH3” solution, respectively) were used. Chlorination was performed by dosing sodium hypochlorite at 8 mg/L as Cl2 to the synthetic solutions. After being stored 4996

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 19, 2004

in the darkness for a certain period (generally 24 h) at room temperature, the resulting solutions were directly introduced to the MIMS system to measure the concentrations of chlorine residuals and CNXs. Samples were also withdrawn, immediately dechlorinated with concentrated sodium thiosulfate, and extracted for THM and HAA analyses. To study the effects of glycine concentrations on DBP formation in chlorination and chloramination, similar chlorination experiments were conducted but using synthetic solutions containing humic acid, glycine (from 0.067 to 2 mg/L as N), and ammonia nitrogen at either 0 or 2 mg/L. Chloramination was also conducted by adding preformed monochloramine (8 mg/L as Cl2) for comparison.

Results and Discussion Substances including ammonia, bromide ions, N-organic compounds, and humic acid in solution react with free chlorine to form various halogens and DBPs. The reaction between free chlorine and bromide ions produces hypobromous acid. The incorporation of ammonia leads to the formation of chloramines, bromamines, and bromochloramine (20). All these halogens may react with humic acid or other precursors to form DBPs. N-organic compounds compete with other substances for free chlorine to form organic chloramines and other byproducts. Therefore, the preferable pathways and products are expected to be dependent on the relative rates of these respective reactions. Chlorination of HA, HA + Gly, and HA + Gly + NH3 Solutions. Figure 1 shows the results of DBP formation and concentrations of chlorine residuals 24 h after chlorination of the HA, HA + Gly, and HA + Gly + NH3 solutions with various concentrations of bromide ions. Humic acid has been known to serve as the precursor to yield THMs and HAAs. However, CNXs were not readily detected after 24 h of chlorination in the presence of humic acid. Although humic acid contains a small percentage of nitrogen sources, which could participate in the chlorination process (7), the interaction did not lead to a significant CNX yield in the current study due to lower formation and/or higher decomposition rates. CNX formation was found to be dependent on the type and structure of N-organic compounds; only some organic nitrogen compounds including glycine, serine, and threonine were reported to yield significant quantities of CNXs (9, 21). CNCl was reported to undergo oxidative decomposition in the presence of free chlorine (22). When using the HA + Gly solution, the highly reactive glycine drew chlorine toward the reaction of glycine chlorination (kglycine-chlorine ) 5.0 × 107 M-1 s-1; 23), therefore reducing the amount of available free chlorine participating in the reactions of oxidation and chlorination of humic acid. As a result, the HAA and THM yields (on both mass and molar bases) were reduced. The small reduction of HAA and THM yields was due to the limited quantities of added glycine (only 0.1 mg/L as N). The effect of glycine at various concentrations is discussed in later sections. Small yields of CNXs (in a few micrograms per liter), compared to the data in Shang et al. (9), were observed since the fast reaction between free chlorine and glycine could lead to large formation of CNXs. The small yields were attributable to the high initial chlorine to nitrogen mass ratio (80:1) used in this study, although some of the initial chlorine was known to be consumed by bromide and HA oxidation. Low CNX yields (less than 24 µg/L) were observed even at high bromide ion concentrations (up to 2 mg/L) after 24 h. The results indicate that both chlorine and bromine can catalyze CNX hydrolysis. Addition of 2 mg/L of ammonia nitrogen in the water matrix further decreased the yields of THMs and HAAs by eliminating the yields of the tri-species, accompanied by markedly suppressing the di-HAA formation. The results

FIGURE 1. DBP formation and residual chlorine as functions of the bromide ion concentration after 24-h chlorination of the three synthetic solutions: HA solution (left bar), HA + Gly solution (middle bar), and HA + Gly + NH3 solution (right bar). ND indicates not detected, and free chlorine was fully depleted. can be explained by the chlorine chemistry involved. Although free chlorine reacted more slowly with ammonia (kammonia-chlorine ) 2.9 × 106 M-1 s-1; 23) than with glycine, after free chlorine satisfied its demand from reacting with glycine, the excessive free chlorine reacted with ammonia nitrogen to form relatively stable, persistent inorganic chloramine residuals in higher quantities as shown in Figure 1. The results of THM and HAA formation also agree with the data reported in the literature (24, 25) that trihalogenated HAAs (TXAAs) occur only if excessive free chlorine (in this study free bromine also) is available. On the other hand, compared with the chlorination of the HA and HA + Gly solutions, substantial increases in the yields of CNXs (on both mass and molar bases) were observed in chlorination of the HA + Gly + NH3 solution after 24 h. This is attributable to a few reasons: (i) increases in CNX formation from the reaction between monochloramine and humic acid, which is known to contribute to the formation of CNCl (26); (ii) decreases in CNX oxidative decomposition due to the transformation of oxidant residuals from free chlorine (strong oxidant) to monochloramine (weak oxidant); and (iii) increases in CNX formation from the formaldehyde monochloramination pathway (27). Effect of Bromide Concentration. Figure 1 also illustrates the differences in species and concentrations of DBPs and chlorine residuals when the three water matrixes were spiked with various concentrations of bromide ions. In general, increasing the bromide concentration significantly increased the mass concentrations of total THMs, total HAAs, and total CNXs. Increasing the bromide concentration also markedly shifted the DBPs toward bromine-containing species and increased the formation of higher bromine-incorporated species including CHBr2Cl, CHBr3, dibromoacetic acid (DBAA), dibromochloroacetic acid (DBCAA), and tribromoacetic acid (TBAA). The formation of lower bromine-incorporated species including CHBrCl2, bromochloroacetic acid (BCAA), and bromodichloroacetic acid (BDCAA) increased with increasing

bromide ion concentrations up to 1 mg/L then decreased at 2 mg/L. Bromine-containing species generally create greater public health concerns (14). In addition, the patterns of THM and HAA speciation were qualitatively similar to the results reported previously in chlorinated water samples in the presence of bromide ions (3, 25). In the absence of ammonia nitrogen, increasing the bromide concentration also decreased the percentage of di-HAAs in total HAAs due to the formation of hypobromous acid from bromide oxidation. Hypobromous acid was reported to be more active in substitution reactions with precursors than was hypochlorous acid (28). On the other hand, in the presence of 2 mg/L ammonia, ammonia reacted with free chlorine to form monochloramine at a far faster rate than the oxidation of bromide at 2 mg/L or less (kbromide-chlorine ) 3.8 × 103 M-1 s-1; 28). Therefore, the formation of hypobromous acid in large quantities was unlikely to occur, which consequently led to lower yields of the tri-HAAs and THMs. Inorganic bromochloramine can be formed from slow monochloramination of bromide ions (kbromide-monochloramine ) 3.5 × 106 M-2 s-1; 29), and it is also known to react with organic precursors to form brominated DBPs (13). Therefore, some increases (only on a mass basis) in the formation of HAAs and THMs and bromine incorporation with increasing bromide concentrations were also observed using the HA + Gly + NH3 solutions. On a molar basis, only the concentration of TXAAs consistently increased with increasing the bromide ion concentration in all solutions. The molar concentrations of total THMs and total HAAs, in general, gradually increased with increasing bromide concentrations; except that the molar concentration of total HAAs in the HA + Gly + NH3 solution fluctuated in the range of 0.15-0.21 µM. On the other hand, in the presence of HA and glycine, a more drastic enhancement in CNX yields, due to large increases in CNBr formation, was obtained with increasing bromide ion (on a mass basis) and/or ammonia concentrations (on both mass and molar bases). The molar concentration of total CNXs, however, VOL. 38, NO. 19, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4997

FIGURE 2. DBP formation and residual chlorine from 1-min (left bar) and 24-h (right bar) chlorination of HA, HA + Gly, and HA + Gly + NH3 solutions in the absence of bromide ions. displayed a small declining trend followed by a relatively larger increase with increasing the bromide ion concentration. CNBr has been reported to be formed from the reaction between organic matter and bromochloramine or the reaction between organic matter and a trace amount of hypobromous acid formed in the beginning and at equilibrium (13). Another two reasons involving glycine nitrogen are provided as possible explanations. First, based on the pathway proposed by Gerritsen et al. (30), the reaction of free chlorine with glycine led to the formation of HCN, which was an intermediate compound for the formation of CNCl after further chlorination. Similarly, HCN can also react with HOBr to form CNBr (30). Second, the decomposition of CNCl and CNBr was less significant with the existence of combined residuals. Much higher CNX yields were observed than in the absence of ammonia. Effect of Contact Time. The effect of contact time on DBP formation was studied by measuring samples taken after 1 min and after 24 h. Figure 2 presents the resulting DBP and residual chlorine concentrations in three water matrixes in the absence of bromide ions. In all three cases, the yields of THM (chloroform) and HAAs (DCAA and TCAA) in the first minute were much less than those generated after 24 h. The formation of CNCl did not follow the same trend as those of THMs and HAAs. CNCl was not formed using the HA solution. When the HA + Gly solution was used, a higher CNCl concentration was observed at 1 min than after 24 h due to the oxidative decomposition of CNCl in the presence of free chlorine (22). In using the HA + Gly + NH3 solution, no CNCl was formed after 1 min but its concentration increased to around 20 µg/L after 24 h. It is presumed that CNCl formation from glycine monochloramination is a slow process and that CNCl oxidative decomposition did not proceed in the presence of monochloramine. These results are consistent with the residual chlorine measurement that the quantities of chlorine residuals were reduced with time. In the absence of ammonia, free chlorine at approximately 7 mg/L as Cl2 was observed at 1 min, but its quantity was reduced down to an undetectable level after 24 h. In the presence of 4998

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 19, 2004

ammonia, inorganic monochloramine was the predominant residual throughout the course, and a smaller decline in concentration with time was observed. Effect of Glycine Concentration. To further illustrate the effect of glycine concentrations on the formation of DBPs under various chlorine disinfection schemes, tests were conducted with solutions containing a constant HA concentration (5 mg/L as DOC), various glycine nitrogen concentrations (from 0.067 to 2 mg/L), two ammonia nitrogen concentrations (0 and 2 mg/L), and different chlorine applying techniques in the absence of bromide ions. Figure 3 displays the formation of THMs, HAAs, CNXs, and concentrations of chlorine residuals 24 h after dosing 8 mg/L (as Cl2) hypochlorous acid to the solutions containing 0 or 2 mg/L ammonia nitrogen (referred to as “without NH3” and “with NH3”, respectively). It is evident that chloroform, DCAA and TCAA formation decreased with the increase of glycine concentrations in both cases. The presence of ammonia greatly reduced their formation and reduced the effect of glycine concentration on their formation. Again, these findings can be explained by the competing reactions involving free chlorine, humic acid, glycine, and ammonia. Higher concentrations of glycine tended to draw higher quantities of free chlorine to the chlorination reaction to form organic chloramines due to glycine’s fastest reaction rate with chlorine as compared to other pairs. Therefore, in the absence of ammonia, glycine’s presence highly reduced the amount of free chlorine available to yield THMs and HAAs. In the presence of 2 mg/L ammonia nitrogen, increasing glycine concentrations shortened the transient free chlorine effect since the additional quantities of glycine, which reacts faster than ammonia with chlorine, fastened the free chlorine disappearance. Meanwhile, glycine competed with ammonia on the free chlorine; hence, the concentrations of monochloramine formed were low. Therefore, DBP production was greatly reduced. On the other hand, larger variations in both the quantity and the pattern of CNCl formation were obtained. The CNCl yields were found to be highly variable and dependent on the initial free chlorine to

FIGURE 3. DBP formation and residual chlorine as functions of the glycine-N concentration after 24-h chlorination of humic acid solutions without ammonium ions (left bar) and with ammonium ions (2 mg/L as N) (right bar) in the absence of bromide ions. ND indicates not detected, and free chlorine was fully depleted. glycine-N ratios and the presence/absence of ammonia nitrogen. The largest formation (506 µg/L) occurred at a free chlorine to glycine-N mass ratio of 8:1 in the absence of ammonia. Reports in the literature agree that monochloramination or chlorination in the presence of ammonia yielded higher CNCl than chlorination in the absence of ammonia (11). However, the results in the current study showed that, at some glycine concentrations (e.g., 1 and 2 mg/L as N), chlorination in the absence of ammonia produced more CNCl than in the presence of ammonia. Therefore, the tendency of CNCl formation cannot be simply generalized as in the literature and further clarification is required. In addition, it is worth mentioning that, in the absence of ammonia, significant quantities of inorganic monochloramine and dichloramine were detected by MIMS at high glycine concentrations of 1 and 2 mg/L as N. The generation of inorganic chloramines due to the cleavage of C-N bonds in glycine molecules after chlorination, decarboxylation, and hydrolysis was also reported by Shang et al. (9) and Alouini and Seux (31). When monochloramination was considered as an alternative to chlorination to reduce DBP formation, its application methods affected the formation of DBPs and the quantities of chloramine residuals (3). The effects of glycine nitrogen concentrations using two different monochloramination application methods were studied. These two application methods were to add free chlorine to ammoniacontaining solutions to form in-situ monochloramine at a free chlorine to ammonia-N mass ratio of 4:1 (referred to as “free chlorine addition”) and to add preformed monochloramine (referred to as “monochloramine addition”). As shown in Figure 4, in both cases, increasing the glycine concentration decreased the yields of chloroform, DCAA, and TCAA, and the concentrations of chloramine residuals but increased

the formation of CNCl. By comparing the two monochloramination application methods, it is clear that the effects of glycine on the DBP formation are more significant when using free chlorine addition, due to the faster reaction rate between free chlorine and glycine nitrogen than that between free chlorine and ammonia nitrogen. Using monochloramine addition led to lower production of the DBPs. On the other hand, using the monochloramine addition approach increased the concentrations of chloramine residuals at low glycine concentrations but decreased the concentrations of chloramine residuals at high glycine concentrations compared with using free chlorine addition. The relatively larger decreases in chloramine residual concentrations with increasing glycine concentrations using monochloramine addition demonstrated the significance of chlorine transfer from inorganic monochloramine to nitrogenous organic compounds proposed by Yoon and Jensen (32). The chlorine transfer resulted in the consumption of inorganic residuals and an increase in CNCl formation. Chlorination of HA Solutions Containing Various NOrganic Compounds. Various organic nitrogen compounds including glycine, glutamic acid, glycylglycine, methylamine, and diethylamine were also used for comparison. Those compounds represent groups of organic N-compounds with different structure and reactivity. Figure 5 showed the formation of THMs, HAAs, and CNXs and concentrations of chlorine residuals 24 h after dosing 8 mg/L (as Cl2) hypochlorous acid to solutions containing 5 mg/L humic acid, 0.5 mg/L (as N) of a model N-organic compound, and 0 or 2 mg/L ammonia nitrogen (referred to as “HA + Org-N” and “HA + Org-N + NH3” solutions, respectively). In all cases, chlorination of the HA + Org-N solutions produced higher quantities of THM, HAA, and CNCl than chlorination of the HA + Org-N + NH3 solutions. There were large VOL. 38, NO. 19, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4999

FIGURE 4. DBP formation and residual chlorine as functions of the glycine-N concentration after 24-h chlorination with ammonium ions (2 mg/L as N) (left bar) and monochloramination (right bar) of humic acid solutions in the absence of bromide ions. Free chlorine was fully depleted.

FIGURE 5. DBP formation and residual chlorine after 24-h chlorination of humic acid and several nitrogenous organic species, including glutamic acid, glycylglycine, methylamine, diethylamine, and glycine without ammonium ions (left bar) and with ammonium ions (2 mg/L as N) (right bar) in the absence of bromide ions. ND indicates not detected, and free chlorine was fully depleted. 5000

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 19, 2004

TABLE 1. Second-Order Rate Constants with Free Chlorine (25 °C) reactant

k (106 M-1 s-1)

ref

ammonium ions bromide ions glycine glycylglycine methylamine diethylamine

2.9 0.0038 50 5.3 190 41.4

23 28 23 23 23 33

variations in the yields of these DBPs, however. In chlorination of HA + Org-N solutions, the yields of THMs and HAAs followed the order of diethylamine > glycylglycine > methylamine ≈ glycine > glutamic acid. The different reactivities of the model compounds with free chlorine may explain the variations. Compounds with higher reactivities react with chlorine faster and, therefore, reduce the formation of THMs and HAAs from chlorination/oxidation of the humic acid. The reaction rate constants (as shown in Table 1) (kglycine-chlorine ) 5.0 × 107 M-1 s-1, kglycylglycine-chlorine ) 5.3 × 106 M-1 s-1, and kmethylamine-chlorine ) 1.9 × 108 M-1 s-1) support the results. However, the rate constant for glutamic acid cannot be found and that for diethylamine (kdiethylamine-chlorine ) 4.14 × 107 M-1 s-1; 33) does not satisfy the proposed explanation. The disagreement is likely attributable to the additional formation of chloroform, DCAA, and TCAA from chlorination of diethylamine. R-Chlorination of diethylamine followed by hydrolysis can produce ethylamine and acetaldehyde through the similar pathway proposed by de Leer et al. (7). Further chlorination and oxidation of ethylamine and acetaldehyde is expected to yield acetonitrile as the intermediate (31) and chloroform, DCAA, and TCAA as some of the final products (8). The formation of CNCl among different nitrogenous organic species showed great variations in chlorinating the HA + Org-N solutions. The variations can be explained by the structural differences among these N-organic compounds. The differences, after chlorination and decomposition of these model compounds, lead to the formation of either hydrogen cyanide (HCN) in the cases of glycine, glycylglycine, and methylamine or organic substituted nitriles (RsCtN) in the cases of diethylamine and glutamic acid. HCN reacts quickly with residual chlorine/monochloramine to yield CNCl but RCN cannot. In addition, the evidence in the yields of CNCl from chlorination of glycylglycine confirmed the fragmentation in the amide linkage at an initial free chlorine to glycylglycine molar ratio of 6:1, in agreement with Hawkins et al. (34) that high chlorine dosage induces the fragmentation of proteins. Chlorination of model N-organic compounds was reported to produce ammonia or monochloramine (7, 9, 31, 35). The released ammonia undergoes further reactions with excess HOCl to form chloramine residuals based on the classical chlorine-ammonia chemistry (9, 34). However, the chloramines may be consumed by participating in the chloramination of aldehydes (27, 31). This gives a possible explanation that, without adding ammonia, inorganic chloramine residuals were only observed after 24 h of chlorination of methylamine and diethylamine solutions.

Acknowledgments This study was supported in part by the Hong Kong Research Grants Council under Grant HKUST6035/01E.

Literature Cited (1) Woo, Y. T.; Lai, D.; McLain, J. L.; Manibusan, M. K.; Dellarco, V. Environ. Health Perspect. 2002, 110 (Suppl. 1), 75-88.

(2) Liang, L.; Singer, P. C. Environ. Sci. Technol. 2003, 37, 29202928. (3) Qi, Y.; Shang, C.; Lo, I. M. C. Water Res. 2004, 38, 2375-2383. (4) Bryant, E. A.; Fulton, G. P.; Budd, G. C. Disinfection Alternatives for Safe Drinking Water; Hazen and Sowyer: New York, 1992. (5) U.S. EPA. National Primary Drinking Water Regulations: Disinfectants and Disinfection By-Products; Notice of Data Availability; Proposed Rule. Fed. Regist. 1998, 62, 59388. (6) Westerhoff, P.; Mash, H. J. Water SRTsAqua 2002, 51 (8), 415448. (7) de Leer, E. W. B.; Erkelens C.; de Galan, L. In Water Chlorination: Chemistry, Environmental Impact and Health Effects; Jolley, R. L., et al., Eds.; Lewis Publishers: Chelsa, MI, 1987; Vol. 6, pp 763-781. (8) Peters, R. J. B.; de Leer, E. W. B.; de Galan, L. Environ. Sci. Technol. 1990, 24, 81-86. (9) Shang, C.; Gong, W.-L.; Blatchley, E. R., III. Environ. Sci. Technol. 2000, 34, 1721-1728. (10) Kononen, D. W. Bull. Environ. Contam. Toxicol. 1988, 41 (3), 371-377. (11) Krasner, S. W.; McGuire, M. J.; Jacangelo, J. G.; Patania, N. L.; Reagan, K. W.; Aieta, E. M. J. Am. Water Works Assoc. 1989, 81 (8), 41-53. (12) U.S. EPA. Fed. Regist. 1996, 61, 24358-24363. (13) Heller-Grossman, L.; Idin, A.; Limoni-Relis, B.; Rebhun, M. Environ. Sci. Technol. 1999, 33, 932-937. (14) Diehl, A. C.; Speitel, G. E., Jr.; Symons, J. M.; Krasner, S. W.; Hwang, C. J.; Barrett, S. E. J. Am. Water Works Assoc. 2000, 92 (6), 76-90. (15) Shang, C.; Blatchley, E. R., III. Environ. Sci. Technol. 1999, 33, 2218-2223. (16) Standard Methods for the Examination of Water and Wastewater, 20th ed.; APHA, AWWA, and WEF: Washington, DC, 1998. (17) Munch, D. J.; Munch, J. W.; Pawlecki, A. M. U.S. EPA Method 552.2 (Revision 1.0); Office of Research and Development, U.S. EPA: Washington, DC, 1995. (18) Munch, D. J.; Hautman, D. P. U.S. EPA Method 551.1 (Revision 1.0); Office of Research and Development, U.S. EPA: Washington, DC, 1995. (19) Krasner, S. W.; Glaze, W. H.; Weinberg, H. S.; Daniel, P. A.; Najm, I. N. J. Am. Water Works Assoc. 1993, 85 (1), 73-81. (20) Gazda, M.; Margerum, D. W. Inorg. Chem. 1994, 33, 118-123. (21) Matsushita, C.; Yamada, Y.; Kuge, Y.; Nakamoto, M.; Maeda, Y.; Munemori, M. 40th National Meeting of the Chemical Society of Japan, Fukuoka, 1979; Abstract No. 3E26. (22) Xie, Y. F.; Reckhow, D. A. Water Res. 1993, 27, 507-511. (23) Margerum, D. W.; Gray, E. T., Jr.; Huffman, R. P. In Organometals and Organometalloids, Occurrence and Fate in the Environment; ACS Symposium Series 82; American Chemical Society: Washington, DC, 1978; pp 278-291 (24) Hwang C. J.; Sclimenti, M. J.; Krasner S. W. In Natural Organic Matter and Disinfection By-products: Characterization and Control in Drinking Water; Barrett, S. E., et al., Eds.; American Chemical Society: Washington, DC, 2000; pp 173-187. (25) Cowman, G. A.; Singer, P. C. Environ. Sci. Technol. 1996, 30, 16-24. (26) Ohya, T., Kanno, S. Chemoshphere 1985, 14, 1717-1722. (27) Pedersen, E. J.; Urbansky, E. T.; Marinas, B. J.; Margerum, D. W. Environ. Sci. Technol. 1999, 33, 4239-4249. (28) Rebhun, M.; Heller-Grossman, L.; Manka, J.; Kimel, D.; Limoni, B. In Water Chlorination: Chemistry, Environmental Impact and Health Effects; Jolley, R. L., et al., Eds; Lewis Publishers: Chelsa, MI, 1987; Vol. 6, pp 665-680. (29) Bousher, A.; Brimblecombe, P.; Midgley, D. Water Res. 1989, 23, 1049-1058. (30) Gerritsen, C. M.; Gazda, M.; Margerum, D. W. Inorg. Chem. 1993, 32, 5739-5748. (31) Alouini, Z.; Seux, R. Water Res. 1987, 21, 335-343. (32) Yoon, J.; Jensen, J. N. Water Environ. Res. 1995, 67, 842-847. (33) Abia, L.; Armesto, X. L.; Canle L. M.; Garcia, M. V.; Santaballa, J. A. Tetrahedron 1998, 54, 521-530. (34) Hawkins, C. L.; Pattison, D. I.; Davies, M. J. Amino Acids 2003, 25, 259-274. (35) Antelo, J. M.; Arce, F.; Parajo, M. J. Phys. Org. Chem. 1996, 9, 447-454.

Received for review March 18, 2004. Revised manuscript received June 30, 2004. Accepted July 12, 2004. ES049580G

VOL. 38, NO. 19, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5001