Amino acids (AAs) have been identified as precursors to disinfection by-products—some more cytotoxic than the currently regulated disinfection by-products. This study measured occurrences and removals of AAs, amino sugars, and bulk organic matter surrogates (dissolved organic carbon [DOC], dissolved organic nitrogen [DON]) in raw water and filter effluents of 16 full-scale water treatment plants influenced by algal activity or wastewater discharge. In raw waters, the average concentration of free AAs—those not bound in larger molecules—was very low, 0.69 µg/L N (< 0.5% of the DON). The average concentration of total AAs—those bound and unbound—was 41.1 µg/L N (15% of the DON). The most dominant (on a molar basis) AA species detected were serine, alanine, and glycine. Water treatment removed, on average, 30, 30, 25, and 65% of DOC, DON, free AAs, and total AAs, respectively. This study was the first large-scale occurrence survey of free and total AAs in United States drinking waters.

Occurrence and removal of amino acids during drinking water treatment BY AARON DOTSON AND PAUL WESTERHOFF

s the building blocks of life, amino acids (AAs) assemble to create proteins, peptides, and other biological macromolecules. Although only 20 species are considered to be “standard amino acids” or ␣-amino acids and are used by cells in biosynthesis, many “nonstandard” amino acids exist and have been shown to play specialized roles in biologically important functions (Voet & Voet, 2004). Proline, although not an ␣-amino acid because of its two amino groups, is still identified as a “standard amino acid.” Amino sugars (ASs) are sugars containing amino groups (e.g., one or more of the OH groups in the sugar is replaced with an amino group). For example, N-acetyl-D-glucosamine is an AS commonly found linked with D-lactic acid as a prominent component of bacterial cell walls. AAs and ASs occur in two forms, free and combined. Combined AAs and ASs are those bound within biological macromolecules (e.g., proteins or peptides); they must be hydrolyzed from their parent structure to be analyzed. Free AAs or ASs are unbound and can be analyzed directly. AAs are known precursors to regulated and unregulated disinfection byproducts (DBPs). Trihalomethanes (THMs) have been shown to form from all AAs, although yields vary depending on the AA species and associated side chain

A

A full report of this project, Occurrence and Formation of Nitrogenous Disinfection By-Products (91250), is free and currently available to Water Research Foundation subscribers by calling 303-347-6121 or logging on to www.waterresearchfoundation.org.

DOTSON & WESTERHOFF | 101:9 • JOURNAL AWWA | PEER-REVIEWED | SEPTEMBER 2009

2009 © American Water Works Association

101

(Hong et al, 2008; Hureiki et al, 1994; Scully et al, 1988; Trehy et al, 1986; Trehy & Bieber, 1981). For example, aromatic side chain–containing tyrosine and tryptophan produce higher THM yields, on a carbon basis, than do other AA species.

N

itrogenous disinfection by-products concern the drinking water industry because they have been found to be more geno- and cytotoxic than many of the currently regulated disinfection by-products. Unregulated nitrogenous DBPs include cyanogen halides (Lee et al, 2006a; Na & Olson, 2006; Na & Olson, 2004; Hirose et al, 1988, 1989), haloacetonitriles (Trehy et al, 1986; Bieber & Trehy, 1983; Trehy & Bieber, 1981), and other classes (Mehrsheikh et al, 2006; Conyers & Scully, 1997; McCormick et al, 1993; Scully et al, 1988). Odorous aldehyde-containing DBPs have also been shown to form from specific AA precursors (e.g., valine, leucine, isoleucine, and phe-

TABLE 1

nylalanine; Hrudey et al, 1988). Nitrogenous DBPs concern the drinking water industry because they have been found to be more geno- and cytotoxic than many of the currently regulated DBPs (Plewa et al, 2008; Muellner et al, 2007). The source, occurrence, and fate of AAs and ASs have been extensively studied from limnological perspectives (Duan & Bianchi, 2007; Kielland et al, 2006; Unger et al, 2005; Shoji et al, 2003; Hagedorn et al, 2000; Volk et al, 1997; Hedges et al, 1994; Thurman, 1985; Lytle & Perdue, 1981; Peake et al, 1972; Peterson et al, 1925). These studies have found that • The percentage of AAs relative to bulk organic surrogates—dissolved organic carbon (DOC), dissolved organic nitrogen (DON), or ultraviolet absorbance at 254 nm (UV254)—vary significantly by location and environmental conditions. • Total AAs—the sum of the free and combined AAs— occur at concentrations between 50 and 1,000 µg/L in rivers, streams, and lakes. Thurman (1985) reported that total AAs accounted for 2.6% of the DOC and 35% of the DON in lakes. Hagedorn and colleagues (2000) found that in catchment runoff the total AAs varied significantly depending on soil type and runoff event, accounting for 20 to > 75% of the DON. • Free AA concentrations can also be related to algal activity (Lytle & Perdue, 1981). Free AAs were found to

Facilities and treatment process components

Carbamazepine ng/L

Chemicals WTP Number

State

Treatment

Pretreatment

Prefilter

Filtration

2006

2007

Algal Counts counts/mL 2006

2007

1

Mich.

S, F/S

P

O3, P

GAC

9.5

NA

2

Calif.

F/S

P, Cl2

O3

DM

NA

352

3

Calif.

F/S

P, Cl2

4

N.J.

HRSBC

P

O3, Cl2

GAC

27

NA

5

Pa.

F/S

KMnO4, Cl2

Cl2

GM

33

5,300

6

Pa.

F/S

Cl2

Cl2

GM

18

19

22,700

6,100

7

Pa.

F/S

Cl2

Cl2

GM

16

3

300

2,000

DM

NA

NA

25

222

8

Colo.

F/S

P, Cl2

DM

1.3

NA

9

Colo.

F/S

P, Cl2

UF

13

NA NA

10

Okla.

F/S

Cl2

GAC

BDL

11

Okla.

F/S

ClO2

ClO2

GM

BDL

NA

12

US*

S, F/S

O3, P

NH2Cl

GM

4.3

NA

O3, P

NH2Cl, P

GM

5.9

NA

GM

NA

NA

GM

NA

NA

GM

3.8

NA

13

US*

S, F/S

14

Va.

F/S

15

Va.

F/S

Cl2

16

Nev.

F/S

Cl2, NH3

O3, Cl2

BDL—below detection limit, Cl2—chlorine, ClO3—chlorine dioxide, DM—dual media, F/S—flocculation/sedimentation, GAC—granular activated carbon, GM—granular media, HRSBC—high-rate sand-blasted clarification, KMnO4—potassium permanganate, NA—not analyzed, not sampled, NH2Cl—chloramines, NH3— ammonia, O3—ozone, P—polymer, S—chemical softening, UF—ultrafiltration, WTP—water treatment plant *State that requested anonymity

102

SEPTEMBER 2009 | JOURNAL AWWA • 101:9 | PEER-REVIEWED | DOTSON & WESTERHOFF

2009 © American Water Works Association

5th/95th percentile Average

5

90th percentile 75th percentile 50th percentile (median) 25th percentile 0.6 10th percentile 0.5

4

0.4

3

0.3

2

0.2

1

0.1

0

0.0

6

DON—mg/L N

FIGURE 1 Occurrence of DOC, DON, free and total AAs, and free and total ASs in WTP influent and filter effluent

100 90 80 70 60 50 40 30 20 10 0

8,000

3.5

600

7,000 6,000 5,000 4,000 3,000 2,000

Total AAs—nmol/L

DOC—mg/L C

1,000 0

3.0

500

2.5

400

2.0 300

1.5 200

1.0

. Total ASs—nmol/L

Growing urbanization has increased centralized wastewater production and the discharge of inorganic and organic nitrogen into drinking water supplies, in turn increasing DBP precursors and nutrient loading, which can lead to algal blooms (Lee et al, 2006b; Pehlivanoglu-Mantas & Sedlak, 2006; Westerhoff & Mash, 2002). Population growth has increased the demand for drinking water, which has forced utilities to begin using source waters that were once considered too impaired because of upstream wastewater discharge or algal growth. Organic nitrogen, which includes materials containing AAs, comes from biological sources either directly or indirectly associated with wastewater or agricultural discharges, or algal activity. Whereas the use of impaired water sources containing elevated concentrations of organic nitrogen has increased, only a few studies have examined AAs in drinking water, and some of those have focused only on free AAs because of the difficulty in quantifying total AAs (Dotson et al, 2008; Chinn & Barrett, 2000; Prevost et al, 1998; Hureiki & Prevost, 1996; Le Cloirec & Martin, 1985). These studies have shown that: • Coagulation (flocculation and sedimentation) removes 34 to 75% of total AAs in California and Canadian drinking waters (Chinn & Barrett, 2000; Hureiki et al, 1996) but does not affect the concentration of free AAs (Le Cloirec & Martin, 1985). • Prechlorination and preozonation each increased the free AA concentration because of the partial hydrolysis of proteins and peptides forming aldehydes and nitriles (Le Cloirec & Martin, 1985). • Biological filtration may increase or reduce the total AA concentration, depending on the water temperature, filter-ripening procedure, and biological activity (Prevost et al, 1998; Hureiki & Prevost, 1996; Le Cloirec & Martin, 1985). In these studies, removal of AAs during drinking water treatment is widely varied and dependent on the water source and particular treatment processes used. These findings clearly identify the need for a cohesive study about AAs that examines a broad array of treatment processes and water sources. This study assessed the occurrence and removal of free AAs, total AAs, and bulk organic matter surrogates at drinking water treatment plants (WTPs) that treat water impaired by upstream wastewater discharge or algal growth. Jar tests were also conducted on selected algal-impaired source waters. Statistical analysis compared the abundances of DOC, DON, AAs, and ASs and evaluated correlations between water quality parameters, and AAs and ASs.

Field sampling. Sixteen participating WTPs, located across the United States, collected samples during a period of algal activity or increased wastewater influence. Three WTPs were sampled twice, once in 2006 and again in 2007, whereas the remaining WTPs were sampled only in 2007. The WTPs were in nine states: California (2), Nevada (1), Colorado (2), Oklahoma (2), Michigan (1), Pennsylvania (2), New Jersey (1), Virginia (2), and two

Total Free AAs—nmol/L

THE NEED FOR THIS RESEARCH

METHODOLOGY

Total Free ASs—nmol/L .

be roughly one-tenth as abundant as combined AAs in streams and lakes (Thurman, 1985; Peake et al, 1972).

100

0.5 0.0

0 Plant Influent

Filter Effluent

Plant Influent

Filter Effluent

AA—amino acid, AS—amino sugar, DOC—dissolved organic carbon, DON—dissolved organic nitrogen, WTP—water treatment plant Total free AAs = the sum of measured free AAs, total AAs = the sum of measured free and combined AAs, total free ASs = the sum of measured free amino sugars, total amino sugars = the sum of measured free and combined ASs Number of samples = 19 (3 in 2006, 16 in 2007)

DOTSON & WESTERHOFF | 101:9 • JOURNAL AWWA | PEER-REVIEWED | SEPTEMBER 2009

2009 © American Water Works Association

103

FIGURE 2 Percentage of free or total AAs accounted for by individual AA species in raw waters 5th/95th percentile Average

Free AAs

90th percentile 75th percentile 50th percentile (median) 25th percentile 10th percentile

30

20

AA Species—%

10

0

Total AAs 20

10

*

in

in hi et M

H

is

on

tid

si Ly

e

e

ne

ne ro

eu ol

si

ne ci

in an yl en

Ph

Is

al

rg A

Ty

e

e in in

lin Va

in ol

e

e

e Pr

uc

in

in

e Le

d

on re

Th

tic ar sp A

G

lu

ta

m

ic

A

ci A

ci

d

e rin Se

ne ni la A

G

ly

ci

ne

0

AA—amino acid, WTP—water treatment plant *Proline was found at significantly greater concentrations at WTPs 4, 5, and 6 than they were at the other sampled WTPs, accounting for 26− 60% of the total AAs. Therefore, these three WTPs are not included in the statistics for the proline box plot. Number of samples = 19 (3 in 2006, 16 in 2007)

WTPs from a state that requested to remain anonymous. Table 1 provides descriptions of the treatment processes used by each of these facilities. Source water samples. Source waters influenced by wastewater discharge were sampled during low natural stream flow. Algal-influenced source waters were sampled when a significant algal bloom was observed in the source water; WTP staff defined “significant” on the basis of their experience with local algal blooms. Many of the facilities had source waters affected by both wastewater and seasonal algal blooms. Raw water, filter effluent samples. Samples were collected from the plant influent (raw water) and filter effluent (treated water prior to disinfection where possible) at all WTPs and between significant treatment processes when possible. WTP operations staff collected grab samples in bottles supplied by the authors; the bottles were returned to the authors’ lab via overnight shipping. The bottles and caps used for utility sampling were prepared by soaking them in 10% (by weight) hydrochloric acid, and then placing them in a 550oC oven for 8 h to combust the remaining organics. Filtering the samples. Immediately upon receipt, the samples (excluding those for AA determination; procedure 104

described in the Analytical methods section) were filtered through 0.45-µm polyethersulfone membrane filters1 that had been rinsed with distilled water2 to remove any easily leached organics, and then the samples were analyzed to determine basic water quality parameters. Laboratory-scale experimentation. Laboratory-scale flocculation/coagulation was performed in 2-L beakers on a programmable jar test apparatus3 following the US Environmental Protection Agency protocol (USEPA, 1999). A 10-g/L stock solution of aluminum sulfate coagulant was manually added to each jar during high-speed mixing. Flocculation. Flocculation was performed with a tapered mixing scheme: 1 min at 150 rpm, and then 5 min each at 75, 50, and 25 rpm. The flocculated water was allowed to settle for 45 min prior to filtration of the decanted supernatant through a preashed (any present organic matter burned into ash at 500oC) 0.7-µm glass fiber filter.4 Filtration. This filtering process simulates a typical filtration process used during full-scale water treatment. The filtered samples were then stored in a refrigerator at 4oC until sample processing and analysis. Analytical methods. Free and total AA concentrations were determined for each sample by high-perfor-

SEPTEMBER 2009 | JOURNAL AWWA • 101:9 | PEER-REVIEWED | DOTSON & WESTERHOFF

2009 © American Water Works Association

DOTSON & WESTERHOFF | 101:9 • JOURNAL AWWA | PEER-REVIEWED | SEPTEMBER 2009

2009 © American Water Works Association

DON—mg/L N

DON—mg/L N

TAA—nmol/L

DOC—mg/L C

TAA—nmol/L

DOC—mg/L C

FAA—nmol/L

FAA—nmol/L

to undergo hydrolysis more rapidly than the measured mance liquid chromatography (HPLC) analysis after AAs, supporting the use of shorter hydrolysis times (e.g., preconcentration. 90 min). In total, 16 AAs and two ASs (glucosamine and Preconcentration. Each 500-mL sample was concengalatosamine) were determined using the specified method trated by vacuum-assisted rotary evaporation5 (70-mbar (Waters Corp., 1993). vacuum at 60oC) and then lyophilized.6 Samples for High-performance liquid chromatography. Separation AA analysis were not filtered through a 0.45-µm filter of the derivatized AAs was performed by an HPLC separabecause parent compounds (proteins/peptides) may be tions module8 equipped with a 3.9 × 150-mm column removed, whereas during water treatment these compounds would be present to interact with applied chemimounted in a 37oC column heater. Quantification was cal oxidants. After recording the weight of the recovered performed by a fluorescence detector9 configured at an lyophilized material, it was placed in a dessicator in the excitation of 250 nm and an emission of 395 nm. Sixteen dark until analysis. amino acids and two amino sugars were separated in 34 Hydrolysis. Free AAs and free ASs were derivatized min using a predetermined three-eluent gradient: (1) a by dissolving 5–10 mg of lyophilized material in borate commercial eluent (2) acetonitrile, and (3) water. Chrobuffer and adding a reagent.7 To determine total AAs matographs for AA standards and a water sample are provided elsewhere (Dotson, 2008). and total ASs, a measured mass of lyophilized material Blank samples were analyzed by following the iden(15–25 mg) was hydrolyzed by liquid-phase 4 N methtified total and free AA procedure on laboratory-generanesulfonic acid (MSA) hydrolysis (90 min at 135oC) ated distilled water every 5 to 10 samples, and comusing a method adapted from the literature (Martens & monly resulted in no measurable fluorescent signal. In Loeffelmann, 2003) and optimized to maximize total a few instances where a blank produced a fluorescent AA recovery (Dotson, 2008; Mitch et al, 2008). The signal, the chromatogram reviewed and determined that hydrolysate (i.e., the solution containing AAs hydrothe resulting signal did not align with known AAs. For lyzed from combined form to free form) was derivatized these cases, analysis was stopped, and the needle and by adding a reagent7 according to the manufacturer’s column were thoroughly rinsed with eluent until no specifications (Waters, 1993). The derivatized solutions signal was observed. of both free and total AAs were centrifuged, and the Dissolved organic carbon and total dissolved nitrosupernatants were transferred to HPLC vials. gen. DOC and total dissolved nitrogen (TDN) were A detailed evaluation of this hydrolysis method revealed accurate and consistent hydrolysis of combined AAs in dissolved organic matter (DOM) and FIGURE 3 Observed correlations between algal counts or wastewater tracer model polypeptides (Dotson, (carbamazepine) and FAA, TAA, DOC, or DON 2008). For example, for lysozyme (a single-chain polypeptide of 129 A B AAs) an average of 95 ± 13% of 0.6 7 140 12,000 FAA DOC TAA the AAs present in the polypeptide DON 120 6 0.5 10,000 was recovered; individual AA 5 100 0.4 8,000 recoveries varied from 74 to 80 4 0.3 6,000 122%. Inorganic salt (sodium 60 3 4,000 0.2 chloride) did not interfere with the 40 2 0.1 2,000 determination of total AAs. 1 20 0 0 Hydrolysis of a DOM isolate, 0 0.0 1 10 1,000 100,000 1 10 1,000 100,000 prepared by dissolving a lyophilized Algal Counts—counts/mL Algal Counts—counts/mL DOM isolate in distilled water, C D 0.6 7 12,000 140 FAA DOC with and without salt (sodium TAA DON 6 120 0.5 10,000 chloride), produced total AA results 5 100 0.4 8,000 within 3% of each other. Triplicate 4 80 0.3 analysis of a lyophilized whole6,000 3 60 water sample, similar to those used 0.2 4,000 2 40 for this work, resulted in an accept0.1 2,000 1 20 able sample variability of ± 12% at 0.0 0 0 0 0.0 10.0 20.0 30.0 0.0 10.0 20.0 30.0 a 95% confidence interval. Carbamazepine—ng/L Carbamazepine—ng/L Although no direct recovery studies were performed with a commerFAA—free amino acids, nmol/L—nanomoles per litre, DOC—dissolved organic carbon, DON—dissolved organic nitrogen, TAA—total amino acids cially available polypeptide containing an AS, free ASs were observed 105

Westerhoff, 2005). Specific UV absorbance (SUVA) was calculated by dividing 100 times the UV absorbance (m-1) measured after 0.45-µm filtration at UV254 by DOC (mg/L C). Wastewater indicators (i.e., caffeine, carbamazepine, and primidone) were measured by liquid chromatography/tandem mass spectrometry in positive ionization electrospray mode after concentration by solid-phase extraction using hydrophilic–lipophilic balanced cartridges (Vanderford et al, 2003). As a result of the correlation observed among the measured wastewater tracers, Table 1 shows only carbamazepine; the reader is directed to the primary study for further information (Mitch et al, 2008). Participating WTPs provided algal types and counts if their laboratories performed such analyses (Table 1).

FIGURE 4 Removal of organics (DOC and AAs) during water treatment

2006 samples 2007 samples

100

Removal of Total AAs—%

80

1:1 Line 60

40

RESULTS AND DISCUSSION 20

0 0

20

40

60

80

100

DOC Removal—% AAs—amino acids, DOC—dissolved organic carbon

determined simultaneously by high-temperature combustion10 (720oC) followed by nondispersive infrared detection and chemiluminescence detection,11 respectively. Nitrite and nitrate were measured by ion chromatography12 with conductivity detection outlined in method 4500-NO 3 - C (Standard Methods, 2001).

P

Occurrence of DON, AAs, and ASs in untreated surface water supplies. Figure 1 shows the statistical distributions of DOC, DON, free and total AAs, and free and total ASs from the survey. The average total AA concentration in the raw water was 41 µg/L N (2.3 µmol total AA/L). Total AAs. Total AA concentration varied more than three orders of magnitude, ranging from a minimum of 0.80 µg/L N to a maximum of 171 µg/L N. Although these raw waters were not filtered through a 0.45-µm filter before analysis, most of the organic carbon was dissolved; 95 ± 11% (95% confidence interval) of the total organic carbon (TOC) was DOC. On the basis of a study of Mississippi River water, the colloidal matter contained two times more organic carbon and organic nitrogen than the particulate matter, on a percentage basis (Rostad et al, 1997). More than likely, the majority of the total AAs and ASs were suspected to be associated with organic nitrogen–enriched biogenic organic colloids. Biogenic organic colloids have been shown to be enriched in organic nitrogen (C/N < 10), and about a third, on average, of this organic nitrogen was accounted for by total AAs (Dotson, 2008). Free amino acids and amino sugars. Free AAs accounted for only a small percentage of the total AAs, averaging 5.9% on a molar basis (minimum 0.23%, maximum 38.2%). The average free AA concentration was 0.69 µg/L N (37.7 nmol/L free AA). ASs were found in lower concentrations than AAs. Free galactosamine was found at one facility, WTP 5, at a concentration of 0.01 µg/L N. Free glucosamine had an average concentration of 0.018 µg/L N and was found in 53% of the samples collected. Upon hydrolysis, the average total galactosamine was determined to be 0.82 µg/L N, whereas the average total glucosamine was higher, 1.75 µg/L N.

opulation growth has increased the demand for drinking water, which has forced utilities to begin using source waters that once were considered too impaired because of wastewater discharge or algal growth.

Ammonia was determined colorimetrically according to method 10205 (Hach Corp., 2007) while spiking in 50 µg/L NH3 as N to reduce the minimum detection limit to 4 µg/L NH3 as N. Dissolved organic nitrogen. DON was calculated by subtracting dissolved inorganic nitrogen (DIN = nitrite + nitrate + ammonia) from TDN. Samples determined to have a DIN-to-TDN ratio of > 0.6 mg N/mg N were dialyzed to more accurately measure DON (Lee & 106

SEPTEMBER 2009 | JOURNAL AWWA • 101:9 | PEER-REVIEWED | DOTSON & WESTERHOFF

2009 © American Water Works Association

half of the samples required dialysis because of elevated concentrations of DIN; dialysis was carried out until a DIN-to-TDN ratio of < 0.6 was acquired. Average DIN concentration was 687 µg/L N and varied from a minimum of 16 to a maximum of 3,072 µg/L N. Furthermore, higher DINto-TDN ratios are indicative of nitrogenimpaired source waters. Source waters impaired by upstream algal blooms or wastewater discharges were clearly nitrogen-enriched (average DON = 290 µg/L N and DOC-to-DON = 12.6) compared with the previously identified US average (average DON = 186 µg/L N and DOCto-DON = 18; Lee et al, 2006b). On average, total AAs accounted for 15% (minimum 0.17%, maximum 47%) of the measured DON concentration and 3.5% (minimum 0.19%, maximum 11.6%) of the DOC. Free AAs accounted for no more than 1.3% of the DON (average 0.28%) and 0.21% of the

Dissolved organic nitrogen. The average raw water DON and DOC were 290 µg/L N (minimum 68, maximum 486 µg/L N) and 3.33 mg/L C (minimum 1.4, max-

A

pplication of chlorine or chlorine dioxide decreased the concentration of total amino acids regardless of the application point but may have promoted the formation of disinfection by-products, some of which are regulated. imum 6.6 mg/L C), respectively, resulting in an average DOC-to-DON ratio of 12.6 mg DOC/mg DON. DIN-to-TDN ratios varied from 0.04 to 0.91, with an average ratio of 0.53. To accurately measure DON,

FIGURE 5 Percentage of free or total AAs accounted for by individual AA species in treated waters 5th/95th percentile Average

30

90th percentile 75th percentile 50th percentile (median) 25th percentile 10th percentile

Free AAs

20

AAs Species—%

10

0

Total AA

* 30

20

10

A

is t H

e in et hi on

M

id in e

ne si Ly

Ty

ro

si

ne

ci ne

in e

le u

an yl al en

Ph

Is o

e in in A rg

lin e

e ol in Pr

in uc

Va

e

e

Th

Le

on re

A

ci

ar tic

A ic

sp

m ta G lu

in

ci d

d

e rin Se

ni ne A la

G

ly

ci

ne

0

AA—amino acid, WTP—water treatment plant *Isoleucine accounted for 96% of total amino acids at WTP 5. Number of samples = 19 (3 in 2006, 16 in 2007)

DOTSON & WESTERHOFF | 101:9 • JOURNAL AWWA | PEER-REVIEWED | SEPTEMBER 2009

2009 © American Water Works Association

107

authors’ method was half as sensitive as it was for other amino acids. Tryptophan has shown that it may account for a notable portion of the AA pool, as indicated by previous fluorescence research (Chen 500 Total AAs with DOC et al, 2003). However, the method characterized DON DON used in this article does not allow 18 Uncharacterized DON Total AAs (N-basis) 400 for quantification by fluorescence UV254 because tryptophan does not form 8 300 a fluorescent derivative. Regardless 7 of the occurrence of these unmea8 200 sured AA species, standard amino 1.0 acids account for less than a third 100 of the DON present in untreated impaired drinking water sources. 0.8 0 Figure 2 illustrates the statistical A B C D distribution of the percent of the measured amino acids accounted 0.6 for by the individual AA species studied. On average, in raw waters glycine (13.5%), alanine (10.2%), 0.4 and serine (8.9%) accounted for the greatest percentage of the measured total amino acids (on a molar basis). 0.2 However, this order did not hold true for the measured free amino acids, in which, on average, serine 0.0 (17.3%), glycine (14.5%), and gluB—Settled after A—Plant influent C—After ozonation D—Filter effluent coagulation tamic acid (8.0%) accounted for the greatest percentages. Regardless of Sample Point the incorporation of an individual AA—amino acid, C/C0—concentration data normalized to plant influent concentration C0 amino acid species into a protein/ (plant influent divided by plant influent), DOC—dissolved organic carbon, DON—dissolved organic nitrogen, UV254—ultraviolet absorbance at 254 nm, WTP—water treatment plant peptide, the five most prominent amino acid species accounted for No chemical oxidants that would oxidize free AAs or total AAs into organic by-products were applied to plant influent before coagulation. nearly half of the measured pool of amino acids (on average 52.3% of free AAs and 45.8% of total AAs). Algae–amino acid correlations. Many of the source DOC (average 0.07%). A few of the source waters sampled waters were evaluated for algal activity (i.e., algal counts). during this study had very low total AAs (< 2% of the DON A correlation was expected on the basis of previous findor < 1% of the DOC). ings that AA concentrations varied with the status of algal Measured individual amino acid species. Only 16 blooms (i.e., lag phase, exponential growth phase, stationof the 20 standard amino acids were directly measured. However, during acid hydrolysis, asparagine and glutamine are converted to aspartic acid and glutamic acid, respectively. Because of this conversion, measured aspartic acid and glutamic acid are the sum of asparagine and aspartic acid and the sum of glutamine and glutamic acid, and are often abbreviated ASX or GLX, respectively. Thus only cystine and tryptophan were not accounted for at all during this study because they are either destroyed or converted to nonmeasurable species during acid ary phase, or decay; Lytle & Perdue, 1981). However, a hydrolysis. Cystine was not expected at measurable conreview of the samples revealed no correlation between centrations because, compared with other standard amino algal counts and AAs (Figure 3, part A), nor was DOC or acids, it is quite rare, and detection of this species by the Normalized Concentration—C/C0

DON—µg N/L

FIGURE 6 Removal of AAs, DOC, DON, and UV254 at WTP 1, using coagulation and intermediate ozonation. Inset: Numbers above bars indicate percentage of DON characterized by nitrogen present in measured total AAs

C

oagulation removed the largest mass of total amino acids compared with other unit processes.

108

SEPTEMBER 2009 | JOURNAL AWWA • 101:9 | PEER-REVIEWED | DOTSON & WESTERHOFF

2009 © American Water Works Association

89.4

91.6

92.8

93.4

93.5

91.0

97.1

97.6

97.9

97.7

97.2

C/C0

mg/L DON

97.1

C/C0

mg/L DON

Removal of amino acids by drinking water treatment. DON correlated with algal counts (Figure 3, part B). During drinking water treatment, total AAs were prefHowever, site-specific trends, similar to those previously erentially removed, on a percentage basis, over bulk observed by Lytle and Purdue (1981), may exist. For organic matter surrogates. example, at WTP 6, total AAs accounted for a high perBefore final disinfection. Between plant influent centage of the DON concentration (41%) when high algal and filter effluent (typically collected prior to final counts were observed (22,700 counts/mL) and a low perdisinfection), the average physical or chemical removal centage of the DON concentration (1.0%) when algal of DOC, DON, free AAs (N-basis), and total AAs counts were lower (6,100 counts/mL). Results were similar (N-basis) was 30% (1.12 mg/L C), 30% (84.4 µg/L at WTP 3: Total AAs were 1.2% of DON when algal N), 25.4% (0.30 µg/L N), and 65.2% (33.8 µg/L N), counts were low (25 counts/mL) and 15% of DON when respectively. No statistically significant correlation was algal counts were high (222 counts/mL). observed between DOC removal and percent DON These site-by-site relationships suggest that environremoval (R2 < 0.2) or percent removal of free AAs and mental factors such as type of algae, growth phase, and sampling location may confound this analysis of a large percent DOC and DON removal (R2 < 0.2). As shown group of samples from a variety of geographical locations, in Figure 4, the percentage of total AAs removed durwatersheds, and impairments. Lack of a general correlaing treatment exceeded the percentages of DOC and tion also may arise because each lab may have used slightly DON (not shown) removed. After treatment, total different analytical methods, which may skew algae counts to their sitespecific predominant algae species. FIGURE 7 Removal of DON, DOC, total AAs, and UV254 in an Arizona (A) Carbamazepine. The wastewater and an Oklahoma (B) reservoirs by alum coagulation tracer carbamazepine was measured Total AAs because wastewater influence was Uncharacterized DON hypothesized to follow a trend simiNormalized DOC Normalized DON lar to that for algal blooms, namely Normalized UV254 that increased indicators would Normalized TAA A result in total AAs accounting for a 1.0 400 greater percentage of DON. 350 This correlation was expected 0.8 because soluble microbial products 300 (SMPs), which contain AAs (as free 250 0.6 AAs, protein fragments, and pep200 tides), are present in wastewater 0.4 150 discharges (Ramesh et al, 2006), 100 and thus higher wastewater indica0.2 tors would represent a higher con50 centration of SMPs. However, no 0 0.0 0 1 2 4 7 14 relationship was observed between B wastewater indicators and bulk 300 1.0 organic matter surrogates or AAs 250 (representatively shown in Fig0.8 ure 3, parts C and D by carbam200 azepine). Algal blooms in upstream 0.6 150 water may have confounded this 0.4 relationship, because they also 100 produce elevated levels of organic 0.2 nitrogen, including proteinaceous 50 materials; the majority of facilities 0 0.0 influenced by wastewater discharge 0 2 3 6 12 25 were also affected by algal blooms. Alum Dose—mg Alum/mg DOC It may also be hypothesized that AA—amino acid, C/C0—concentration data normalized to plant influent concentration C0 correlations are lacking because (plant influent divided by plant influent), DOC—dissolved organic carbon, DON—dissolved AAs and ASs are likely to be more organic nitrogen, UV254—ultraviolet absorbance at 254 nm biologically available and thus less Numbers in figure indicate percent DON not accounted for (uncharacterized) by total AAs. persistent in the environment than carbamazepine. DOTSON & WESTERHOFF | 101:9 • JOURNAL AWWA | PEER-REVIEWED | SEPTEMBER 2009

2009 © American Water Works Association

109

(null hypothesis rejected, pDOC or DON, two-tailed = 0.002). Free AAs accounted for statistically similar percentages of bulk organic matter surrogates before and after treatment (null hypothesis accepted, pDOC, two400 DOC Total AAs with tailed = 0.11, p DON, two-tailed = 0.10). DON characterized DON Although organic matter surrogates Total AAs (N-basis) Uncharacterized DON 37 were removed during treatment, the UV254 300 ratio of DOC-to-DON was unaf3 fected (null hypothesis accepted, pDON, two-tailed = 0.10). 200 10 Results of measuring AAs. Figure 5 shows that in treated drinking waters, the same three AAs (serine, 100 alanine, and glycine) accounted for 1.0 the largest percentage of the mea0 sured AAs, free or total, on a molar A B C basis. Serine accounted for a larger 0.8 percentage (16.6% on average) of free AAs than total AAs (6.3%). Glycine accounted for the largest 0.6 percentage, on average, of the measured AAs (17.4% of free AAs and 18.6% of the total AAs). Overall, 0.4 eight of the 16 measured AA species accounted for, on average, < 5% of the measured total AA concentration 0.2 (on a molar basis), and nine of the 16 accounted for < 5% of the measured free AA concentration (on a 0.0 molar basis). B—Settled after A—Plant influent C—Filter effluent Amino acid removal by individual coagulation unit processes. On the basis of results Sample Point from WTP 1 and laboratory jar tests, AA—amino acid, C/C0—concentration data normalized to plant influent concentration C0 flocculation/sedimentation (coagula(plant influent divided by plant influent), DOC—dissolved organic carbon, DON—dissolved tion) was the most significant proorganic nitrogen, UV254—ultraviolet absorbance at 254 nm, WTP—water treatment plant cess for removing total AAs. Although all participating facilities used some form of coagulation, emphasis was placed on samples from WTP 1 because no AAs accounted for 4.1% of the DON and 1.0% of the agent that would oxidize free AAs or total AAs into DOC on average. organic by-products was used prior to coagulation. After treatment. Although TOC and DOC were not Coagulation. Figure 6 shows that at WTP 1, 29% measured after treatment, DOC would be expected to (120 µg/L N, as shown in the inset of Figure 6) of the nearly equal TOC, because typical drinking water treatDON and 70% (53.1 µg/L N) of the total AAs were ment processes efficiently remove particulate and large removed during coagulation. Normalized removal of colloidal organic matter. The remaining total AAs are UV254 nearly mirrored that of total AAs, whereas less likely associated with smaller proteins or peptides and/or protein or peptide fragments in DOM fractions that are removal of bulk organic matter surrogates (DOC and challenging to remove (e.g., neutrals and hydrophilics). DON) was observed. The similarity of UV254 and total Bulk organic matter surrogates. Statistical hypothesis AA removal was observed only in waters with higher testing (t-test, p-value < 0.005) was used to determine the SUVA values (SUVA WTP 1 = 2.7). As illustrated by the effect of water treatment processes on organic matter and inset in Figure 6, the DON portion characterized as total AAs. The percentage of bulk organic matter surrogates AAs was reduced most during coagulation, and the accounted for by the total AAs in the treated water was remaining unit processes did not significantly affect the less than that in the raw water and statistically different distribution of the DON pool. Normalized Concentration—C/C0

DON—µg/L N

FIGURE 8 Removal of AAs, DOC, DON, and UV254 at WTP 11, applying chemical oxidants to plant influent before coagulation. Inset: Numbers above the bars indicate percentage of DON characterized by nitrogen present in measured total AAs

110

SEPTEMBER 2009 | JOURNAL AWWA • 101:9 | PEER-REVIEWED | DOTSON & WESTERHOFF

2009 © American Water Works Association

Normalized Concentration—C/C0

DON—µg/L N

N) total AA removal and 22% (40.2 µg/L N) DON During laboratory-scale jar tests from an Oklahoma removal occurred. reservoir and an Arizona reservoir, total AA removal The observed increase in non-AA DON may be associreached 46 and 51%, respectively. When compared with ated with the oxidation of AA-yielding DON-containing bulk organic matter surrogates, though, total AAs were species not removed during coagulation. Although measelectively removed. As shown in Figure 7, an increase in sured concentrations of total AAs removed are greater alum dosages to > 7 mg alum/mg DOC, removed no than those removed by coagulation alone, the application additional total AAs but continued to remove bulk of an oxidant in the presence of higher concentrations organic matter surrogates. of total AAs before coagulation will result in the formaThe jar-test removal of UV254 mirrored total AA tion of halogenated by-products. These by-products are removal for the Oklahoma reservoir (SUVA = 3.0) but not unlikely to be removed by other unit processes and may the Arizona reservoir (SUVA = 1.7). The similar removal of contribute significantly to the disinfection by-products UV254 and total AAs in high-SUVA water may be related found in the distribution system. to the removal of high-molecular-weight (MW) material, Ozonation before coagulation. Figure 9 shows that because high SUVA indicates greater concentrations of high primary ozonation (before coagulation) at WTP 12 MW humic DOM (i.e., macromolecular aromatic organproduced an increase in total AAs (164% or 68.5 µg/L ics). DOC and DON were removed to a similar degree to N) and lesser increases in DOC (4% or 0.23 mg/L C) coagulant dosages of < 7 mg alum/mg DOC; higher coaguand DON (42% or 181 µg/L N). A similar trend was lant dosages removed more DOC than DON. The preferential physical removal of total AAs during coagulation supports the observation FIGURE 9 Removal of AAs, DOC, DON, and UV254 at WTP 12, using primary ozonation before coagulation. Inset: Numbers above by Lee and Westerhoff (2006) that bars indicate percentage of DON characterized by nitrogen present coagulation preferentially removes in measured total AAs high-MW (> 1 kD) material over low-MW material (< 1 kD). On the basis of this study it is not surDOC Total AAs with DON characterized DON prising that AAs are preferentially Total AAs (N-basis) 700 Uncharacterized DON removed because polypeptides UV254 18 and proteins comprise combined 600 AAs and are typically larger than 4 500 7 5.5 kD (Voet & Voet, 2004); they 10 400 would be part of large MW material easily removed by coagulation. 300 3.0 This suggests that the remaining 200 AAs are associated with a NOM fraction not amenable to coagula100 2.5 tion (i.e., neutral, basic, or colloi0 dal organic matter). Furthermore, A B C D physically removing total AAs dur2.0 ing coagulation should reduce the formation of DBPs from AAs. 1.5 Applying oxidants to raw water. When chemical oxidants were applied to the raw water, 1.0 total AA removal during coagulation increased. For example, Figure 8 shows that at WTP 11, 0.5 nearly complete removal of total AAs (93%, 109 µg/L N) occurred 0.0 during coagulation when chlorine A—Plant influent B—Primary ozone C—Settled after D—Filter effluent dioxide was applied to the raw coagulation water. This corresponded to 21% Sample Point (66 µg/L N) DON removal. The AA—amino acid, C/C0—concentration data normalized to plant influent concentration C0 same trend occurred at WTP 10, (plant influent divided by plant influent), DOC—dissolved organic carbon, DON—dissolved where chlorine was added as a organic nitrogen, UV254—ultraviolet absorbance at 254 nm, WTP—water treatment plant preoxidant and 83% (42.6 µg/L DOTSON & WESTERHOFF | 101:9 • JOURNAL AWWA | PEER-REVIEWED | SEPTEMBER 2009

2009 © American Water Works Association

111

coagulation) was practiced, total AAs and organics decreased. At WTP 1, DON and total AA concentrations were reduced by 43 and 4.4 µg/L N, respectively, which resulted in a 1% reduction in the percent Total AAs with DOC DON accounted for by the total 300 characterized DON DON Uncharacterized DON Total AAs (N-basis) AAs (Figure 6). A similar trend in 18 UV254 DON and total AA removal 3 9 occurred during intermediate ozo200 6 nation at WTP 2 (Figure 10). Filtration. Full-scale filtration accounted for only minimal differ100 ences in total AAs. WTP 1 operated its filters in a biological mode. This filtration process reduced the total 1.0 0 AA concentration by 1.5 µg/L N A B C D and DON by 45 µg/L N (Figure 6). WTP 2, which also operated its fil0.8 ters in a biological mode, exhibited an increase of 3.6 µg/L N total AAs while removing 28 µg/L N DON 0.6 (Figure 10). Other adequately sampled facilities removed between 0.2 and 74 µg/L N DON and between 0.4 –11.55 (addition) and 1.18 µg/L N total AAs, although oxidants or polymers may have been added prior 0.2 to filtration. Unidentified dissolved organic nitrogen. The standard total AAs accounted for a minimal amount 0.0 A—Plant influent B—Settled after C—Afer ozonation D—Filter effluent (average of 15%) of the DON in raw coagulation waters and even less in treated waters (average of 4.1%). This unidentified AA—amino acid, C/C0—concentration data normalized to plant influent concentration C0 (plant influent divided by plant influent), DOC—dissolved organic carbon, DON—dissolved DON clearly warrants identification organic nitrogen, UV254—ultraviolet absorbance at 254 nm and further study because DON has been determined to be a precursor to nitrogenous DBPs (Lee et al, 2007) that are more geno- and cytotoxic than the currently observed at WTP 13, which also practiced primary regulated carbon-containing DBPs (Plewa et al, 2008; ozonation. This trend is likely related to the oxidaMuellner et al, 2007). tion of particulate organics, particularly algal cells, This unidentified DON could exist as a variety of which causes the cells to shrink and release their contents organic nitrogen-containing compounds, such as nonstan(Hammes et al, 2007). dard amino acids (e.g., lanthionine and dehydroalanine), Postponing the application of ozone until after coagulaamines (e.g., methylamine and dimethylamine), amides tion would likely improve water quality and plant effi(e.g., acetamide), heterocyclic nitrogen (e.g., pyrrole and ciency because the organic matter load to be oxidized pyridine), nucleic acids (e.g., DNA and RNA), organic would be smaller thus reducing the amount of ozone chloramines, or nitrogen-containing pharmaceuticals and required, as illustrated by the facilities using intermediate personal care products (e.g., caffeine). Of these nitrogenozonation (WTPs 1, 2, 4, and 16). Three of these plants containing compound classes, heterocyclic nitrogen may were adequately sampled for detailed analysis; at WTP 16, be the next most dominant portion of the organic nitrogen one of the samples was lost because of a broken bottle. present in raw waters, accounting for upward of 25% Intermediate ozonation. The point at which ozona(Vairavamurthy & Wang, 2002; Schulten & Schnitzer, tion is used during the water treatment process can affect 1997; Tuschall & Brezonik, 1980). Currently available water quality changes from bulk organic matter surroanalytical techniques (15N-NMR and pyrolysis gas chrogates and AAs. When intermediate ozonation (after Normalized Concentration—C/C0

DON—µg/L N

FIGURE 10 Removal of AAs, DOC, DON, and UV254 at WTP 2, using coagulation and intermediate ozonation. Inset: Numbers above bars indicate percentage of DON characterized by nitrogen present in measured total AAs

112

SEPTEMBER 2009 | JOURNAL AWWA • 101:9 | PEER-REVIEWED | DOTSON & WESTERHOFF

2009 © American Water Works Association

matography/mass spectrometry) used to determine nitrogen heterocycles are not quantitative and may result in destruction of some organic compounds (Vairavamurthy &Wang, 2002; Hatcher et al, 2001; da Cunha et al, 2000; Mahieu et al, 2000; Thorn et al, 1996; Gadel & Bruchet, 1987). Recently, improved recovery and degradation of nitrogenous species in NOM through the use of microscale sealed vessel (MSSV) pyrolysis over flash pyrolysis has clearly revealed the presence of heterocyclic nitrogen compounds in NOM isolates (e.g., pyrroles and porphyrins; Berwick et al, 2007). Further analytical method develop-

P

30% (1.1 mg/L C) of DOC. Total AAs accounted for a smaller percentage of bulk organic matter surrogates in treated waters, showing that these species are preferentially removed over other bulk organic matter surrogates. • Coagulation preferentially removed total AAs compared with bulk organic matter surrogates on a percentage basis. Furthermore, coagulation removed the largest mass of total AAs compared with other unit processes (e.g., filtration, disinfection). • The location and type of oxidant applied during treatment affected the total AA concentration. Primary ozonation increased the concentration of total AAs and bulk organic matter surrogates by oxidizing particulate organics into dissolved organics. In contrast, intermediate ozonation decreased total AAs by oxidizing dissolved organics. Application of chlorine or chlorine dioxide decreased the concentration of total AAs regardless of the application point but may have promoted the formation of disinfection by-products, some of which are regulated. • The presence of AAs throughout the drinking water treatment train has the potential to form nitrogen-containing disinfection by-products (N-DBPs)—by-products more geno- and cytotoxic than the currently regulated carbon-containing disinfection by-products (C-DBPs)— and at concentrations (e.g. ng/L) of concern. For example, if 100% of the AA-associated nitrogen incorporated as one of the two nitrogens in N-nitrosodimethylamine, a by-product could form at a concentration of public health concern.

ostponing the application of ozone until after coagulation would likely improve water quality and plant efficiency because the organic matter load to be oxidized would be smaller, thus reducing the amount of ozone required. ment will be required to quantitatively measure additional organic nitrogen-containing species.

CONCLUSION In measuring specific organic nitrogen species (free and total AAs, free and total ASs) and bulk organic matter surrogates (DOC, DON, and UV254) at 16 full-scale US drinking water treatment plants affected by upstream algal blooms or wastewater discharges, the authors observed the following: • Algal- or wastewater-affected drinking waters are more nitrogen-enriched (DOC-to-DON = 12.6) than previously observed waters that were not targeted as being under such influences. • In raw waters, free AA concentrations varied between 0.12 and 2.5 µg/L N with an average concentration of 0.38 µg/L N, whereas total AA concentrations varied between 29 and 171 µg/L N with an average concentration of 41.1 µg/L N. • Free AAs accounted for < 1.3% of the DON, whereas total AAs accounted for 15% of the DON on average and varied between 0.17 and 47%. • Total ASs accounted for, on average, 0.8% of the DON and did not exceed 2.6%. • Serine, glycine, and alanine were statistically most likely to occur in the five most dominant AA species in a given water sample. AA species found in treated drinking waters varied slightly from those found in raw waters; in treated waters tyrosine was likely to be found as a dominant free AA and threonine as a dominant total AA. • On average, drinking water treatment removed 65% (34 µg/L N) of total AAs, 30% (84.4 µg/L N) of DON, and

ACKNOWLEDGMENT This work was supported by the Water Research Foundation (project 3014). The authors thank the Water Research Foundation and the USEPA for their financial, technical, and administrative assistance in funding and managing this research. The authors thank Stuart Krasner and the chemists of the Metropolitan Water District of Southern California for their analysis of wastewater tracers and coordination of sample programs, as well as each of the participating utilities for their support. ABOUT THE AUTHORS

Aaron Dotson (to whom correspondence should be addressed) is a research associate at the University of Colorado, 1111 Engineering Dr., ECOT 441, UCB 428, Boulder, CO 80309, aaron. [email protected]. He has a BS degree from the University of Arizona, Tucson, and MSE and PhD degrees

DOTSON & WESTERHOFF | 101:9 • JOURNAL AWWA | PEER-REVIEWED | SEPTEMBER 2009

2009 © American Water Works Association

113

from Arizona State University, Tempe. Before pursuing an academic career, Dotson worked for Malcolm Pirnie Inc., where he was part of the process and planning group, designing water treatment plant facilities. His doctoral work focused on the character of nitrogenrich dissolved organic matter and its propensity to form nitrogenous disinfection by-products. Paul Westerhoff is professor and chair of engineering at Arizona State University, Tempe.

3Model

PB-901 jar tester, Phipps & Bird, Richmond, Va. Glass Microfibre Filters GF/F, Whatman Inc., Piscataway,

4Whatman

N.J. 5RE-300 Rotary Evaporator, Yamato Scientific Corp., Santa Clara, Calif. 6FreeZone 6, Labconco Corp., Kansas City, Mo. 7AccQ·TagTM Chemistry Kit, Waters Corp., Milford, Mass. 8Alliance 2695 Separations Module, Waters Corp., Milford, Mass. 92475 Multi-Wavelength Fluorescence Detector, Waters Corp., Milford, Mass. 10TOC-V CSH Total Organic Carbon Analyzer, Shimadzu Corp., Kyoto, Japan 11TMN-1 Total Nitrogen Analyzer, Shimadzu Corp., Kyoto, Japan 12Dionex DX-120 Ion Chromatograph, Sunnyvale, Calif.

Date of submission: 11/17/2008 Date of acceptance: 03/12/2009

FOOTNOTES 1GE

Osmonics, Minnetonka, Minn. Infinity, Barnstead International, Dubuque, Iowa

If you have a comment about this article, please contact us at [email protected].

2NANOpure

REFERENCES Berwick, L.; Greenwood, P.; Kagi, R.; & Croué, J.P., 2007. Thermal Release of Nitrogen Organics From Natural Organic Matter Using Micro Scale Sealed Vessel Pyrolysis. Organic Geochem., 38:7:1073. Bieber, T. & Trehy, M.L.,1983. Dihaloacetonitriles in Chlorinated Natural Waters. Water Chlorination, Environmental Impact, and Health Effects, Book 1 Chemistry and Water Treatment (R. Jolley, R. Bull, W. Davis, & S. Katz, editors). Ann Arbor Science Publishers, Ann Arbor, Mich. Chen, W., Westerhoff, P., Leenheer, J.A. & Booksh, K., 2003. Fluorescence Excitation—Emission Matrix Regional Integration to Quantify Spectra for Dissolved Organic Matter. Envir. Sci. & Technol., 37:24:5701. Chinn, R. & Barrett, S., 2000. Occurrence of Amino Acids in Two Drinking Water Sources. Natural Organic Matter and Disinfection Byproducts (S. Barret, S. Krasner, & G. Amy, editors). Oxford University Press, New York. Conyers, B. & Scully, F.E., 1997. Chloramines .5: Products and Implications of the Chlorination of Lysine in Municipal Wastewaters. Envir. Sci. & Technol., 31:6:1680. da Cunha, L.C.; Serve, L.; Gadel, F.; & Blazi, J.L., 2000. Characterization of Riverine Particulate Organic Matter by Pyrolysis-Gas Chromatography-Mass Spectrometry. Sci. of the Total Envir., 256:2:191. Dotson, A., 2008. Structure and Treatability of Organic Nitrogenenriched Drinking Water. Doctoral dissertation, University of Arizona, Tempe.

Catchments—the Significance of Water Flow Paths. Biogeochemistry, 50:2:137. Hammes, F.; Meylan, S.; Salhi, E.; Koster, O.; Egli, T.; & Von Gunten, U., 2007. Formation of Assimilable Organic Carbon (AOC) and Specific Natural Organic Matter (NOM) Fractions During Ozonation of Phytoplankton. Water Res., 41:7:1447. Hatcher, P.G.; Dria, K.J.; Kim, S.; & Frazier, S.W., 2001. Modern Analytical Studies of Humic Substances. Soil Sci., 166:11:770. Hedges, J.I.; Cowie, G.L.; Richey, J.E.; Quay, P.D.; Benner, R.; Strom, M.; & Forsberg, B.R., 1994. Origins and Processing of Organic Matter in the Amazon River as Indicated by Carbohydrates and Amino Acids. Limnol. & Oceanog., 39:4:743. Hirose, Y.; Maeda, N.; Ohya, T.; Nojima, K.; & Kanno, S., 1989. Formation of Cyanogen Chloride by the Reaction of Peptides and Proteins With Hypochlorous Acid in the Presence of Ammonium Ion. Chemosphere, 18:11-12:2101. Hirose, Y.; Maeda, N.; Ohya, T.; Nojima, K.; & Kanno, S., 1988. Formation of Cyanogen Chloride by the Reaction of Amino Acids With Hypochlorous Acid in the Presence of Ammonium Ion. Chemosphere, 17:5:865. Hong, H.C.; Wong, M.H.; & Liang, Y., 2008. Amino Acids as Precursors of Trihalomethane and Haloacetic Acid Formation During Chlorination. Archives of Envir. Contamination & Toxicol., 56:4:638. Hrudey, S.E.; Gac, A.; & Daignault, S.A., 1988. Potent Odor-Causing Chemicals Arising From Drinking-Water Disinfection. Water Sci. & Technol., 20:8:55.

Dotson, A.; Westerhoff, P.; Chen, B.; & Lee, W., 2008. Organic Nitrogen Occurrence and Characterization (T. Karanfil, S.W. Krasner, P. Westerhoff, & Y. Xie, editors). American Chemical Society, Washington.

Hureiki, L. & Prévost, M., 1996. Evolution of Amino Acids in Two Drinking Water Treatment Plants. Revue Des Sciences De L’eau, 9:3:297.

Duan, S. & Bianchi, T.S., 2007. Particulate and Dissolved Amino Acids in the Lower Mississippi and Pearl Rivers (USA). Marine Chemistry, 107:2:214.

Hureiki, L.; Croué, J.P.; & Legube, B., 1994. Chlorination Studies of Free and Combined Amino Acids. Water Res., 28:12:2521.

Gadel, F. & Bruchet, A., 1987. Application of Pyrolysis-Gas Chromatography-Mass Spectrometry to the Characterization of Humic Substances Resulting From Decay of Aquatic Plants in Sediments and Waters. Water Res., 21:10:1195. Hach Corp., 2007. Method 10205. Nitrogen, Ammonia, Salicylate Method. Hach Corp., Loveland, Colo. Hagedorn, F.; Schleppi, P.; Waldner, P.; & Fluhler, H., 2000. Export of Dissolved Organic Carbon and Nitrogen From Gleysol Dominated

114

Kielland, K.; McFarland, J.; & Olson, K., 2006. Amino Acid Uptake in Deciduous and Coniferous Taiga Ecosystems. Plant and Soil, 288:1:297. Le Cloirec, C. & Martin, G., 1985. Evolution of Amino Acids in Water Treatment Plants and the Effect of Chlorination on Amino Acids. Proc. Conf. on Water Chlorination: Envir. Impact & Health Effects. Oak Ridge, Tenn. Lee, W. & Westerhoff, P., 2006. Dissolved Organic Nitrogen Removal During Water Treatment by Aluminum Sulfate and Cationic Polymer Coagulation. Water Res., 40:20:3767.

SEPTEMBER 2009 | JOURNAL AWWA • 101:9 | PEER-REVIEWED | DOTSON & WESTERHOFF

2009 © American Water Works Association

Lee, W. & Westerhoff, P., 2005. Dissolved Organic Nitrogen Measurement Using Dialysis Pretreatment. Envir. Sci. & Technol., 39:3:879. Lee, W.; Westerhoff, P.; & Croué, J.P., 2007. Dissolved Organic Nitrogen as a Precursor for Chloroform, Dichloroacetonitrile, N-Nitrosodimethylamine, and Trichloronitromethane. Envir. Sci. & Technol., 41:15:5485. Lee, J.H.; Na, C.Z.; Ramirez, R.L.; & Olson, T.M., 2006a. Cyanogen Chloride Precursor Analysis in Chlorinated River Water. Envir. Sci. & Technol., 40:5:1478. Lee, W.; Westerhoff, P.; & Esparza-Soto, M., 2006b. Occurrence and Removal of Dissolved Organic Nitrogen in US Water Treatment Plants. Jour. AWWA, 98:10:102. Lytle, C.R. & Perdue, E.M., 1981. Free, Proteinaceous, and Humic-Bound Amino Acids in River Water Containing High Concentrations of Aquatic Humus. Envir. Sci. & Technol., 15:2:224. Mahieu, N.; Olk, D.C.; & Randall, E.W., 2000. Accumulation of Heterocyclic Nitrogen in Humified Organic Matter: A N-15-NMR Study of Lowland Rice Soils. European Jour. of Soil Sci., 51:3:379. Martens, D.A. & Loeffelmann, K.L., 2003. Soil Amino Acid Composition Quantified by Acid Hydrolysis and Anion Chromatography-Pulsed Amperometry. Jour. Agricultural & Food Chem., 51:22:6521. McCormick, E.F.; Conyers, B.; & Scully, F.E., 1993. N-Chloroaldimines. 2. Chlorination of Valine in Model Solutions and in a Wastewater. Envir. Sci. & Technol., 27:2:255. Mehrsheikh, A.; Bleeke, M.; Brosillon, S.; Laplanche, A.; & Roche, P., 2006. Investigation of the Mechanism of Chlorination of Glyphosate and Glycine in Water. Water Res., 40:16:3003. Mitch, W.A.; Krasner, S.; Westerhoff, P.; & Dotson, A., 2008. Occurrence and Formation of Nitrogenous Disinfection By-Products, Water Research Foundation, Denver, in press. Muellner, M.G.; Wagner, E.D.; McCalla, K.; Richardson, S.D.; Woo, Y.T.; & Plewa, M.J., 2007. Haloacetonitriles versus Regulated Haloacetic Acids: Are Nitrogen-Containing DBPs More Toxic? Envir. Sci. & Technol., 41:2:645. Na, C.Z. & Olson, T.M., 2006. Mechanism and Kinetics of Cyanogen Chloride Formation From the Chlorination of Glycine. Envir. Sci. & Technol., 40:5:1469. Na, C.Z. & Olson, T.M., 2004. Stability of Cyanogen Chloride in the Presence of Free Chlorine and Monochloramine. Envir. Sci. & Technol., 38:22:6037. Peake, E.; Hodgson, G.W; & Baker, B.L., 1972. Hydrogeochemistry of Surface Waters of MacKenzie River Drainage Basin, Canada. 2. Contribution of Amino-acids, Hydrocarbons and Chlorins to Beaufort Sea by MacKenzie River System. Geochim. & Cosmochim.—Acta, 36:8:867. Pehlivanoglu-Mantas, E. & Sedlak, D.L., 2006. Wastewater-derived Dissolved Organic Nitrogen: Analytical Methods, Characterization, and Effects—A Review. Envir. Sci. & Technol., 36:3:261. Peterson, W.H.; Fred, E.B.; & Domogalla, B.P., 1925. The Occurrence of Amino Acids and Other Organic Nitrogen Compounds in Lake Water. Jour. Biological Chemistry, 63:2:287. Plewa, M.J.; Wagner, E.D.; Muellner, M.G.; Hsu, K.M.; & Richardson, S.D., 2008. Comparative Mammalian Cell Toxicity of N-DBPs and C-DBPs. Disinfection By-products in Drinking Water: Occurrence, Formation, Health Effects, and Control (T. Karanfil, S.W. Krasner, P. Westerhoff, Y. Xie, editors). American Chemical Society, Washington. Prévost, M.; Gauthier, C.; Hureiki, L.; Desjardins, R.; & Servais, P., 1998. Removal of Amino Acids, Biodegradable Organic Carbon, and Chlorine Demand by Biological Filtration in Cold Water. Envir. Technol., 19:903.

Ramesh, A.; Lee, D.J.; & Hong, S.G., 2006. Soluble Microbial Products (SMP) and Soluble Extracellular Polymeric Substances (EPS) From Wastewater Sludge. Applied Microbiol. & Biotechnol., 73:1:219. Rostad, C.E.; Leenheer, J.A.; & Daniel, S.R., 1997. Organic Carbon and Nitrogen Content Associated With Colloids and Suspended Particulates From the Mississippi River and Some of its Tributaries. Envir. Sci. & Technol., 31:11:3218. Schulten, H.R. & Schnitzer, M., 1997. The Chemistry of Soil Organic Nitrogen: A Review. Biology & Fertility of Soils, 26:1:1. Scully, F.E.; Howell, G.D.; Kravitz, R.; Jewell, J.T.; Hahn, V.; & Speed, M., 1988. Proteins in Natural Waters and Their Relation to the Formation of Chlorinated Organics During Water Disinfection. Envir. Sci. & Technol., 22:5:537. Shoji, T.; Yamamoto, Y.; Nishikawa, D.; Kurihara, K.; & Ueda, H., 2003. Amino Acids in Stream Water Are Essential for Salmon Homing Migration. Fish Physiology & Biochem., 28:1-4:249. Standard Methods for the Examination of Water and Wastewater, 2001 (20th ed., suppl.). APHA, AWWA, and WEF, Washington. Thorn, K.A.; Pettigrew, P.J.; & Goldenberg, W.S., 1996. Covalent Binding of Aniline to Humic Substances. 2. N-15 NMR Studies of Nucleophilic Addition Reactions. Envir. Sci. & Technol., 30:9:2764. Thurman, E.M., 1985. Organic Geochemistry of Natural Waters. Martinus Nijohoff/Dr. W. Junk, Publishers, Dordrecht, the Netherlands. Trehy, M.L. & Bieber, T.L.,1981. Detection, Identification, and Quantitative Analysis of Dihaloacetonitriles in Chlorinated Natural Waters. Advances in the Identification and Analysis of Organic Pollutants in Water (L. Keith, editor). Ann Arbor Science Publishers Inc., Ann Arbor, Mich. Trehy, M.L.; Yost, R.A.; & Miles, C.J., 1986. Chlorination By-products and Amino Acids in Natural Waters. Envir. Sci. & Technol., 20:11:1117. Tuschall, J.R. & Brezonik, P.L., 1980. Characterization of Organic Nitrogen in Natural Waters—Its Molecular Size, Protein Content, and Interactions With Heavy Metals. Limnol. & Oceanog., 25:3:495. Unger, D.; Gaye-Haake, B.; Neumann, K.; Gebhardt, A.C.; & Ittekkot, V., 2005. Biogeochemistry of Suspended and Sedimentary Material in the Ob and Yenisei Rivers and Kara Sea: Amino Acids and Amino Sugars. Continental Shelf Res., 25:4:437. USEPA (US Environmental protection Agency), 1999. Enhanced Coagulation and Enhanced Precipitative Softening Guidance Manual. EPA 815-R-99-012. Ofce. of Water, Washington. Vairavamurthy, A. & Wang, S., 2002. Organic Nitrogen in Geomacromolecules: Insights on Speciation and Transformation With K-edge XANES Spectroscopy. Envir. Sci. & Technol., 36:14:3050. Vanderford, B.J.; Pearson, R.A.; Rexing, D.J.; & Snyder, S.A., 2003. Analysis of Endocrine Disruptors, Pharmaceuticals, and Personal Care Products in Water Using Liquid Chromatography/Tandem Mass Spectrometry. Analytical Chemistry, 75:6265. Voet, D. & Voet, J.G., 2004. Biochemistry (3rd edition). John Wiley & Sons Inc., Hoboken, N.J. Volk, C.J.; Volk, C.B.; & Kaplan, L.A., 1997. Chemical Composition of Biodegradable Dissolved Organic Matter in Streamwater. Limnol. & Oceanog., 42:1:39. Waters Corp., 1993. Waters AccQ·Tag Chemistry Package Instruction Manual. Millipore Corp., Billerica, Mass. Westerhoff, P. & Mash, H., 2002. Dissolved Organic Nitrogen in Drinking Water Supplies: A Review. Jour. Water Supply Res. & Technol.– Aqua, 51:8:415.

DOTSON & WESTERHOFF | 101:9 • JOURNAL AWWA | PEER-REVIEWED | SEPTEMBER 2009

2009 © American Water Works Association

115