Ultraviolet Irradiation Effects Incorporation of Nitrate

Journal of Environmental Quality

TECHNICAL REPORTS SURFACE WATER QUALITY

Ultraviolet Irradiation Effects Incorporation of Nitrate and Nitrite Nitrogen into Aquatic Natural Organic Matter Kevin A. Thorn* and Larry G. Cox

U

ltraviolet (UV) light is being used increasingly

One of the concerns regarding the safety and efficacy of ultraviolet radiation for treatment of drinking water and wastewater is the fate of nitrate, particularly its photolysis to nitrite. In this study, 15N NMR was used to establish for the first time that UV irradiation effects the incorporation of nitrate and nitrite nitrogen into aquatic natural organic matter (NOM). Irradiation of 15N-labeled nitrate in aqueous solution with an unfiltered medium pressure mercury lamp resulted in the incorporation of nitrogen into Suwannee River NOM (SRNOM) via nitrosation and other reactions over a range of pH from approximately 3.2 to 8.0, both in the presence and absence of bicarbonate, confirming photonitrosation of the NOM. The major forms of the incorporated label include nitrosophenol, oxime/nitro, pyridine, nitrile, and amide nitrogens. Natural organic matter also catalyzed the reduction of nitrate to ammonia on irradiation. The nitrosophenol and oxime/nitro nitrogens were found to be susceptible to photodegradation on further irradiation when nitrate was removed from the system. At pH 7.5, unfiltered irradiation resulted in the incorporation of 15N-labeled nitrite into SRNOM in the form of amide, nitrile, and pyridine nitrogen. In the presence of bicarbonate at pH 7.4, Pyrex filtered (cutoff below 290–300 nm) irradiation also effected incorporation of nitrite into SRNOM as amide nitrogen. We speculate that nitrosation of NOM from the UV irradiation of nitrate also leads to production of nitrogen gas and nitrous oxide, a process that may be termed photochemodenitrification. Irradiation of SRNOM alone resulted in transformation or loss of naturally abundant heterocyclic nitrogens.

for disinfection of wastewater and primary disinfection of drinking water (Oppenlander, 2003; White, 2010). Nitrate is a common contaminant in both sources of water, and its fate on UV irradiation has therefore been the focus of numerous studies, particularly its photolysis to nitrite (Mark et al., 1996; Sharpless and Linden, 2001). In the United States, the maximum allowable contaminant levels (MCL) in drinking water are 10 ppm for nitrate (10 mg nitrate-N L−1 or 44.3 mg nitrate L−1) and 1 ppm for nitrite (1 mg nitrite-N L−1 or 3.3 mg nitrite L−1). These limits have not gone unchallenged. For example, epidemiological studies relating incidences of cancer to concentrations of nitrite in drinking water in China have called into question whether the nitrite MCL of 1 ppm is too high (Tickell, 2008). Because nitrite is more toxic than nitrate (it oxidizes hemoglobin to methemoglobin, which cannot transfer oxygen to tissues), factors affecting its photoproduction rates have been determined under various UV treatment conditions for drinking waters, including the effects of inorganic carbon and natural organic matter (NOM) (Buchanan et al., 2006; Sharpless and Linden, 2001; Thomson et al., 2004). Natural organic matter was shown to increase initial quantum yields of nitrite from nitrate at pH 8 and 6, whereas bicarbonate decreased yields at pH 8 but not pH 6 (Sharpless and Linden, 2001). One question that has not been addressed is whether the photolysis products of nitrate can become incorporated into NOM, a process that would ultimately affect nitrite quantum yields and the mass balance of nitrogen in the treatment system, and what the structural forms of the incorporated nitrogen may be. As summarized by Goldstein and Rabani (2007), the absorption spectrum of nitrate is dominated by a weak n → π* band around 302 nm (ε = 7.2 M−1 cm−1) and a stronger π → π* band at 200 nm (ε = 9900 M−1 cm−1). Excitation in the π → π* band (λ < 280 nm) is considered to proceed by the two primary processes R1 and R2, whereas excitation in the n → π* band (λ > 280 nm) proceeds through R1 and R3 (Mack and Bolton, 1999). The main contribution to nitrite formation at λ < 280 nm is through the decomposition of peroxynitrite.

Copyright © 2012 by the American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

U.S. Geological Survey, Denver Federal Center, MS 408, Denver, CO 80225-0046. Assigned to Associate Editor Robert Cook.

J. Environ. Qual. 41:865–881 (2012) doi:10.2134/jeq2011.0335 Received 14 Sept. 2011. Supplemental data file is available online for this article. *Corresponding author ([email protected]). © ASA, CSSA, SSSA 5585 Guilford Rd., Madison, WI 53711 USA

Abbreviations: CP/MAS, cross polarization/magic angle spinning; DEPT, distortionless enhancement by polarization transfer; FID, free induction decay; FT-ICR, Fourier transform ion cyclotron resonance; FTIR, Fourier transform infrared; N-DBP, nitrogen-containing disinfection by-product; NMR, nuclear magnetic resonance; NOE, nuclear Overhauser enhancement; NOM, natural organic matter; SRNOM, Suwannee River natural organic matter.

865

NO3− + hν → ∙NO2 + ∙O− (λ < 280 nm) (∙O− + H2O ↔ ∙OH + OH−)

[R1]

NO3− + hν → ONOO− (λ < 280 nm)

[R2]

NO3− + hν → NO2− + 1/2 O2 (λ > 280 nm)

[R3]

Nitrite (absorption maximum near 354 nm) and undissociated nitrous acid in turn undergo photolysis primarily to nitric oxide and hydroxyl radical after absorption of light in the 300- to 400-nm wavelength region; photolysis of nitrite to nitrogen dioxide and hydrated electron is a minor parallel reaction [R4– R7] (Fischer and Warneck, 1996). Secondary reactions have been summarized by Fischer and Warneck (1996). In the absence of other solutes, hydroxyl radical reacts with nitrite or nitrous acid to form nitrogen dioxide, which largely combines with nitric oxide to form dinitrogen trioxide. NO2− + hν (+H+) → ∙NO + ∙OH (300–400 nm)

[R4]

HNO2 + hν → ∙NO + ∙OH (300–400 nm)

[R5]

NO2− + hν → ∙NO2 + eaq− (300–400 nm)

[R6]

NO2− + ∙OH → ∙NO2 + OH −

[R7]

HNO2 + ∙OH → ∙NO2 + H2O Formation of nitrite, peroxynitrite, nitric oxide, and nitrogen dioxide can lead to nitrosation and nitration reactions of organic compounds in aqueous solution, depending on pH and concentrations of reactants. Nitrite and peroxynitrate in their protonated forms as nitrous acid and peroxynitrous acid act as nitrosating and nitrating species, respectively. Nitrogen dioxide (and nitric acid oxidized to nitrogen dioxide) can affect free radical and electrophilic nitration reactions. Details of the reaction mechanisms for these nitrosating and nitrating species can be found in several reviews (Olah et al., 1989; Williams, 1988; Williams, 2004). Nitration of catechol, hydroquinone, and naphthalene and both nitration and nitrosation of phenol have been documented after UV irradiation of these compounds in water in the presence of nitrate (Dzengel et al., 1999; Machado and Boule, 1995; Niessen et al., 1988; Vione et al., 2001; Vione et al., 2005). Aqueous phase nitrosation of phenol, resorcinol, and dimethylamine and nitration of naphthalene, hydroquinone, and catechol have been documented after UV irradiation of these compounds in the presence of nitrite (Machado and Boule, 1995; Ohta et al., 1982; Suzuki et al., 1984; Vione et al., 2005). Enhancement of aromatic photonitration at acidic pH has been demonstrated for a number of compounds on nitrate photolysis (Minero et al., 2007). The susceptibility of NOM to nitrosation is well known (Stevenson, 1994; Thorn and Mikita, 2000). Nitrogen-15 NMR analyses have demonstrated that soil and aquatic humic substances undergo carbon and nitrogen nitrosation when reacted with sodium nitrite in the approximate pH range of 3 to 6 in laboratory reactions (Thorn and Mikita, 2000); nitrosation of organic matter has been reported to occur in soils with pH values as high as 7.8 (Nelson, 1982). Nitrosative decarboxylation of aromatic carboxylic acids and demethoxylation of aromatic methyl ether groups are among 866

the numerous reactions that occur during nitrosation of NOM. Reaction of nitrite with NOM to produce nitric oxide, nitrous oxide, and nitrogen gas is one of the processes associated with chemodenitrification (Stevenson, 1994; Thorn and Mikita, 2000). The objectives of this study were to determine if the photolysis products of nitrate and nitrite become incorporated into aquatic NOM subjected to UV irradiation, via nitrosation or nitration reactions. The effects of UV irradiation on aquatic NOM are complex and have been extensively investigated over the past several decades (Steinberg, 2003). Absorption of solar and UV lamp radiation by aquatic NOM results in the loss of its absorbance in the UV and visible regions (photobleaching) and in a decrease in its molecular weight, concomitant with the formation of low-molecular-weight degradation products, including carbon dioxide, carbon monoxide, ammonium, amino acids, and numerous aldehydes, ketones, and carboxylic acids (Allard et al., 1994; Buffam and McGlathery, 2003; Bushaw et al., 1996; Kieber et al., 1990; Miller, 1998; Stedmon et al., 2007; Steinberg, 2003; Tarr et al., 2001; Wang et al., 2000). Production of bicarbonate via photochemical decarboxylation is a significant process (Miller and Zepp, 1995). A wavelength dependence for photochemical production of dissolved inorganic carbon from NOM has been documented (Graneli et al., 1998; Wang et al., 2009). In Suwannee River humic acid, for example, UV-B, UV-A, and visible wavelengths accounted for 31.8, 32.6, and 25.6%, respectively, of dissolved inorganic carbon production under simulated solar irradiation (Wang et al., 2009). Photodegradation occurs through direct photolysis, where light absorption by NOM leads to bond cleavage, and through indirect or sensitized processes, wherein NOM may react with photochemically generated reactive oxygen species, including hydroxyl radical (Goldstone et al., 2002; Steinberg, 2003). Natural organic matter can act as a source and a sink for hydroxyl radical (Vione et al., 2006). Carbon-13 NMR studies have documented successive losses of carboxylic acid, ketone, quinone, aromatic, and O-alkyl and methoxyl carbons, coupled to the appearance of bicarbonate, formate, acetate, and succinate, as aquatic humic, fulvic, and XAD-4 acids are photochemically degraded with unfiltered UV light (Thorn et al., 2010). Only a few reports have examined the photochemical production of nitrite directly from NOM, with conflicting results. In one study, nitrite was not detected in control experiments of irradiated Suwannee River NOM (SRNOM) (Sharpless and Linden, 2001), whereas nitrite was detected in irradiated aquatic humic substances isolated from North Carolina coastal waters (Kieber et al., 1999). During the initial phases of UV irradiation of NOM in the presence of nitrate, therefore, processes of photochemical release of bicarbonate, ammonium, amino acids, and other low molecular weight carbonyl compounds from NOM potentially compete with incorporation of nitrogen into NOM via nitrosation and nitration reactions. In this study, SRNOM was subjected to UV irradiation with an unfiltered medium pressure mercury lamp in the presence of 15N-labeled sodium nitrate, at initial pH values of 3.2 and 8.0, to establish incorporation of nitrate nitrogen. The samples were analyzed by liquid-state 15N NMR and compared with spectra of SRNOM nitrosated with 15 N-labeled sodium nitrite without UV irradiation. The initial Journal of Environmental Quality

irradiation experiments were performed at concentrations of SRNOM and nitrate higher than normally encountered in water treatment to maximize incorporation of labeled nitrogen for NMR detection. In subsequent experiments, SRNOM was irradiated in the presence of nitrate and sodium bicarbonate at concentrations more relevant to water treatment. Suwannee River NOM was also reacted separately with 15N-labeled nitrogen dioxide and nitric acid as control reactions for free radical and electrophilic nitration, respectively. To establish UV light–induced condensation of nitrite-N, SRNOM was irradiated in the presence of Na15NO2 at pH 7.5 without a filter and in the presence of Na15NO2 and bicarbonate at pH 7.4 with a Pyrex filter, which cuts off radiation below 290 to 300 nm. Solid-state cross polarization/magic angle spinning (CP/MAS) spectra were also recorded on SRNOM before and after unfiltered UV irradiation in the absence of nitrate and nitrite to examine the photodegradation of the NOM naturally abundant nitrogen functionality (see Supplemental Figure S1 for experimental flow chart).

Materials and Methods Materials Sodium nitrate and sodium nitrite, 99 atom% 15N each, were purchased from ISOTEC. Nitric acid (40%; 99 atom% 15N) was purchased from Cambridge Isotope Laboratories. Nitrogen dioxide, 99 atom% 15N, was purchased from ICON Stable Isotopes. Suwannee River NOM (reverse osmosis isolate) was purchased from the International Humic Substances Society. Most experiments were performed with SRNOM that was filtered to lower the ash content from 7.0 to 4.3%. About 300 to 400 mg of the NOM was added to 0.5 L deionized water, passed through a 1-μm glass fiber filter (Gelman) and then a 0.45-μm Type HA Millipore filter, and freeze dried. Elemental analyses were performed by Huffman Laboratories.

Photochemical Reactor The photochemical reactor (Ace Glass Inc.) consisted of a 450 W, medium pressure, quartz, mercury-vapor lamp housed in a quartz immersion well and equipped with a 1.0-L reaction vessel. The reaction vessel was maintained at 23 to 25°C with a water cooling jacket and remained open to the atmosphere through three ports to maintain oxic conditions. Light filters were not used except for the experiment in which SRNOM was irradiated in the presence of sodium nitrite and sodium bicarbonate, described below.

Irradiation of Suwannee River Natural Organic Matter in the Presence of Nitrate Suwannee River NOM was irradiated in the presence of sodium nitrate without pH adjustment (pH 3.2), in the presence of nitrate with pH adjusted to 8.0, and with pH adjusted to 7.5 in the presence of sodium nitrate and sodium bicarbonate. For the first reaction, SRNOM (400 mg) and Na15NO3 (80 mg) were dissolved in 900 mL deionized H2O and irradiated for 4 h (195 mg C/L and 65 mg nitrate L−1). The initial and final pH readings were 3.2 and 4.3, respectively. The reaction solution was then freeze dried. After recording of an inverse-gated decoupled 15N NMR spectrum in aqueous solution, the sample was dialyzed against a 100-Da MWCO cellulose acetate membrane, H+–saturated on an Dowex MSC-1 cation exchange resin, freeze dried, and redissolved in DMSO-d6 for ACOUSTIC and distortionless enhancement by polarization transfer (DEPT) 15N NMR spectra. For the second reaction, 300 mg of SRNOM was adjusted to pH 8.0 with 0.1 mol L−1 NaOH in a total volume of 900 mL H2O, charged with 80 mg Na15NO3, and irradiated for 4 h, after which the pH had decreased to 7.0 (146 mg C L−1 and 65 mg nitrate L−1). The solution was then freeze dried and subjected to 15N NMR analyses before and after dialysis as in the first reaction. For the third reaction, 30 mg of SRNOM, 84 mg of NaHCO3, and 30 mg of Na15NO3 in 900 mL of H2O were irradiated for 4 h (16 mg C L−1, 24 mg nitrate L−1, and 68 mg bicarbonate L−1). The initial and final pH readings were 7.5 and 6.9, respectively. The reaction was performed in quadruplicate and the combined reaction solutions were then freeze dried. Solidstate 15N NMR spectra were recorded on the sample before and after dialysis, and a liquid-state 15N NMR spectrum was recorded on a portion of the sample in aqueous solution before dialysis. A 126-mg portion of the SRNOM sample irradiated with nitrate at pH 3.2 and then dialyzed was subjected to an additional 2 h of irradiation to examine the photolability of the incorporated nitrogen.

Irradiation of Suwannee River Natural Organic Matter in the Presence of Nitrite

Filtered SRNOM (300 mg) dissolved in 900 mL deionized H2O was irradiated to 62% of initial absorbance at 465 nm (62%IA465nm) in duplicate reactions and freeze dried for solidstate CP/MAS 15N and 13C NMR analysis. The initial and final pH readings were 3.4 and 4.4, respectively.

Suwannee River NOM was irradiated in the presence of Na15NO2 at pH 7.5 and in the presence of Na15NO2 and NaHCO3 at pH 7.4, the latter irradiation performed using a Pyrex filter. In the first reaction, 65 mg of Na15NO2 was added to a solution of 324 mg of SRNOM adjusted to pH 7.5 in a total volume of 900 mL (175 mg C L−1 and 48 mg nitrite L−1). The solution was irradiated for 4 h (final pH, 7.3), freeze dried, and redissolved in 75% D2O for 15N NMR analysis. In the second reaction, a solution of 30 mg of Na15NO2 added to 30 mg SRNOM and 84 mg NaHCO3 dissolved in 900 mL H2O was irradiated for 4 h using the Pyrex light filter (16 mg C L−1, 22 mg nitrite L−1, and 68 mg bicarbonate L−1). The initial and final pH readings were 7.4 and 7.1, respectively. The reaction was performed in quadruplicate, the combined reaction solutions were freeze dried, and the powder was analyzed by solid-state CP/MAS 15N NMR.

Control Irradiation of Nitrate

Nitrosation of Suwannee River Natural Organic Matter

Irradiation of Suwannee River Natural Organic Matter Alone

15

A solution of 85 mg of Na NO3 in 900 mL H2O was irradiated for 4 h, freeze dried, and redissolved in 2 mL 75% D2O for 15N NMR analysis. www.agronomy.org • www.crops.org • www.soils.org

Suwannee River NOM was reacted with Na15NO2 at high concentrations without pH adjustment and at dilute 867

concentrations with pH adjusted to 7.5. In the first reaction, 300 mg of SRNOM in 3 mL H2O was added to 86 mg Na15NO2 in 1 mL H2O and stirred for 3 d in the dark. The initial pH was approximately 3.5. The solution was passed through a 0.45-μm HA filter and freeze dried. After recording of an inverse-gated decoupled 15N NMR spectrum in 75% D2O, the sample was dialyzed against a 100 Da MWCO cellulose acetate membrane, H+–saturated on an MSC-1 cation exchange resin, freeze dried, and redissolved in DMSO-d6 for ACOUSTIC and DEPT 15N NMR spectra. In the second reaction, a solution of 324 mg SRNOM adjusted to pH 7.5 in a total volume of 900 mL H2O was charged with 65 mg Na15NO2, stirred for 26.5 h in the dark, freeze dried, and redissolved in 75% D2O for 15N NMR analysis.

Reaction of Suwannee River Natural Organic Matter with Nitrogen Dioxide To trap the gas as a liquid, 15NO2 was bubbled for 1 min into a two-necked flask containing 50 mL methanol (dried over 3A molecular sieves) cooled to 0°C in an ice bath. A solution of 300 mg of SRNOM (dried in a dessicator under vacuum) in 50 mL of dried methanol was added to the two-necked flask. After approximately 1 h of stirring in the ice bath under a gentle stream of nitrogen gas, the solution was allowed to come to room temperature and was then evaporated under an increased flow of nitrogen. The sample was dissolved in DMSO-d6 for 15N NMR analysis.

Reaction of Suwannee River Natural Organic Matter with Nitric Acid A 2.5-gram solution of 40% H15NO3 was added to a solution of 204 mg SRNOM dissolved in 2 mL H2O (final nitric acid concentration of 22.2%) and stirred for 48 h in a screw cap vial. The sample was then evaporated under a stream of nitrogen gas, dialyzed against a 100 Da MWCO cellulose acetate membrane, and freeze dried. The SRNOM took on a noticeable orange color from treatment with nitric acid. The sample was dissolved in DMSO-d6 for 15N NMR analysis.

NMR Spectrometry Solid-state CP/MAS 15N and 13C NMR spectra were recorded on a Chemagnetics CMX-200 spectrometer at 20.3 and 50.3 MHz, respectively, using a 7.5-mm ceramic probe (zirconium pencil rotors). Acquisition parameters for 15N included a 30.0 KHz spectral window, 17.051-ms acquisition time, 2.0- or 5.0-ms contact time, 0.2-s pulse delay, and spinning rate of 5.0 KHz. The 15N NMR chemical shifts are reported in ppm downfield from ammonia, taken as 0.0 ppm. Acquisition parameters for 13C included a 30.0 KHz spectral window, 17.051-ms acquisition time, 5-ms contact time, 0.5-s pulse delay, and spinning rate of 5.0 KHz. Liquid-state 15N NMR spectra were recorded on a VARIAN 300 MHz NMR spectrometer at 30.4 MHz using a 10 mm broadband probe. Inverse-gated decoupled 15N NMR spectra used a 35,111.7 Hz (1154.3 ppm) spectral window, 0.2-s acquisition time, and 2.0-s pulse delay. ACOUSTIC (Patt, 1982) 15N NMR spectra were recorded using a 35,111.7 Hz (1154.3 ppm) spectral window, 0.2-s acquisition time, 0.5-s pulse delay, and tau delay of 0.1-ms. ACOUSTIC spectra of samples 868

dissolved in DMSO-d6 were recorded with and without the paramagnetic relaxation reagent chromium (III) acetylacetonate, in the former case to shorten T1’s (spin lattice relaxation times) and quench nuclear Overhauser enhancement (NOE) effects. The DEPT 15N NMR spectra (distortionless enhancement by polarization transfer) were recorded with a 26,000-Hz spectral window, 0.2 s-acquisition time, 1.0-s delay for proton relaxation, and 1JNH of 90.0 Hz. Neat formamide (112.4 ppm) was used as an external reference standard. Single-pulse, liquid-state spectra of the dialyzed samples exhibited significant baseline roll due to acoustic ringing, and for this reason the ACOUSTIC sequence was used.

Results Solid-State Cross Polarization/Magic Angle Spinning Natural Abundance 15N and 13C NMR Spectra of Suwannee River Natural Organic Matter before and after Irradiation The 15N NMR spectra of the naturally abundant nitrogens in SRNOM were recorded before and after irradiation with the unfiltered medium pressure mercury lamp to identify potentially reactive sites toward nitrosation and nitration and to determine how the naturally abundant nitrogen is photochemically degraded as a control for the irradiation experiments performed in the presence of labeled nitrate and nitrite (Fig. 1A, 1B). The low concentration of naturally abundant nitrogen in aquatic NOM (1.1% in SRNOM) makes detection of 15N by CP/MAS in the solid state a challenge because of the large number of transients required; dilution of the sample in water or organic solvent imposes a greater challenge for detection in the liquid state, even with polarization transfer or indirect detection techniques. Furthermore, limitations regarding the quantitative accuracy of the CP/MAS experiment with respect to naturally abundant nitrogen in NOM need to be considered. Potential problems affecting quantitation or detection have been documented, and these include the underestimation of nitrogens not directly bonded to protons, the large chemical shift anisotropy of pyridine-like nitrogens, and chemical exchange or molecular motion phenomena associated with enamino-imino systems such as porphyrins and phthalocyanins (Earl, 1987; Kelemen et al., 2002; Smernik and Baldock, 2005; Thorn and Cox, 2009). With these caveats as background, the natural abundance spectrum of the SRNOM is seen to consist of a broad band of resonances from approximately 0 to 230 ppm, with peak maxima at 34, 122, 133, 171, and 194 ppm (Fig. 1A). The assignments are: 34 ppm, amino sugar and amino acid free amine groups; 122 ppm, aminoquinones (amino acid-quinones such as glycylaminoquinone) and amides, including secondary amide nitrogens of amino acids involved in peptide linkages; 133 ppm, amide and heterocyclic nitrogens; 171 and 194 ppm, heterocyclic nitrogen. Amino acid analyses have not been reported for SRNOM, so it is not possible to determine what percentage of the amide/aminoquinone peak at 122 ppm is comprised of peptide or amino acid-quinone linkages and what the relative breakdown of the aminosugar/amino acid free amine peak at 34 ppm is. Reported amino acid analyses for the IHSS Suwannee River fulvic acid (Standard I) and reference Suwannee River humic acid, including the nitrogen-containing side-chain Journal of Environmental Quality

Fig. 1. Solid-state cross polarization/magic angle spinning (CP/MAS) 15N and 13C nuclear magnetic resonance (NMR) spectra of Suwannee River natural organic matter (SRNOM) before and after irradiation to 62%IA465nm. Ct, contact time; LB, line broadening.

amino acids arginine and histidine, account for 6.8 and 14.2% of the total nitrogen, respectively, in these samples (Thorn and Cox, 2009). Plausible structural configurations for amino acids in NOM are short-chain peptides covalently bonded to NOM molecules through aminoquinone or glycosidic linkages (Thorn and Mikita, 2000). A variety of nonpeptide structures may contribute to the peaks at 122 and 133 ppm, including N-acetyl and anilide nitrogens (Schmidt-Rohr et al., 2004), whereas heterocyclic nitrogens, such as indoles and pyrroles, occur in the region from approximately 130 to 160 ppm. In other words, amide, aminoquinone, and heterocyclic nitrogens are not completely resolved from one another. Structures that would correlate to the upfield shoulder of the peak at 87 ppm include primary amines of purine and pyrimidine bases such as adenine, guanine, and cytidine, other aromatic amines, and N-glycosyl nitrogens. The amino sugar/amino acid free amine peak at 34 ppm was not observed in the spectrum of SRNOM that we reported previously (Thorn, 2002; Thorn and Cox, 2009), which was recorded on an unfiltered sample. Here, the combination of a reduction in ash content through filtering and a longer accumulation time enabled detection of this peak as well as an overall improvement in signal to noise ratio. A compilation of 15N NMR chemical shifts relevant to the natural abundance spectra and to spectra of reacted samples is presented in Supplemental Fig. S1. Without the use of an internal intensity standard for spin counting experiments, a direct measurement of the loss of www.agronomy.org • www.crops.org • www.soils.org

nitrogen peak intensities as a result of photodegradation of the sample to 62%IA465nm is not possible (Fig. 1B). Comparison of the 15N NMR spectra reveals a loss of the heterocyclic nitrogen peak at 171 ppm and a shift of the peak maximum from 122 to 137 ppm on irradiation. This change in the chemical shift position of the peak maximum may indicate a loss of amide and aminoquinone nitrogens with respect to heterocyclic nitrogens; however, the chemical shifts of these nitrogen functionalities are not sufficiently resolved from one another to definitively make this conclusion. Photochemical rearrangements unassociated with nitrogen losses may also be taken into consideration. When viewed with a narrower line broadening of 1.0 Hz, the spectrum of the irradiated sample showed no evidence of discreet peaks corresponding to ammonium or amino acids. If formed, these could be at concentrations below the detection limit of the NMR experiment, or alternatively, could have recondensed with the NOM (Tarr et al., 2001; Thorn and Mikita, 1992; Thorn and Cox, 2009). Although condensation of ammonia with NOM is favored at alkaline pH (Thorn and Mikita, 1992), we have observed fixation reactions down to pH 4 (Thorn, unpublished results). The solid state CP/MAS 13C NMR spectra of the SRNOM (Fig. 1C, 1D) indicate a loss of O-alkyl (66 ppm), anomeric (106 ppm), aromatic (106, 115, 129, 155 ppm), and carboxyl/ amide/ester carbons (174 ppm) with respect to C-alkyl carbons (39 ppm) as a result of the photodegradation, consistent with previous reports (Osburn et al., 2001; Thorn et al., 2010). 869

Quinones occur in the chemical shift region from approximately 190 to 169 ppm and thus overlap with the ketone and carboxyl/ amide/ester carbon peaks. A separate experiment in which SRNOM was similarly irradiated indicated loss of quinones, via oximation analysis (Supplemental Figs. S3–S5). A loss of quinones is assumed in the present sample. The loss of aromatic carbons is consistent with the loss of heterocyclic nitrogens at 171 ppm observed in the 15N NMR spectra. The decrease in the carboxyl/amide/ester carbon peak would also be consistent with the possible loss of amide and aminoquinone nitrogens suggested by the 15N NMR spectra. The mechanisms for the reported photochemical release of ammonium and amino acids from aquatic NOM are not fully understood. The results presented here indicate a role for 15N NMR in determining the reaction pathways, especially at higher

field. In the 15N NMR spectra of the SRNOM reacted with 15N labeled reagents that follow, in both the liquid and solid state, the naturally abundant nitrogens of the NOM are not observed under the acquisition parameters used.

Nitrosation of Suwannee River Natural Organic Matter with Na15NO2 Liquid-state 15N NMR spectra of SRNOM reacted with Na NO2 under concentrations and pH (3.5) that favor nitrosation of the NOM are shown in Fig. 2. The inverse-gated decoupled spectrum of the sample (Fig. 2A) before dialysis exhibits a nitrate peak at 375.6 ppm and ammonium peak at 20.8 ppm (Table 1), in addition to peaks corresponding to the organic reaction products. A peak corresponding to nitrogen gas is also visible at 308.3 ppm. Nitrite is stable at pH approximately 7 or higher (Cai et al., 2001) but undergoes disproportionation to nitric oxide and nitrate at acidic pH, which is the fate of the excess nitrite that does not react with the SRNOM via nitrosation or nitration: 15

3HNO2 → HNO3 + 2 ∙NO + H2O

Fig. 2. Liquid-state 15N nuclear magnetic resonance (NMR) spectra of Suwannee River natural organic matter (SRNOM) reacted with Na15NO2 at pH 3.5. (A) Inverse-gated decoupled spectrum of sample before dialysis recorded in 75% D2O. (B) ACOUSTIC spectrum of sample after dialysis recorded in DMSO-d6. (C) Distortionless enhancement by polarization transfer (DEPT) spectrum of sample after dialysis recorded in DMSO-d6; all nitrogens bonded to protons in positive phase. LB, line broadening. 870

[R8]

This reaction accounts both for the absence of residual nitrite and the presence of nitrate in the spectrum. The peak at 266.5 ppm may correspond to cyanide. Cyanide formation has been associated with nitrosation of aromatic compounds at pH values ranging from 2 to 4, although a reaction mechanism has not been definitively established (Zheng et al., 2004). The ACOUSTIC spectrum (Fig. 2C) was recorded on the sample after it was dialyzed to decrease the concentration of nitrate, which overlaps with a signal of main interest. The description of this spectrum is based on our previously reported analyses of NOM samples reacted with sodium nitrite (Thorn and Mikita, 2000) and more recent literature. Reaction of the sample with nitrite has resulted in a twofold increase in its nitrogen content and in a decrease in its C/N ratio from 47.7 to 20.7 (Table 2). Peaks resulting directly from nitrosation include N-nitrosamines or N-nitrosamides (565–530 ppm; from nitrosation of naturally abundant 2° amines or 2° amides), nitrosophenols (i.e., quinone monoximes; 430–390 ppm; from nitrosation of phenolic carbons), and ketoximes and aldoximes (390–340 ppm; from nitrosation of activated methylene and methyl carbons, respectively) (Fig. 3). Possible substrate sites for N-nitrosation include indole nitrogens (Kirsch and de Groot, 2009) and peptide nitrogens (Bonnett et al., 1975; Challis et al., 1994), evidence for which is provided by the natural abundance spectrum (Fig. 1A). Tertiary amines, such as the pyrrolidine nitrogen of nicotine, are also known to undergo N-nitrosation (Hecht et al., 1978; Sleiman et al., 2010). Although primary amines and amides tend to undergo deamination to N2 via diazonium intermediates (Fig. 3), there are examples of primary aromatic amines that undergo nitrosation to form heterocyclic structures containing pyrazole and triazole Journal of Environmental Quality

type nitrogens, whose 15N NMR chemical shifts overlap with other peaks in the spectrum (Supplemental Fig. S1). For instance, 2-methyl-4-nitroaniline undergoes nitrosation to 5-nitro-1Hindazole (Fieser, 1967) and 2,3-diaminonaphthalene undergoes nitrosation to 2,3-naphthotriazole (Kono et al., 1995). The reaction also of nitric oxide with aromatic ortho diamines to form flourescent benzotriazole structures has been exploited as a probe for nitric oxide detection. The 15N NMR chemical shifts of oximes and nitro groups overlap in the approximate range from 380 to 320 ppm (Fig. 4). Therefore, the occurrence of nitrogen, oxygen, or aromatic carbon nitration reactions cannot be ruled out. Formation of nitro groups could occur via several pathways, including oxidation of initially formed aromatic nitroso compounds in the presence of excess nitrous acid, nitration reactions initiated by the production of nitric oxide or nitric acid from disproportionation of nitrous acid, and nitric oxide formed via reduction of nitrous acid by constituents in NOM. Ortho-diphenols, such as quercetin and chlorogenic acid, have the capacity to reduce nitrous acid to nitric oxide (Takahama, 2008); these are plausible structural units in NOM. The remaining peaks in the spectrum of Fig. 2B upfield of approximately 350 ppm can most plausibly be assigned as products from Beckmann reactions of the initially formed oximes (Gawley, 1987; Thorn et al., 1992) or condensation reactions with ammonia and possibly hydroxylamine. One of the most significant secondary reaction products is the nitrile peak from about 235 to 285 ppm. Nitriles arise from Beckmann fragmentations of the initially formed oximes or quinone monooximes (nitrosophenols) (Fig. 3); these oximes are likely to undergo fragmentation to nitriles are those adjacent to quaternary carbon centers. The broad set of resonances from about 70 to 225 ppm in the ACOUSTIC spectrum (Fig. 2B) is resolved into primary amide (90–120 ppm), lactam and secondary amides (120–180 ppm), and nonprotonated nitrogens (180–230 ppm) by comparison with the DEPT spectrum in Fig. 2C. Primary amides also arise from Beckmann fragmentations of ketoximes, and Beckmann rearrangements of aldoximes. Lactams and secondary amides originate from Beckmann rearrangements of ketoximes (Fig. 3). Assignments for the nonprotonated nitrogens from 180 to 230 ppm may

Table 1. 15N nuclear magnetic resonance chemical shifts for nitrogen compounds. Compound

15

N nuclear magnetic resonance shift ppm 20.7† 27.3† 77.6‡

NH4NO3 NH4Cl K+NCO− NH2OH.HCl pH 3.0 pH 4.9 pH 5.9 pH 6.4 pH 7.0 pH 8.0 N2O (NO) (N=) NaCN N2 NaNO3 NH4NO3 HNO3 NaNO2 HNO2 ONOO−

81.0§ 82.0 89.7 96.6 100.4 104.1 232.6‡ 144.4 274.5§ 306.2‡ 373.9§ 376.3† 375.8‡ 607.1§ 560.7‡ 554.6¶

† Levy and Lichter (1979). ‡ Witanowski et al., (1993). § Thorn (determined in this lab). ¶ Butler et al. (1997).

include imidazole, pyrazole, isocyanide, imidate, and amidine. A summary of assignments is presented in Table 3. In view of the uncertainty over the relative importance of nitrosation versus nitration reactions because of the chemical shift overlap of oxime and nitro groups, the N-nitrosamine/nitrosamide, nitrosophenol, and nitrile nitrogens may be considered signature peaks for the occurrence of nitrosation. The mechanism of ammonia formation during the course of reaction with sodium nitrite is uncertain. Ammonium could arise from direct reduction of nitrite or hydroxylamine, mediated by hydroquinone and catechol moieties in the NOM, with assistance

Table 2. Elemental analyses of reacted Suwannee River natural organic matter. SRNOM† Carbon Hydrogen Nitrogen Oxygen Sulfur Phosphorus NR/N0# C/N

SRNOM HNO3

SRNOM NaNO3/UV‡

SRNOM NaNO2§

————————————————————— % (w/w)¶ ————————————————————— 52.47 42.90 49.37 46.14 4.19 3.97 4.71 4.41 1.10 2.86 2.91 2.23 42.69 49.71 42.17 46.68 0.65 0.54 0.79 0.52 0.02 0.02 0.05 0.02 2.6 2.6 2.0 47.7 15.00 17.0 20.7

† Suwannee River natural organic matter. ‡ Corresponds to pH 3.2 reaction (see Fig. 6A and 6B). § Corresponds to pH 3.5 reaction (see Fig. 2). ¶ Reported on a dry, ash free basis. # %N of reacted sample/%N of unreacted sample. www.agronomy.org • www.crops.org • www.soils.org

871

Fig. 3. Nitrosation and related reactions.

from trace metals comprising the residual ash content or from decomposition of hydroxylamine (Segal and Wilson, 1949). 3NH2OH → N2 + NH3 + 3H2O

[R9]

Hydroxylamine itself is not detectable in Fig. 2A but could be a transitory species. Ammonium could also result from hydrolysis of secondary reaction products, such as the primary amides. Previous studies have indicated that the reaction products from condensation of ammonia with the carbonyl groups of 872

NOM occur in the region from approximately 30 to 350 ppm (Thorn and Mikita, 1992). The 15N NMR spectra of ammonia fixated NOM samples show major peaks corresponding to aminohydroquinone (~40 ppm), primary amide (~112 ppm), indole (~135 ppm), pyrrole (~160 ppm), heterocyclic (~170 ppm), imidazole, pyrazole (~253 ppm), and pyridine and pyrazine (~314 ppm) nitrogens (Thorn and Mikita, 1992; Thorn and Mikita, 2000). Spectra illustrating the reaction of ammonia and hydroxylamine with SRNOM are reported in Journal of Environmental Quality

Supplemental Figs. S2 through S5. Because the 15N NMR chemical shifts of the reaction products from condensation of ammonia and hydroxylamine with carbonyl groups and from Beckmann reactions of oximes overlap, quantitation of the relative importance of the two sets of reactions is not possible. For the remainder of the discussion, the Beckmann and condensation reactions are referred to as secondary reactions. The N2 gas observed in Fig. 2A at 308.3 ppm could arise from decomposition of diazonium intermediates resulting from nitrosation of primary amines or amides or from decomposition of hydroxylamine [R9].

Irradiation of Suwannee River Natural Organic Matter in the Presence of Na15NO3 Liquid-state 15N NMR spectra of a control irradiation of Na15NO3 and the product mixtures from SRNOM irradiated with Na15NO3 at initial pH values of 3.2 and 8.0 are shown in Fig. 5. In the control reaction (Fig. 5A), after 4 h of irradiation in deionized water, approximately 60% of the nitrate (peak at 373.9 ppm) not lost through volatilization has been converted to nitrite (peak at Fig. 4. Nitrogen-15 nuclear magnetic resonance chemical shift ranges for oxime, nitro, and 607.1 ppm), as determined from electronic nitroso nitrogens. integration of the spectrum (Fig. 5A). No other peaks are visible in the spectrum. Use These would include photolysis of the nitrosation, nitration, and of a Pyrex filter would prevent photolysis of the nitrate to nitrite secondary reaction products, photolysis of hydroxylamine (Behar by blocking irradiation below 290 to 300 nm. The spectrum of et al., 1972), etc. The ammonium observed in the spectra of Fig. SRNOM irradiated with Na15NO3 at pH 3.2 (Fig. 5B) shows 5B and 5C arises from the labeled nitrate. Ammonium resulting discreet peaks corresponding to ammonium (21.7 ppm) and from photolysis of the naturally abundant nitrogen in the NOM residual nitrate (375.6 ppm), while the spectrum of the sample would overlap the nitrate derived ammonium but would remain irradiated at pH 8.0 (Fig. 5C) shows ammonium, nitrate and buried under its intensity, according to the acquisition parameters nitrite, the latter stable at the final pH of the reaction. A trace of the liquid-state NMR analyses. Whatever the mechanism, N2 gas peak becomes evident in the pH 3.2 irradiation spectrum NOM can be viewed as a catalyst for photoreduction of nitrate to (Fig. 5B) when a narrower line broadening is applied to the ammonium under the conditions of these reactions. free induction decay (FID). None of the spectra contains peaks The ACOUSTIC spectrum (NOE eliminated) of the downfield of the nitrite peak at 607.7 ppm. Photodegradation dialyzed SRNOM/Na15NO3/UV/pH 3.2 sample (Fig. 6B; reactions may be responsible for ammonium formation in the solid-state spectrum in Fig. 7C) contains a distribution of peaks irradiation experiments in addition to the possible mechanisms similar to the spectrum of the directly nitrosated sample (Fig. discussed in conjunction with the direct nitrosation of SRNOM. 2B). The concentration of nitrogen in the sample increased Table 3. Assignments for 15N nuclear magnetic resonance spectra of samples reacted with 15N-labeled reagents. Chemical shift range

Assignment

ppm 565–530 430–390 390–340 340–280

N-nitrosamide; N-nitrosamine nitrosophenol (quinone monoxime); furazan ketoxime; nitro; aldoxime; benzotriazole pyridine; indazole; azoxy; indophenol; phenoxazinone; imine; isoquinoline; dihydroisoquinoline; 1Δ-pyrroline

280–230 230–180 180–150 150–120 120–80

nitrile; benzotriazole; oxazole; pyrazole imidazole; pyrazole; imidate; amidine; isocyanide; imide singly protonated nitrogen; hydroxamic acid; indazole; imide; pyrrole singly protonated nitrogen; 2° amide; lactam; indole doubly protonated nitrogen; 1° amide; aminoquinone

www.agronomy.org • www.crops.org • www.soils.org

873

Fig. 5. Liquid-state inverse-gated decoupled 15N nuclear magnetic resonance spectra. (A) Control irradiation of Na15NO3. (B) Irradiation of Suwannee River natural organic matter (SRNOM) with Na15NO3 at initial pH of 3.5, before dialysis. (C) Irradiation of SRNOM with Na15NO3 at initial pH of 8.0, before dialysis. Solvent = 75% D2O. LB, line broadening.

by a factor of 2.6, whereas the C/N ratio decreased from 47.7 to 17.0 (Table 2). The peaks include, most importantly, the nitrosophenol (quinone monoxime) nitrogens at 391.3 ppm and the oxime/nitro nitrogens at 364.2 ppm. The N-nitrosamide/Nnitrosamine nitrogens observed at 590 to 530 ppm in Fig. 2B, however, are not present in Fig. 6B. This may be explained by photochemical transformation or degradation of the naturally abundant nitrogens in SRNOM that act as the substrate sites for N-nitrosation or instability of N-nitroso groups toward irradiation. The spectrum of the irradiated sample also contains nitrile nitrogens at 247.4 and 260.1 ppm and the broad resonance of other secondary reaction products from about 70 to 225 ppm (Fig. 6B). This latter peak is again resolved into nitrogens bonded to two protons (primary amides, 108 ppm), nitrogens bonded to one proton (secondary amide, lactam, and heterocyclic; 133.6 and 164.0 ppm), and nonprotonated nitrogens (~175 to 225 ppm) by comparison with the DEPT spectrum of Fig. 6A. The peak from approximately 280 to 340 ppm, with maximum at 308.4 ppm, was also observed in NOM samples directly nitrosated with Na15NO2 at pH 6 (Thorn and Mikita, 2000). This chemical shift region encompasses a number of nitrogen functional groups, including pyridine, pyrazine, imine, isoquinoline, dihydroisoquinoline, 1 Δ-pyrroline, and indophenol (Table 3). 874

Fig. 6. Liquid-state 15N nuclear magnetic resonance (NMR) spectra. (A) Distortionless enhancement by polarization transfer (DEPT) spectrum, Suwannee River natural organic matter (SRNOM) irradiated with Na15NO3 at initial pH of 3.5, after dialysis, recorded in DMSO-d6; nitrogens bonded to two protons in negative phase and one proton in positive phase. (B) ACOUSTIC spectrum, SRNOM irradiated with Na15NO3 at initial pH of 3.5, after dialysis, recorded in DMSO-d6. (C) ACOUSTIC spectrum, SRNOM irradiated with Na15NO3 at initial pH of 8.0, after dialysis, recorded in DMSO-d6. LB, line broadening.

Photochemical transformations not operable under conditions of direct nitrosation with NaNO2 may contribute to the secondary reaction products visible in Fig. 6. Aromatic nitro compounds, such as nitrobenzene and TNT (2,4,6-trinitrotoluene), are known to undergo photochemical transformation to aromatic amine, amide, lactam, nitrile, and azoxybenzene compounds, among others (Pennington et al., 2007; Thorn, unpublished results). If aromatic nitration of SRNOM occurs on irradiation in the presence of nitrate, it is conceivable that the nitro groups could undergo subsequent transformation to these other functional groups. Photochemical Beckmann rearrangements (e.g., conversion of cyclic ketoximes to lactams, aryl aldoximes to amides, etc.) would be another example (Gawley, 1987). The ACOUSTIC spectrum (NOE retained) of SRNOM irradiated in the presence of Na15NO3 at initial pH of 8.0 (Fig. 6C) is almost identical to the spectrum from the pH 3.2 experiment, encompassing the same organic nitrogen functional groups in the range from approximately 70 to 430 ppm. The presence of the nitrosophenol (394.6, 417.8 ppm) and nitrile (266.6 ppm) Journal of Environmental Quality

peaks confirms the occurrence of nitrosation reactions. Thus, UV irradiation is seen to effect the incorporation of nitrate nitrogen into SRNOM over a range of pH values and in similar structural forms. The amount of nitrogen incorporation at pH 8 is surprising. The dark control reaction of SRNOM with Na15NO2 at pH 7.5 resulted in essentially negligible incorporation of nitrogen (vide infra). This suggests that in the condensation of nitrate nitrogen with the NOM during irradiation at pH 8, reactive species such as nitrogen dioxide or nitric oxide may be more significant than nitrous acid. The NOE of the primary amide peak at 108.6 ppm was incompletely quenched by the addition of a paramagnetic relaxation reagent (Fig. 6B), resulting in an attenuation of the peak intensity. Therefore, the ACOUSTIC spectrum of the pH 8.0 reaction recorded without PRA is presented. The inversion of peaks in Fig. 6C is due to negative NOE.

Irradiation of Suwannee River Natural Organic Matter in the Presence of Na15NO3 and Bicarbonate To establish the phenomenon of nitrogen incorporation under conditions that more accurately represent water treatment conditions, SRNOM was irradiated at pH 7.5 in the presence of nitrate and bicarbonate at concentrations more realistic for all three constituents (16 mg C L−1 of SRNOM, 24 mg NO3− L−1, and 68 mg HCO3− L−1). In previous nitrate photolysis studies, bicarbonate was shown to decrease nitrite yields at pH 8 but not at pH 6, presumably because carbon dioxide reacts with the peroxynitrite intermediate in nitrite formation (Sharpless and Linden, 2001). The pH dependence was explained by the fact that it is peroxynitrite (pKb = 7.2) and not peroxynitrous acid, which reacts with carbon dioxide (Sharpless and Linden, 2001). ONOO− + CO2 → ONOOCO2−

[R10]

In this experiment, bicarbonate produced from photodegradation of SRNOM and possibly from nitrosative decarboxylation of the NOM would add to the starting concentration of bicarbonate. The ACOUSTIC spectrum was recorded without decoupling to prevent sample heating from the high salt content, thus the relative decrease in resolution and signal to noise ratio (Fig. 7A). The spectrum exhibits ammonium, nitrate, and nitrite peaks and once again confirms incorporation of nitrogen into the SRNOM by virtue of the presence of the organic nitrogen peaks from approximately 40 to 420 ppm. Qualitatively, the same resonances are present as in Fig. 6B and 6C. Amides and lactams are the peaks of major intensity. Under the reaction conditions of this irradiation, however, formation of the nitrosophenol and oxime/nitro nitrogens with respect to other nitrogens upfield of about 350 ppm is diminished compared with the previous two irradiations. This is more clearly apparent by comparing the solidstate CP/MAS 15N NMR spectrum of the dialyzed sample with the spectrum of the SRNOM irradiated with Na15NO3 at pH 3.2 in the absence of the bicarbonate, also recorded after dialysis (Fig. 7B and 7C). Determination of whether the change in distribution of nitrogen condensation products is an effect of the bicarbonate, differences in concentrations of reactants, or both, requires further analysis. The consideration of underestimation of nitrogens not directly bonded to protons with respect to nitrogens bonded to protons applies to these solid state CP/MAS spectra. www.agronomy.org • www.crops.org • www.soils.org

Fig. 7. Nitrogen-15 nuclear magnetic resonance (NMR) spectra. (A) Liquid-state ACOUSTIC (no decoupling) 15N NMR spectrum of Suwannee River natural organic matter (SRNOM) irradiated with Na15NO3 in the presence of bicarbonate at initial pH of 7.5, recorded in 75% D2O before dialysis. (B) Solid-state state cross polarization/ magic angle spinning (CP/MAS) 15N spectrum of dialyzed sample. (C) Solid-state CP/MAS 15N spectrum of SRNOM irradiated with Na15NO3 at initial pH of 3.5, after dialysis. LB, line broadening.

Photochemical Lability of Incorporated Nitrogen The sample of SRNOM irradiated with nitrate at pH 3.2 was further dialyzed to remove the remaining nitrate and ammonium and subjected to an additional 2 h of irradiation with the unfiltered medium pressure mercury lamp to determine the 875

susceptibility of the incorporated nitrogen to photodegradation. The ACOUSTIC (NOE retained) 15N NMR spectra indicate that the nitrosophenol, oxime, and nitro groups were degraded, whereas nitrogens upfield of about ~330 ppm survived the irradiation (Fig. 8). In other words, amide, lactam, nitrile, and pyridine nitrogens and other unidentified nitrogens in the region from 70 to 330 ppm were resistant to photodegradation for the duration of the irradiation. Similar results were obtained with the sample of SRNOM directly nitrosated with NaNO2 at pH 3.5 and then irradiated for 2 h (Supplemental Fig. S6). 15

N NMR Spectra of Suwannee River Natural Organic Matter Reacted with Nitrogen Dioxide and Nitric Acid

Suwannee River natural organic matter was reacted separately with 15N-labeled nitrogen dioxide and nitric acid to generate 15 N NMR spectra that would serve as controls for free radical and electrophilic nitration, respectively. It was anticipated that a comparison of the control spectra with spectra of SRNOM reacted with sodium nitrite and SRNOM irradiated in the presence of sodium nitrate could provide insight into the extent of nitration in the latter reactions. Reaction of SRNOM with NO2 was performed in methanol rather than water to prevent hydrolysis:

Fig. 8. Liquid-state ACOUSTIC 15N nuclear magnetic resonance (NMR) spectra. (A) Suwannee River natural organic matter (SRNOM) irradiated with Na15NO3 at initial pH of 3.5, after dialysis, recorded in DMSO-d6. (B) After second irradiation. 876

2 ∙NO2 + H2O → HNO3 + HNO2

[R11]

The inverse-gated decoupled spectrum exhibits two peaks at 369.6 and 365.1 ppm, presumed to be comprised mainly of nitro groups, a nitrosophenol shoulder from 430 to 390 ppm, residual nitric acid peak at 376.9 ppm, and ammonium at 22.7 ppm (Fig. 9). Additionally, the nitrogen peaks from about 70 to 285 ppm present in the spectra of Fig. 4, 5, and 6, and attributable to the secondary Beckmann or condensation reactions products, are observed, but at very low signal intensities. A peak of low intensity corresponding to pyridine-like nitrogen is also observed at 308.4 ppm. The spectrum of SRNOM treated with nitric acid displays nitro peaks at 368.9, 367.5, and 364.9 ppm, a nitrosophenol shoulder from 430 to 390 ppm, a peak at 338.9 ppm, residual nitric acid peak at 376.6 ppm, and an ammonium peak at 22.7 ppm (Fig. 10A and 10B). It also contains the peaks associated with the secondary Beckmann and condensation reactions. These peaks, visible at 106.2, 165.5, 182.8, 211.0, and 255.1 ppm, are again of low intensity but are more pronounced than in the spectrum of the SRNOM reacted with the nitrogen dioxide. Polarization transfer enhances the signal intensity of the amide and lactam peaks at 106.2, 131.6, and 165.5 ppm, as evident in the DEPT spectrum (Fig. 10C). Under the strongly acidic conditions of the nitric acid treatment, hydrolysis of secondary reaction products, such as the amides, is a likely contributing

Fig. 9. Liquid-state inverse-gated decoupled 15N nuclear magnetic resonance (NMR) spectra of Suwannee River natural organic matter (SRNOM) reacted with 15NO2; recorded in DMSO-d6. (A) Full plot. (B) Horizontal expansion. LB, line broadening. Journal of Environmental Quality

mechanism for production of the ammonia. Treatment of the SRNOM with nitric acid increased its nitrogen content by a factor of 2.6 and decreased its C/N ratio from 47.7 to 15.0 (Table 2). The occurrence of nitrosation reactions during treatment with nitric acid is most likely explained by the NOM catalyzed reduction of nitric acid to nitrous acid. In the case of treatment with nitrogen dioxide, we speculate that nitrosation reactions occurred as a result of two possibilities, the first being that the NO2 hydrolyzed to nitrous acid, from trace moisture in the NOM or methanol, despite precautions taken to exclude water. If this is the case, nitric acid likewise formed during hydrolysis could also be the active nitrating species, rather than the starting nitrogen dioxide, a possibility supported by the overall similarity between the two sets of spectra. The alternative explanation is that the NOM catalyzed reduction of NO2 to nitrous acid. In this case, the implication would be that the same spectra result from reaction of SRNOM with nitric acid and nitrogen dioxide, although the mechanisms differ. The peaks from 350 to 390 ppm in both sets of spectra are interpreted as being comprised mainly of nitro groups, with the possibility of some contribution from oxime nitrogens. The results of these reactions between nitrogen dioxide and nitric acid with SRNOM were reproduced with Suwannee River fulvic acid. The results of the nitric acid treatment were likewise reproduced with the IHSS Pahokee Peat fulvic acid, but with the occurrence of nitrogen-nitrosation (Thorn, unpublished results). These results illustrate the experimental difficulty in generating control spectra that show nitration versus nitrosation and reflect model compound studies where both nitration and nitrosation reactions occur with nitrating or nitrosating reagents. The large ratio of intensity of the nitro peaks with respect to the peaks representing secondary reaction products appears to be a characteristic of the spectra corresponding primarily to nitration from reaction with nitric acid or nitrogen dioxide when compared with the spectra of the samples irradiated with nitrate or directly reacted with nitrite. The reaction of SRNOM acid with 15N-labeled nitric oxide in methanol under a N2 atmosphere was also examined, but reproducible results have not been obtained as of this writing. At least one previous report inferred nitrosation of humic acid during treatment with nitric acid. Green and Manahan (1979) investigated the nitrohumic acids produced from nitric acid oxidation of bituminous coal. Their nitrohumic acids showed polarographic reduction waves attributable to reduction of nitroso groups in addition to nitro groups, the latter to a significant extent occurring on aromatic rings substituted with hydroxyl groups. Although an IR spectrum of the nitrohumic acid exhibited –NO2 stretching bands at 1540 and 1350 cm−1, the authors noted the potential contribution to the 1540 cm−1 band by nitroso groups, which absorb in the 1500 to 1600 cm−1 region. Nitrosation was hypothesized to occur through formation of nitrous acid or through autoreduction of nitro groups by the humic acid. We observed very weak peaks at 1549 and 1381 cm−1 in an FTIR spectrum of the nitric acid–treated SRNOM but were unable to find evidence for nitration or nitrosation in FTIR spectra of the directly nitrosated (pH 3.5) or nitrate-irradiated (pH 3.2) SRNOM, an indication of the limits of IR for determination of these reactions (Thorn, unpublished results). www.agronomy.org • www.crops.org • www.soils.org

15

N NMR Spectra of Suwannee River Natural Organic Matter Irradiated in the Presence of Na15NO2

The final set of experiments examined the ability of UV irradiation to effect incorporation of nitrite nitrogen into aquatic NOM. Liquid-state inverse-gated decoupled 15N NMR spectra of a control reaction of SRNOM reacted with Na15NO2 in the dark at pH 7.5 for 26.5 h versus SRNOM irradiated in the presence of Na15NO2 at initial pH of 7.5 for 4.0 h are compared in Fig. 11. There is no discernible evidence for nitrosation at pH 7.5 in the spectrum of the control reaction (Fig. 11A). Only the residual nitrite and a relatively minor nitrate peak are clearly visible; a trace nitrogen gas peak becomes evident when a narrower line broadening was used. When the spectrum of the dialyzed sample was re-recorded, a very weak peak became visible in the oxime/nitrosophenol region. Thus, the extent of nitrogen incorporation into the SRNOM in the dark reaction can be considered negligible. Irradiation unequivocally results in the incorporation of nitrite nitrogen into the SRNOM (Fig. 11B). The nitrite has been completely consumed or degraded, as indicated by its absence from the spectrum. An ammonium peak is present at 23.3 ppm, demonstrating that NOM can catalyze

Fig. 10. Liquid-state 15N nuclear magnetic resonance (NMR) spectra of Suwannee River natural organic matter (SRNOM) reacted with H15NO3; recorded in DMSO-d6. (A) Inverse-gated decoupled spectrum, full plot. (B) Inverse-gated decoupled spectrum, horizontal expansion. (C) Distortionless enhancement by polarization transfer (DEPT) spectrum; all nitrogens bonded to protons in positive phase. LB, line broadening. 877

the photoreduction of nitrite to ammonium. The spectrum exhibits the major peaks upfield of approximately 330 ppm that are present in the spectra of SRNOM irradiated in the presence of nitrate: amide, lactam, and nonprotonated nitrogens in the broad peak from 70 to 225 ppm; nitrile nitrogens from 225 to 275 ppm; and pyridine nitrogens from 290 to 330 ppm. The major difference from the nitrate-irradiated spectra is the absence of the nitrosophenol, oxime, and nitro groups. Peaks at 44.5 and 77.7 ppm possibly correlate to hydrazine and cyanate, respectively. Because nitrosation of the NOM is negligible at pH 7.5 in the dark control reaction, this pH was chosen for execution of the photolysis reaction. At lower pH, the occurrence of photolytic and nonphotolytic reactions would preclude a definitive resolution between the two processes. To account for the spectrum of Fig. 11B, the following scheme is proposed, based in part on the observation that nitrosophenol, oxime, and nitro groups are photolabile. Photolysis of nitrite leads to formation of nitrogen dioxide, which hydrolyses to nitrate, which in turn undergoes photolysis back to nitrite. During this cycling, nitrosophenol, oxime, and nitro groups are formed in the NOM, which subsequently undergo secondary reactions to form amide, lactam, nitrile, and pyridine nitrogens. Residual nitrosophenol, oxime, and nitro groups are photodegraded. The use of a Pyrex filter cuts off radiation below approximately 290 to 300 nm; therefore, the results shown in Fig. 12 may

be representative of reactions that can take place in surface waters exposed to natural sunlight. Here, because of the lower concentrations of SRNOM and nitrate used in conjunction with bicarbonate, a solid-state CP/MAS spectrum was recorded to detect reaction products formed after 4 h irradiation at initial pH of 7.4. The spectrum shows a broad peak from approximately 60 to 205 ppm, with a maximum at 108.7 ppm, indicating that nitrite is incorporated into the NOM significantly in the form of amide and lactam nitrogen. The downfield region of the peak from approximately 135 to 205 ppm likely encompasses heterocyclic nitrogens. In contrast to the previous reaction, the nitrite (broad peak at 610.6 ppm) and its degradation product nitrate (376.0 ppm) have not been completely consumed. If this process occurs naturally in sunlit surface waters, then it would constitute a previously unrecognized pathway for the formation of amide nitrogens in aquatic NOM as well as the other nitrogens comprising the peak from 60 to 205 ppm. In this regard, it is interesting to note that photolysis of nitrite in marine surface waters has been investigated as a process contributing to tropospheric nitric oxide, which plays a role in regulating the concentrations of ozone and hydroxyl radical (Olasehinde et al., 2010). The results of our experiments indicate that mass balance studies of nitrite photolysis in surface waters should take into account the possibility for NOM to act as a sink for the photolysis products.

Discussion This study is the first to document that UV irradiation results in the incorporation of nitrate and nitrite nitrogen into aquatic NOM. Condensation of nitrate photolysis products with NOM occurred over the pH range from 3.2 to 8.0 and in one experiment at concentrations of NOM, nitrate, and bicarbonate that are realistic for water treatment conditions. Although the structural forms of nitrogen incorporated into NOM via nitrate photolysis can reasonably be identified, elucidation of the reaction mechanisms and identification of the reactive nitrogen species requires further experimentation. Among the

Fig. 11. Liquid-state inverse-gated decoupled 15N nuclear magnetic resonance (NMR) spectra. (A) Suwannee River natural organic matter (SRNOM) reacted with Na15NO2 at initial pH of 7.5 (dark control reaction, 26 h). (B) Suwannee River natural organic matter irradiated with Na15NO2 at initial pH of 7.5 (4 h irradiation). Solvent = 75% D2O. LB, line broadening. 878

Fig. 12. Solid-state cross polarization/magic angle spinning (CP/MAS) 15 N nuclear magnetic resonance (NMR) spectrum of Suwannee River natural organic matter (SRNOM) irradiated with Na15NO2 in presence of bicarbonate at initial pH of 7.4. A Pyrex filter was used with medium pressure mercury lamp. LB, line broadening. Journal of Environmental Quality

questions that need to be addressed are the extents to which nitrogen dioxide, nitric oxide, hydroxylamine, and ammonia react with the NOM. The NMR data provide evidence for nitrosation reactions, confirming photonitrosation of NOM in the presence of nitrate, but the relative importance of nitration reactions is unresolved. This uncertainty could be addressed through the use of 17O-labeled nitrate in conjunction with 17O NMR, where the chemical shifts of nitro and oxime oxygens are clearly separated. The mechanism of ammonia formation during UV irradiation of nitrate in the presence of NOM also needs to be elaborated. These questions surrounding reaction mechanisms also pertain to the UV-catalyzed reduction and incorporation of the nitrite nitrogen. The NMR spectra reflect the distribution of nitrate nitrogen incorporated into the NOM under the conditions of these laboratory experiments for the starting concentrations, pH, and irradiation time specified. The potential for additional incorporation of nitrate nitrogen into the NOM would be limited by the continuing photochemical degradation of the carbon functionality of the NOM. Our previous work indicated that UV irradiation by itself does not result in complete mineralization of the NOM but that photodegradation reaches an endpoint yielding hydrophobic, decarboxylated, and predominantly C- and O-alkyl structures that are resistant to further photolysis (Thorn et al., 2010). We have seen that the nitrosophenol, oxime, and nitro groups incorporated into the NOM are photolabile. The stability toward further irradiation of the other broad range of nitrogen functionalities incorporated into the NOM needs to be determined. If photodegraded NOM molecules containing nitroso and nitro groups survive UV irradiation during water treatment, these would be of interest for their potentially toxicological properties. The susceptibility toward photodegradation of the amide, lactam, nitrile, and pyridine nitrogens incorporated into NOM through nitrite photolysis also needs to be evaluated. Determination of the balance between photochemical loss of naturally abundant nitrogen and incorporation of nitrate and nitrite nitrogen in NOM during UV irradiation under water treatment conditions is of interest from the perspective of nitrogen-containing disinfection by-product (N-DBP; e.g., haloacetamides, haloacetonitriles, N-nitrosodimethylamine, etc.) formation, when irradiation is followed by chlorination or chloramination. Neither the source nitrogen constituents in NOM nor the reaction mechanisms for N-DPB production are well understood. How incorporated nitrate and nitrite nitrogen may affect the quantity and distribution of N-DBPs is an additional research problem open for investigation. The possibility of conversion of naturally abundant primary amide or amine nitrogen in NOM into nitrogen gas when irradiated in the presence of nitrate may be taken into consideration for an accounting of the gaseous loss of nitrogen during water treatment with UV light. The sequence of reactions consisting of nitrate photolysis to nitrite, nitrosation of NOM primary amides or amines to diazonium ions, and decomposition of diazonium ions to nitrogen gas is a process that we may term “photo-chemodenitrification.” Confirmation of this process could be achieved through application of gas chromatography-isotope ratio mass spectrometry in conjunction with the use of 15N-labeled nitrate, where the N2 gas www.agronomy.org • www.crops.org • www.soils.org

would be comprised of an unlabeled nitrogen from the NOM and a labeled nitrogen from the nitrate. Production of nitrous oxide via photo-chemodenitrification is also conceivable, but in this instance both nitrogens of the N2O would arise from the labeled nitrate. Here, an oxime resulting from nitrosation of the NOM undergoes nitrosative deoximation (Freeman, 1973) to nitrous oxide on reaction with a second molecule of nitrous acid (Fig. 3).

Environmental Relevance To gain insight into the reaction mechanisms for incorporation of nitrate and nitrite nitrogen into NOM during photolysis, SRNOM was reacted individually with HNO3, NO2, and NO. These reactions are of fundamental interest in a variety of other research areas, as are abiotic reactions of nitrogen compounds with NOM in general. Recent identification of nitrated metabolites of pharmaceuticals and endocrine disruptors, such as acetaminophen (Chiron et al., 2010), 17α-ethinylestradiol (Gaulke et al., 2008), and nonylphenol isomers in sewage sludges (Telscher et al., 2005) and agricultural soils (Zhang et al., 2009) has led to speculation on the mechanisms of nitration for these compounds in their respective environmental matrices. Chiron et al. (2010) proposed that production of nitric oxide by nitrifying bacteria in activated sludge and its subsequent conversion to peroxynitrite in the presence of superoxide anion led to the nitration of acetaminophen. The authors speculated that if organic contaminants underwent nitration, the more abundant NOM should as well. There is an interest in the atmospheric nitration of the organic carbon fraction of carbonaceous aerosols (Deguillaume et al., 2008). Nitrated analogs of fulvic acids in the form of esters of nitric acid and possibly nitrosated molecules have been detected in atmospheric aerosol by FT-ICR mass spectrometry (Reemtsma et al., 2006), but the formation pathways remain open to speculation. Stemmler et al. (2006) recently demonstrated that visible light initiates the photoreduction of nitrogen dioxide to nitrous acid at the surface of humic acid. They proposed that this organic surface photochemistry reaction in soils could account for the observed daytime increases of nitrous acid in the troposphere, which is important because of the role of nitrous acid as a significant photochemical precursor of hydroxyl radical. Humic acid has also been shown to catalyze the reduction of nitrogen dioxide to nitrous acid in ice (Bartels-Rausch et al., 2010). The overall impact of NOM and humic-like substances on nitrate and nitrite photolysis in ice and snowpack has yet to be determined. The fate of nitrate contaminating acid forest soils through atmospheric deposition has been of long-term interest to ecologists ( Janssens et al., 2010). Retention of nitrate nitrogen at rates too high to be accounted for by biological immobilization has led researchers to investigate the significance of abiotic immobilization of the nitrate by soil organic matter (Davidson et al., 2003; TorresCañabate et al., 2008). In one hypothesis, soil organic matter is nitrosated by nitrite produced from iron(II)-catalyzed reduction of the deposited nitrate (Davidson et al., 2003). The potential importance to the cycling of nitrogen in the biosphere of co-denitrification reactions, in which nitrogen in NOM (primary amine and primary amide) is mobilized into N2 and N2O through reaction with nitrite and nitric oxide, and of reactions in which inorganic nitrogen is immobilized by NOM has recently been 879

highlighted (Spott et al., 2011). A further understanding of the reactions of nitric acid, nitrous acid, nitrogen dioxide, and nitric oxide with NOM will help to account for the reaction pathways underlying all these observations.

Conclusions The reduction and incorporation of nitrate nitrogen into aquatic NOM as a result of exposure to unfiltered UV irradiation from a medium-pressure mercury lamp has been demonstrated over the range of pH from 3 to 8 in the presence and absence of bicarbonate. Nitrosation has been confirmed, whereas nitration is probable but unconfirmed. Nitrate nitrogen was incorporated into the NOM as nitrosophenol, oxime, and nitro groups from nitrosation and nitration reactions and from pyridine, nitrile, lactam, amide, and other nitrogens, plausibly from secondary reactions of the initially formed adducts. The distribution of condensation products was affected by the concentrations of NOM, nitrate, and bicarbonate, but the relationships have not been determined. The nitrosophenol, oxime, and nitro nitrogens were found to be more susceptible to photodegradation on further irradiation than the pyridine, nitrile, amide, and lactam nitrogens. Incorporation of nitrite into NOM in the form of amide, lactam, nitrile, and pyridine nitrogens on unfiltered irradiation at pH 7.5 and amide, lactam, and heterocyclic nitrogens on Pyrex filtered irradiation at pH 7.4 in the presence of bicarbonate has also been demonstrated.

References Allard, B., H. Boren, C. Pettersson, and G. Zhang. 1994. Degradation of humic substances by UV irradiation. Environ. Int. 20:97–101. doi:10.1016/0160-4120(94)90072-8 Bartels-Rausch, T., M. Brigante, Y.F. Elshorbany, M. Amman, B. D’Anna, C. George, K. Stemmler, M. Ndour, and J. Kleffmann. 2010. Humic acid in ice: Photo-enhanced conversion of nitrogen dioxide into nitrous acid. Atmos. Environ. 44:5443–5450. doi:10.1016/j.atmosenv.2009.12.025 Behar, D., D. Shapira, and A. Treinin. 1972. Photolysis of hydroxylamine in aqueous solution. J. Phys. Chem. 76:180–186. doi:10.1021/j100646a006 Bonnett, R., R. Holleyhead, B.L. Johnson, and E.W. Randall. 1975. Reaction of acidified nitrite solutions with peptide derivatives: Evidence for nitrosamine and thionitrite formation from 15N NMR studies. J. Chem. Soc. Perkin Trans. 1(22):2261–2264. doi:10.1039/p19750002261 Buchanan, W., F. Roddick, and N. Porter. 2006. Formation of hazardous by-products resulting from the irradiation of natural organic matter: Comparison between UV and VUV irradiation. Chemosphere 63:1130– 1141. doi:10.1016/j.chemosphere.2005.09.040 Buffam, I., and K.J. McGlathery. 2003. Effect of ultraviolet light on dissolved nitrogen transformations in coastal lagoon water. Limnol. Oceanogr. 48:723–734. doi:10.4319/lo.2003.48.2.0723 Bushaw, K.L., R.G. Zepp, M.A. Tarr, D. Schulz-Jander, R.A. Bourbonniere, R.E. Hodson, W.L. Miller, D.A. Bronk, and M.A. Moran. 1996. Photochemical release of biologically labile nitrogen from dissolved organic matter. Nature 381:404–407. doi:10.1038/381404a0 Butler, A.R.T., D. Short, and J. Ridd. 1997. Tyrosine nitration and peroxonitrite (peroxynitrite) isomerisation: 15N CIDNP NMR studies. Chem. Commun. 669–670. doi:10.1039/a700676d Cai, Q., W. Zhang, and Z. Yang. 2001. Stability of nitrite in wastewater and its determination by ion chromatography. Anal. Sci. 17:917–920. doi:10.2116/analsci.17.917 Challis, B.C., N. Carman, M.H.R. Fernandes, B.R. Glover, F. Latif, P. Patel, J.S. Sandhu, and S. Shuja. 1994. Peptide nitrosations. In: R.N. Loeppky and C.J. Michejda, editors, Nitrosamines and related N-nitroso compounds chemistry and biochemistry. American Chemical Society, Washington, DC. p. 74–92. Chiron, S., E. Gomez, and H. Fenet. 2010. Nitration processes of acetamenophen in nitrifying activated sludge. Environ. Sci. Technol. 44:284–289. doi:10.1021/es902129c Davidson, E.A., J. Chorover, and D.B. Dail. 2003. A mechanism of abiotic immobilization of nitrate in forest ecosystems: The ferrous wheel hypothesis. 880

Glob. Change Biol. 9:228–236. doi:10.1046/j.1365-2486.2003.00592.x Deguillaume, L., M. Leriche, P. Amato, P.A. Ariya, A.-M. Delort, U. Poschl, N. Chaumerliac, H. Bauer, A.I. Flossmann, and C.E. Morris. 2008. Microbiology and atmospheric processes: Chemical interactions of primary biological aerosols. Biogeosciences 5:1073–1084. doi:10.5194/ bg-5-1073-2008 Dzengel, J., J. Theurich, and D.W. Bahnemann. 1999. Formation of nitroaromatic compounds in advanced oxidation processes: Photolysis versus photocatalysis. Environ. Sci. Technol. 33:294–300. doi:10.1021/ es980358j Earl, W.L. 1987. Effects of molecular mobility on high resolution solid state NMR spectra: Model systems. In: R.L. Wershaw and M.A. Mikita, editors, NMR of humic substances and coal. Lewis Publishers, Chelsea, MI. Fieser, M. 1967. Fiesers’ reagents for organic synthesis. Vol. 1. Wiley, Hoboken, NJ. Fischer, M., and P. Warneck. 1996. Photodecomposition of nitrite and undissociated nitrous acid in aqueous solution. J. Phys. Chem. 100:18749– 18756. doi:10.1021/jp961692+ Freeman, J.P. 1973. Less familiar reactions of oximes. Chem. Rev. 73:283–292. doi:10.1021/cr60284a001 Gaulke, L.S., S.E. Strand, T. Kalhorn, and H. Stensel. 2008. 17 α-ethinylestradiol transformation in the presence of ammonia oxidizing bacteria. Environ. Sci. Technol. 42:7622–7627. Gawley, R.E. 1987. The Beckmann reactions. In: A.S. Kende, editor, Organic reactions. Wiley, Hoboken, NJ. p. 1–420. Goldstein, S., and J. Rabani. 2007. Mechanism of nitrite formation by nitrate photolysis in aqueous solutions: The role of peroxynitrite, nitrogen dioxide, and hydroxyl radical. J. Am. Chem. Soc. 129:10597–10601. doi:10.1021/ ja073609+ Goldstone, J.V., M.J. Pullin, S. Bertilsson, and B.M. Voelker. 2002. Reactions of hydroxyl radical with humic substances: Bleaching, mineralization, and production of bioavailable carbon substrates. Environ. Sci. Technol. 36:364–372. doi:10.1021/es0109646 Graneli, W., M. Lindell, B.M. De Faria, and F. De Assis Esteves. 1998. Photoproduction of dissolved inorganic carbon in temperate and tropical lakes: Dependence on wavelength band and dissolved organic carbon concentration. Biogeochemistry 43:175–195. doi:10.1023/A:1006042629565 Green, J.B., and S.E. Manahan. 1979. Polarographic characterization of nitrohumic acids prepared by nitric acid oxidation of coal. Anal. Chem. 51:1126–1129. doi:10.1021/ac50044a009 Hecht, S.S., C.-H.B. Chen, R.M. Ornaf, E. Jacobs, J.D. Adams, and D. Hoffmann. 1978. Chemical studies on tobacco smoke. 52. Reaction of nicotine and sodium nitrite: Formation of nitrosamines and fragmentation of the pyrrolidine ring. J. Org. Chem. 43:72–76. doi:10.1021/jo00395a017 Janssens, I.A., . W. Dieleman1, S. Luyssaert, J-A. Subke, M. Reichstein, R. Ceulemans, P. Ciais, A. J. Dolman, J. Grace, G. Matteucci, D. Papale, S. L. Piao, E-D. Schulze, J. Tang, and B.E. Law. 2010. Reduction of forest soil respiration in response to nitrogen deposition. Nat. Geosci. 3:315–322. doi:10.1038/ngeo844 Kelemen, S.R., M. Afeworki, M.L. Gorbaty, P.J. Kwiatek, M.S. Solum, J.Z. Hu, and R.J. Pugmire. 2002. XPS and 15N NMR study of nitrogen forms in carbonaceous solids. Energy Fuels 16:1507–1515. doi:10.1021/ef0200828 Kieber, R.J., Z. Xianliang, and K. Mopper. 1990. Formation of carbonyl compounds from UV-induced photodegradation of humic sustances in natural waters: Fate of riverine carbon in the sea. Limnol. Oceanogr. 35:1503–1515. doi:10.4319/lo.1990.35.7.1503 Kieber, R.J., A. Li, and P.J. Seaton. 1999. Production of nitrite from the photodegradation of dissolved organic matter in natural waters. Environ. Sci. Technol. 33:993–998. doi:10.1021/es980188a Kirsch, M., and H. de Groot. 2009. N-Nitrosomelatonin: Synthesis, chemical properties, potential prodrug. J. Pineal Res. 46:121–127. doi:10.1111/j.1600-079X.2008.00655.x Kono, Y., H. Shibata, Y. Kodama, and Y. Sawa. 1995. The suppression of the N-nitrosating reaction by chlorogenic acid. Biochem. J. 312:947–953. Levy, G., and R.L. Lichter. 1979. Nitrogen-15 nuclear magnetic resonance spectroscopy. John Wiley & Sons, New York. Machado, F., and P. Boule. 1995. Photonitration and photonitrosation of phenolic derivatives induced in aqueous solution by excitation of nitrite and nitrate ions. J. Photochem. Photobiol. Chem. 86:73–80. doi:10.1016/1010-6030(94)03946-R Mack, J., and J.R. Bolton. 1999. Photochemistry of nitrate and nitrite in aqueous solution: A review. Photochem. Photobiol. Chem. 128:1–13. doi:10.1016/S1010-6030(99)00155-0 Mark, G., H.-G. Korth, H.-P. Schuchmann, and C. von Sonntag. 1996. The photochemistry of aqueous nitrate ion revisited. J. Photochem. Photobiol. Journal of Environmental Quality

Chem. A 101:89–103. doi:10.1016/S1010-6030(96)04391-2 Miller, W.L. 1998. Effects of UV radiation on aquatic humus: Photochemical principles and experimental considerations. In: D.O. Hessen and L.J. Tranvik, editors, Aquatic humic substances: Ecology and biogeochemistry. Springer, Berlin, Germany. p. 125–143. Miller, W.M., and R.G. Zepp. 1995. Photochemical production of dissolved inorganic carbon from terrestrial organic matter: Significance to the oceanic organic carbon cycle. Geophys. Res. Lett. 22:417–420. doi:10.1029/94GL03344 Minero, C., F. Bono, F. Rubertelli, D. Pavino, V. Maurino, E. Pelizzetti, and D. Vione. 2007. On the effect of pH in aromatic photonitration upon nitrate photolysis. Chemosphere 66:650–656. doi:10.1016/j. chemosphere.2006.07.082 Nelson, D.W. 1982. Gaseous losses of nitrogen other than through denitrification. In: F.J. Stevenson, editor, Nitrogen in agricultural soils. ASA,CSSA, and SSSA, Madison, WI. p. 327–363. Niessen, R., D. Lenoir, and P. Boule. 1988. Phototransformation of phenol induced by excitation of nitrate ions. Chemosphere 17:1977–1984. doi:10.1016/0045-6535(88)90009-4 Ohta, T., J. Suzuki, Y. Iwano, and S. Suszuki. 1982. Photochemical nitrosation of dimethylamine in aqueous solution containing nitrite. Chemosphere 11:797–801. doi:10.1016/0045-6535(82)90108-4 Olah, G.A., R. Malhotra, and S.C. Narang. 1989. Nitration methods and mechanisms. VCH, New York. Olasehinde, E.F., K. Takeda, and H. Sakugawa. 2010. Photochemical production and consumption mechanisms of nitric oxide in seawater. Environ. Sci. Technol. 44:8403–8408. doi:10.1021/es101426x Oppenlander, T. 2003. Photochemical purification of water and air. WileyVCH, Weinheim, Germany. Osburn, C.L., D.P. Morris, K.A. Thorn, and R.E. Moeller. 2001. Chemical and optical changes in freshwater dissolved organic matter exposed to solar radiation. Biogeochemistry 54:251–278. doi:10.1023/A:1010657428418 Patt, S.L. 1982. Alternating compound one eighties used to suppress transients in the coil. J. Magn. Reson. 49:161–163. Pennington, J.C., K.A. Thorn, L.G. Cox, D.K. MacMillan, S. Yost, and R.D. Laubscher. 2007. Photochemical degradation of composition B and its components. ERDC/EL TR-07–16. US Army Corps of Engineers, Washington, DC. Reemtsma, T., A. These, P. Ventatachari, X. Xia, P.K. Hopke, A. Springer, and M. Linscheid. 2006. Identification of fulvic acids and sulfated and nitrated analogues in atmospheric aerosol by electrospray fourier transform ion cyclotron resonance mass spectrometry. Anal. Chem. 78:8299–8304. doi:10.1021/ac061320p Schmidt-Rohr, K., J.-D. Mao, and D.C. Olk. 2004. Nitrogen-bonded aromatics in soil organic matter and their implications for a yield decline in intensive rice cropping. Proc. Natl. Acad. Sci. USA 101:6351–6354. doi:10.1073/ pnas.0401349101 Segal, W., and P.W. Wilson. 1949. Hydroxylamine as a source of nitrogen for azotobacter Vinelandii. J. Bacteriol. 57:55–60. Sharpless, C.M., and K.G. Linden. 2001. UV photolysis of nitrate: Effects of natural organic matter and dissolved inorganic carbon and implications for UV water disinfection. Environ. Sci. Technol. 35:2949–2955. doi:10.1021/es002043l Sleiman, M., L.A. Gundel, J.F. Pankow, P. Jacob, B.C. Singer, and H. Destaillats. 2010. Formation of carcinogens indoors by surface-mediated reactions of nicotine with nitrous acid, leading to potential thirdhand smoke hazards. Proc. Natl. Acad. Sci. USA 107:6576–6581. doi:10.1073/ pnas.0912820107 Smernik, R.J., and J.A. Baldock. 2005. Does solid-state 15N NMR spectroscopy detect all soil organic nitrogen? Biogeochemistry 75:507–528. doi:10.1007/s10533-005-2857-8 Spott, O., R. Russow, and C.F. Stange. 2011. Formation of hybrid N2O and hybrid N2 due to codenitrification: First review of a barely considered process of microbially mediated N-nitrosation. Soil Biol. Biochem. 43:1995–2011. doi:10.1016/j.soilbio.2011.06.014 Stedmon, C.A., S. Markager, L. Tranvik, L. Kronberg, T. Slätis, and W. Martinsen. 2007. Photochemical production of ammonium and transformation of dissolved organic matter in the Baltic Sea. Mar. Chem. 104:227–240. doi:10.1016/j.marchem.2006.11.005 Steinberg, C.E.W. 2003. Ecology of humic substances in freshwaters. Springer, Berlin, Germany. Stemmler, C., M. Ammann, C. Donders, J. Kleffmann, and C. Geoge. 2006. Photosensitized reduction of nitrogen dioxide on humic acid as a source of nitrous acid. Nature 440:195–198. doi:10.1038/nature04603 Stevenson, F.J. 1994. Humus chemistry: Genesis, composition, reactions. 2nd ed. John Wiley & Sons, New York. www.agronomy.org • www.crops.org • www.soils.org

Suzuki, J., N. Yagi, and S. Suzuki. 1984. Photochemical nitrosation of phenol in aqueous nitrite solution. Chem. Pharm. Bull. (Tokyo) 32:2803–2808. doi:10.1248/cpb.32.2803 Takahama, U.H.S. 2008. Reduction of nitrous acid to nitric oxide by coffee melanoidins and enhancement of the reduction by thiocyanate: Possibility of its occurrence in the stomach. J. Agric. Food Chem. 56:4736–4744. doi:10.1021/jf703660k Tarr, M.A., W. Wang, T.S. Bianchi, and E. Engelhaupt. 2001. Mechanisms of ammonia and amino acid photoproduction from aquatic humic and colloidal matter. Water Res. 35:3688–3696. doi:10.1016/ S0043-1354(01)00101-4 Telscher, J.S., U.B. Schmidt, and A. Shaeffer. 2005. Occurance of a nitro metabolite of a defined nonylphenol isomer in soil/sewage sludge mixtures. Environ. Sci. Technol. 39:7896–7900. doi:10.1021/es050551v Thomson, J., A. Parkinson, and F.A. Roddick. 2004. Depolymerization of chromophoric natural organic matter. Environ. Sci. Technol. 38:3360– 3369. doi:10.1021/es049604j Thorn, K.A., and M.A. Mikita. 1992. Ammonia fixation by humic substances: A nitrogen-15 and carbon-13 NMR study. Sci. Total Environ. 113:67–87. doi:10.1016/0048-9697(92)90017-M Thorn, K.A., and M.A. Mikita. 2000. Nitrite fixation by humic substances: Nitrogen-15 nuclear magnetic resonance evidence for potential intermediates in chemodenitrification. Soil Sci. Soc. Am. J. 64:568–582. doi:10.2136/sssaj2000.642568x Thorn, K.A. 2002. Reduction and incorporation of nitrate nitrogen into aquatic NOM upon UV irradiation examined by N-15 NMR. Preprints of Papers, Division of Environmental Chemistry, American Chemical Society 42(1):537–542. Thorn, K.A., J.B. Arterburn, and M.A. Mikita. 1992. 15N and 13C NMR investigation of hydroxylamine-derivatized humic substances. Environ. Sci. Technol. 26:107–116. doi:10.1021/es00025a011 Thorn, K.A., and L.G. Cox. 2009. N-15 NMR spectra of naturally abundant nitrogen in soil and aquatic natural organic matter samples of the International Humic Substances Society. Org. Geochem. 40:484–499. doi:10.1016/j.orggeochem.2009.01.007 Thorn, K.A., S.J. Younger, and L.G. Cox. 2010. Order of functionality loss during photodegradation of aquatic humic substances. J. Environ. Qual. 39:1416– 1428. doi:10.2134/jeq2009.0408 Tickell, O. 2008. Nitrites and cancer. Ecologist 38:14–15. Torres-Cañabate, P., E.A. Davidson, E. Bulygina, R. García-Ruiz, and J.A. Carreira. 2008. Abiotic immobilization of nitrate in two soils of relic Abies pinsapo-fir forests under Mediterranean climate. Biogeochemistry 91:1– 11. doi:10.1007/s10533-008-9255-y Vione, D., V. Maurino, C. Minero, and E. Pelizzetti. 2001. Phenol photonitration upon UV irradiation of nitrite in aqueous solution I: Effects of oxygen and 2-propanol. Chemosphere 45:893–902. doi:10.1016/ S0045-6535(01)00035-2 Vione, D., V. Maurino, C. Minero, and E. Pelizzetti. 2005. Nitration and photonitration of naphthalene in aqueous system. Environ. Sci. Technol. 39:1101–1110. doi:10.1021/es048855p Vione, D., G. Falletti, V. Maurino, C. Minero, E. Pelizzetti, M. Malandrino, R. Ajassa, R.-J. Plariu, and C. Arsene. 2006. Sources and sinks of hydroxyl radicals upon irradiation of water samples. Environ. Sci. Technol. 40:3775–3781. doi:10.1021/es052206b Wang, W., M.A. Tarr, T.S. Bianchi, and E. Engelhaupt. 2000. Ammonium photoproduction from aquatic humic and colloidal matter. Aquat. Geochem. 6:275–292. doi:10.1023/A:1009679730079 Wang, X., T. Lou, and H. Xie. 2009. Photochemical production of dissolved inorganic carbon from Suwannee River humic acid. Chin. J. Oceanology Limnol. 27:570–573. doi:10.1007/s00343-009-9156-5 White, G.C. 2010. White’s handbook of chlorination and alternative disinfectants. 5th ed. John Wiley & Sons, Hoboken, NJ. Williams, D.L.H. 1988. Nitrosation. Cambridge Univ. Press, Cambridge, UK. Williams, D.L.H. 2004. Nitrosation reactions and the chemistry of nitric oxide. Elsevier, Amsterdam. Witanowski, M., L. Stefaniak, and G.A. Webb. 1993. Nitrogen NMR spectroscopy. Academic Press, New York. Zhang, H.S., M. Spitellera, K. Guentherb, G. Boehmlerc, and S. Zuehlke. 2009. Degradation of a chiral nonylphenol isomer in two agricultural soils. Environ. Pollut. 157:1904–1910. doi:10.1016/j.envpol.2009.01.026 Zheng, A., D.A. Dzombak, and R.G. Luthy. 2004. Effects of nitrosation on the formation of cyanide in publicly owned treatment works secondary effluent. Water Environ. Res. 76:197–204. doi:10.2175/106143004X141735

881