Oxidation of Amino Groups by Hydroxyl Radicals in Relation to the

Jun 7, 2002 - Reaction mechanisms are proposed which account for the observed inorganic nitrogen end product depending on the oxidation degree of the ...
2 downloads 7 Views 112KB Size
Download PDF
Environ. Sci. Technol. 2002, 36, 3083-3089

Oxidation of Amino Groups by Hydroxyl Radicals in Relation to the Oxidation Degree of the r-Carbon N. KARPEL VEL LEITNER,* P. BERGER, AND B. LEGUBE Laboratoire de Chimie de l’Eau et de l’Environnement, UMR CNRS 6008, Ecole Supe´rieure d’Inge´nieurs de Poitiers, Universite´ de Poitiers, 40 Avenue du Recteur Pineau, 86022 Poitiers Cedex, France

Nitrogen organic compounds constitute a large class of aqueous pollutants. These compounds include not only azoic structures, nitrogen heterocycles, and nitrous groups but also amides and amines. This work consisted in studying the OH° induced oxidation of simple primary amines in dilute aqueous solution with special attention to mineralization of the nitrogen group as a function of the nature of the R-carbon. H2O2/UV and γ-irradiation processes were used for the production of OH° radicals, and the molecules studied were one R-amino acid i.e., glycine (HOOCCH2NH2), and two primary amides i.e., acetamide (CH3CONH2) and oxamic acid (HOOCCONH2). It was shown that the oxidation of glycine leads to the formation of ammonia, whereas the acetamide molecule is first oxidized into oxamic acid ending in complete mineralization with production of nitrates. Reaction mechanisms are proposed which account for the observed inorganic nitrogen end product depending on the oxidation degree of the carbon atoms of the molecules. It follows that the present study will allow for prediction of the fate of nitrogen resulting from the oxidation of primary amino groups by OH° radicals.

FIGURE 1. Removal of glycine during the photolysis of hydrogen peroxide in the presence (a) and in the absence (b) of oxygen: glycine removal (0) and production of ammonium ions (×); [Gly]o ) [H2O2]o ) 1 mM; pH ) 7.8; photon flux ) 2.78 J s-1.

Introduction Nitrogen functional groups are included in the structure of numerous organic compounds. These compounds as far various as humic substances, proteins, pesticides, dyes, ... are found in surface waters and industrial wastewaters (1, 2). Therefore their behavior in water treatment plants and especially the determination of the byproducts formed during their chemical oxidation are of interest. Numerous data are available relative to the reactions of chlorine with organic nitrogen compounds (3-6) but much less concern with ozone and hydroxyl radicals. Before the study of complexe molecules our work focused on primary aliphatic amines with various functional groups in R position. Literature indicates that the molecule of ozone reacts mainly on the unprotonated nitrogen of amines and amino acids yielding hydroxylamine, oxime, carboxylic acids, and then nitrates as the major end-product (7-10). Studies on the ozonation of glycine in aqueous solution showed the formation of nitrate and ammonium ions as well as formic acid and carbon dioxide (10, 11). From investigations on the ozonation of serine, Le Lacheur and Glaze (12) reported that the oxidation by molecular ozone yields nitrates after decarboxylation of the amino acid, whereas under radicalpromoting conditions, ammonia is formed. Similarly, we * Corresponding author phone: 33 5 49 45 39 16; fax: 33 5 49 45 37 68; e-mail: [email protected]. 10.1021/es0101173 CCC: $22.00 Published on Web 06/07/2002

 2002 American Chemical Society

FIGURE 2. Evolution of hydrogen peroxide during the removal of glycine by the H2O2/UV process in the presence and in the absence of oxygen (experimental conditions of Figure 1). demonstrated (13) that the ammonia found by Duguet et al. (10, 11) from the ozonation of glycine arose from a OH° radical induced pathway. Like ozone, the OH° radical attack is dependent on the protonation of the nitrogen atom and consequently on the pH of the solution. The initial attack was found to occur on the carbon chain when the nitrogen is protonated and both on the nitrogen or the R-carbon when the nitrogen is unprotonated (14). Thus, both HN°-CH2COO- and H2N-C°H-COO- are assumed to be formed from the initial attack of OH° on the glycine anions. As the number of C-H bonds is increased in the molecule, competition appears for OH° radicals between C-H and N-H bonds (15). These sites of attack are still a matter of controversy in the literature for short chain molecules. Main pathways will be described and discussed in the later part (Figure 6) of the present article in light of our results. Recently, Bonifacic et al. (16) indicated that the initial step in the OH° induced mechanism was oxidation of the amino group, producing VOL. 36, NO. 14, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3083

FIGURE 3. Removal of glycine by the γ-irradiation/N2O process in the presence and in the absence of oxygen; production of ammonium ions; [Gly]o ) 1 mM; pH ) 7.6.

FIGURE 4. Removal of glycine by the γ-irradiation/H2O2 process in the presence ([O2]0 ) 280 µM) and in the absence of oxygen: production of ammonium ions; [Gly]o ) [H2O2]o ) 1 mM; pH ) 7.6.

FIGURE 6. Removal of acetamide (0) during the photolysis of H2O2. Production of oxamic acid (4) and ammonium ions (×) in the presence (a) and in the absence (b) of oxygen: [acetamide]o ) 1 mM; [H2O2]o ) 2 mM; pH ) 8.0; photon flux ) 2.78 J s-1.

TABLE 1. Studied Nitrogen Organic Compounds: Formula and Rate Constants for the Reaction with •OH Radicals name

p Ka

k•OH (M-1 s-1) (22)

2.34 (25 °C) 1.7 107 (pH 5.8-6) 9.60 (25 °C) 1.9 109 (pH 10) acetamide 0.63 (25 °C) 1.9 × 108 (pH 5.5) H3C-CO-NH2 oxamic acid (oxalic acid monamide) 2.5 (24 °C) ≈ 105 (pH 8)a -OOC-CO-NH + 11.8 (24 °C) 3 a

FIGURE 5. Evolution of hydrogen peroxide during the removal of glycine by the γ-irradiation/H2O2 process in the presence and in the absence of oxygen (experimental conditions of Figure 4). - and HN°-CH -COO- from electron 2N°-CH2-COO 2 transfer and H-atom abstraction, respectively. As previously stated by Monig et al. (17), these authors showed that the amino radical cation +H2N°-CH2-COO- rapidly decomposes into CO2 and °CH2NH2. The other primary radical HN°CH2-COO- converts into the decarboxylating +H2N°-CH2COO-, releases the CO2°- radical and HNdCH2, or transforms into a carbon-centered radical H2N-C°H-COO- (18). Decarboxylations are considered to be the major pathway. According to these authors, the C-centered H2N-C°H-COOradical is not generated in a direct H-atom abstraction by OH°. An unstable imino structure is sometimes suggested from this radical decomposition (19-21). Generally, the fate of the primary radicals formed is not mentioned. Moreover, most of the experiments are conducted under oxygen-free conditions. Our research consisted of determining the influence of the functional groups bound to the nitrogen of primary amines on the reactions induced by OH° radicals, in the presence and in the absence of oxygen. Therefore three simple -NH2 containing molecules were +H

3084

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 14, 2002

Deduced from our results.

chosen: glycine, acetamide, and oxamic acid (Table 1). Simultaneously to the decay of the initial molecule induced by the photolysis of hydrogen peroxide or by γ-irradiation, the formation of inorganic nitrogen and carboxylic acids was followed. Mechanisms were proposed that account for the organic compounds and the nitrate or ammonium ions observed as end-products.

Experimental Section Materials and Methods. Glycine (purity > 99%; Sigma), oxamic acid (purity > 98%; Fluka), acetamide (purity > 99%; Prolabo), and hydrogen peroxide (30%; Fluka) were used as received. Experiments were realized at 20 ( 0.5 °C with solutions prepared in ultrapure water delivered by a Millipore MilliRO-MilliQ system (DOC < 0.1 mg C.L-1) and buffered with phosphates (ionic strengh ) 10-2 M; pH ≈ 8). H2O2/UV Process. H2O2/UV experiments were carried out in a 4-L cylindrical batch reactor surrounded with a 20 °C isothermal jacket. This batch reactor was equipped with a low-pressure mercury vapor lamp (Hanau NN 15/20) in the axial position leading to an optical width of l ) 6.25 cm (23). This lamp emits a monochromatic radiation at 253.7 nm.

The incident photonic flux Io ) 5.9 × 10-6 E.s.-1 was determined by chemical actinometry using 5 × 10-2 M hydrogen peroxide as actinometer and the simplified relation - d[H2O2]/dt ) φIo/V for high absorbance with φ ) 1 for the overall reaction of H2O2 photolysis. Some experiments were carried out in the absence of oxygen by bubbling nitrogen through a fritted glass fitting at the bottom of the reactor. Thirty-mL samples were taken from the reactor for analysis before irradiation and after various exposure times. γ-Irradiation. Considering the small reactivity of oxamic acid with OH° radicals (k ≈ 105 M-1 s-1 from competitive kinetic experiments) by comparison with hydrogen peroxide (kOH°/H2O2 ) 2.7 × 107 M-1 s-1; kOH°/HO2- ) 7.5 × 109 M-1 s-1 ; pKa ) 11.6), the kinetic of oxamic acid oxidation by the H2O2/UV process was very small due to the important scavenging effect of H2O2. Therefore experiments on oxamic acid were performed using the γ-irradiation process in preference to the H2O2/UV system to get significant removal of the acid. This process was also used for some experiments with glycine. One-liter buffered solutions of the organic molecule were irradiated using a AECL Gammacell 220 60Co source (average dose rate: 71.6 Gy/min) at the Austrian Research Center of Seibersdorf. The bottom of the reactor was equipped with a fritted glass for nitrous oxide (N2O) bubbling alone or with 20% oxygen during irradiation (24). Solutions were first saturated by sparging for 15 min with the gas mixture before the beginning of the experiment and then continuously during irradiation. The interaction of water with ionizing radiation such as γ-rays results in the formation of a suite of radical species among which the primary OH° radicals and the hydrated electrons (e-aq) are predominant (see reaction 1). When nitrous oxide is introduced it reacts rapidly with the solvated electrons (ke-/N2O ) 9.1 × 109 M-1 s-1) to form nitrogen gas and the hydroxyl radical (see reaction 2). Saturation of a solution with N2O thus doubles the OH° radical yield and favors oxidizing conditions upon irradiation. Molecular oxygen also scavenges the solvated electrons (k ) 1.9 × 1010 M-1 s-1) and the hydrogen atom (k ) 2.1 × 1010 M-1 s-1) forming the superoxide radical anion O2°- and the hydroperoxyl radical HO2° (see reactions 3 and 4). But because of the much higher solubility of N2O (2.4 × 10-2 M at 1 atm) compared to O2 (1.3 × 10-3 M), when a 4:1 mixture of these gases is used reactions 3 and 4 can be neglected. For some experiments nitrous oxide bubbling was replaced by introduction of hydrogen peroxide in the solutions at the beginning of the experiments. As N2O, H2O2 reacts rapidly with the solvated electrons (ke-/H2O2 ) 1.1 × 1010 M-1 s-1) to form hydroxyl radicals (reaction 5). The oxygen-free conditions were obtained by bubbling nitrogen into the solution before the experiment and keeping a nitrogen flux at the surface of the solution during the experiment.

Analytical Methods

e- + O2 f O2°- k ) 1.9 × 1010 M-1 s-1

(3)

k ) 2.1 × 1010 M-1 s-1

(4)

For the analysis of hydrogen peroxide, the colorimetric method with titanium chloride as reagent was used (23). The spectrophotometric measurements were performed by using a Safas Double Energy System 190 spectrophotometer set at 410 nm and a 1 cm path quartz cell. Glycine was determined by HPLC with a spectrofluorimetric detection (detector F-1050 Merck) provided with a software system (Millennium 2010 Waters). After a 1 to 5 dilution, 50 µL samples were derivatized with 18 µL of a solution of orthophthaldialdehyde (100 mg of orthophthaldialdehyde, 9 mL of methanol, 1 mL of borate buffer, 100 µL of mercaptoethanol) with 29 µL of borate buffer and then 3 µL of HCl (0.75 N) using an automatic injector (AS 4000 Merck). Twenty µL of the derivatized sample were then injected into a C18 column (Delta Pak C18 Waters 5 µm-100 Å-3.9 × 150 mm) using methanol in ultrapure water (30:70) as an eluent solution at an isocratic flow rate of 0.9 mL min-1 (13). Acetamide was analyzed in aqueous solution with a Chrompack (CP 9000) Gas Chromatograph equipped with a FFAP capillary column (Alltech Econo-Cap; 30 m × 0.25 mm) and a flame ionization detector. Nonanoic acid (500 µM) in methanol was used as internal standard. The column temperature was programmed from 50 °C with a 4 min hold to 220 °C with a 6 °C/min rise. For a 0.5 µL injection volume in the on-column mode, the detection limit was close to 10 µM. Oxamic acid and some byproducts (oxalic and formic acids) were quantified by HPLC combined either with a photodiode array detector (PDA 996 Waters) equipped with the Millennium 2015 Waters software system or with a UV detector set at 210 nm (LDC Spectromonitor 3100). The column used was an ionic column (Supelcogel C-610H) with a 0.1% concentrated phosphoric acid solution in deionized water as an eluent at an isocratic flow of 0.8 mL min-1. One hundred µL samples were injected by an automatic injector (Waters 717) allowing a detection limit of 0.1 µM. Inorganic ions (NO2-, NO3-, NH4+) were analyzed using a capillary electrophoresis apparatus (CIA Waters) equipped with a fused-silica capillary (60 cm × 75 µm), an autosampler, a data acquisition system (Millennium 2010 Waters), and a UV detector set at 185 nm. For ammonium ions, the electrolyte was made of UV CAT 3 (1 g/L; Waters) in aqueous solution with a pH value adjusted in the range 4.2-5.0. The hydrostatic sample injection mode was used for the highest concentrations of ammonium ions (>0.04 mM). The injection time was 30 s with a run voltage of 20 kV. For lower concentrations, the electromigration sample injection mode was used with a sample potential of 5 kV allowing a detection limit of 4 µM. Lithium ions were used as internal standard at a concentration of 0.2 mM for the hydrostatic mode and 20 µM for the electromigration mode. For nitrite and nitrate ions, a fused-silica capillary (60 cm × 100 µm) was also used with a solution of OFM-OH- (25 mL/L; Waters) and sodium sulfate (1420 mg Na2SO4/L) in water as electrolyte. The injection time used was 50 s with a run voltage of 12 kV in the hydrostatic mode. Each analysis started in the isomigration mode i.e., with a constant current of 39 µA during 60 s to ensure a good reproducibility of the retention times. These conditions were optimal for the determination of concentrations in the range of 1-800 µM.

(5)

Results

H2O f e-; OH°; H°

G(e-) ) 0.27; G(OH°) ) 0.28; G(H°) ) 0.07 (1) G: energy efficiency in µmol/J

-

e + N2O + H2O f OH° + OH- + N2

k ) 9.1 × 109 M-1 s-1 (2)

H° + O2 f HO2°

e- + H2O2 f OH° + OH-

k ) 1.1 × 1010 M-1 s-1

The unit of the absorbed radiation dose is the Gray (Gy). 1 Gy is defined as an energy absorption of 1 J kg-1. Experiments were renewed for each dose selected for samples analysis.

OH° Induced Oxidation of Glycine. H2O2/UV Process. The decomposition of glycine induced by the photolysis of hydrogen peroxide was studied in the presence and in the absence of dissolved oxygen. Under the experimental conditions applied (1 mM of glycine), removal of glycine due to VOL. 36, NO. 14, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3085

TABLE 2. Organic Byproducts and Ammonium Ions Identified from the Oxidation of Glycine by the H2O2/UV System in the Presence of Oxygen (Experimental Conditions: See Figure 1) glycine removed (mM) oxalic acid produced (mM) formic acid produced (mM) ammonium ions produced (mM)

0.388 0.064 0.021 0.309

0.454 0.085 0.025 0.353

0.496 0.080 0.021 0.372

direct photolysis was found negligible. The data indicate that whatever the oxygen level, ammonium ions are the main nitrogen endproduct of the OH° induced oxidation of glycine by the use of the H2O2/UV system. In oxygen-free solutions ammonium ions are formed according to the stoichiometry of 1 mole per mole of glycine removed (Figure 1b). A lower yield was observed in the presence of oxygen since only 0.75 mole of NH4+ was found per mole of glycine decomposed (Figure 1a). Under this latter condition the total nitrogen content of the solution (not presented) does not change during irradiation. No nitrite and only very low concentrations of nitrate ions were detected ([NO3-] < 5 µM) indicating the formation of nitrogen organic byproducts in oxygen containing solutions. The organic byproducts identified, oxalic and formic acids, were found only in the presence of oxygen (Table 2). Under these conditions, the carbon mass balance was almost complete (13). In oxygen-free solution, the carbon mineralized only represents 0-30% of the glycine removed. During the experiment, the concentration of dissolved oxygen dropped from 294 µM (t ) 0 min) to 231 µM (t ) 62 min) and then increased gradually up to 278 µM by the end of the irradiation period (t ) 360 min). The initial decrease is consistent with reactions of oxygen with the intermediates formed from the attack of OH° on the glycine molecule. The subsequent oxygen increase is associated with a lower consumption of O2 due to the slowing down of glycine removal and additionally with the gradual balance with the air volume surmounting the solution. Considering the evolution of H2O2, the rate of H2O2 removal was twice as low as in the presence of oxygen than that for the photolysis of oxygen free solutions (Figure 2). Assuming a simplified firstorder kinetic for diluted solutions, these rates are respectively 1.90 × 10-4 s-1 and 3.50 × 10-4 s-1. From the integrated form of the general kinetic expression - d[H2O2]/dt ) φIo/V (1 e-2.3l[H2O2]) for direct photolysis, the slope - 2.3lφIo/V of the linear relation ln(10l[H2O2]-1) ) f(t) enables the determination of apparent quantum yields φ for the removal of hydrogen peroxide. These values were equal to 0.50 and 0.90 in the presence and in the absence of dissolved oxygen, respectively. γ-Irradiation Process. The γ-irradiation experiments with N2O bubbling showed that the glycine removal is favored in oxygen-free solution (Figure 3). In the presence like in the absence of oxygen, the organic nitrogen is totally converted into ammonium ions (1 mole of NH4+ produced per mole of glycine removed). Under both conditions, oxalic and formic acids were identified (Table 3). The concentration of oxalic acid was higher in the presence of oxygen and reached 0.44 mol/mol of glycine removed for the 2000 Gy radiation dose. Concerning the experiments with the γ-irradiation/H2O2 system, the results indicate a smaller production of ammonium ions in the presence of oxygen (0.85 mol/mol of glycine removed against 0.94 mol/mol in the absence of oxygen; Figure 4). When oxygen is absent, more solvated electrons are available for reaction with hydrogen peroxide to yield hydroxyl radicals. Therefore, hydrogen peroxide was removed more rapidly in the absence of oxygen, and that is why 0.5 mM of H2O2 was added to the solution after the dose 1000 Gy (Figure 5). The formation of oxalic and formic acids was higher than the concentrations determined during γ-irradiation with N2O. 3086

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 14, 2002

TABLE 3. Organic Byproducts and Ammonium Ions Identified from the Oxidation of Glycine by the γ-Irradiation/N2O Process in the Presence and in the Absence of Oxygen (Experimental Conditions: See Figure 3) dose (Gy)

2000

5000

8000

0.191 0.084 0.024 0.164

0.489 0.134 0.015 0.490

0.615 0.165 0.014 0.600

Without Oxygen glycine removed (mM) 0.291 oxalic acid produced (mM) 0.013 formic acid produced (mM) 0.014 ammonium ions produced (mM) 0.292

0.625 0.070 0.023 0.633

0.821 0.133 0.020 0.813

With Oxygen glycine removed (mM) oxalic acid produced (mM) formic acid produced (mM) ammonium ions produced (mM)

TABLE 4. Organic Byproducts and Ammonium Ions Identified from the Oxidation of Glycine by the γ-Irradiation/H2O2 Process in the Presence and in the Absence of Oxygen (Experimental Conditions: See Figure 4) dose (Gy)

2000

5000

0.267 0.029 0.026 0.205

0.594 0.100 0.014 0.506

Without Oxygen glycine removed (mM) 0.274 oxalic acid produced (mM) 0.035 formic acid produced (mM) 0.030 ammonium ions produced (mM) 0.270

0.572 0.125 0.014 0.563

With Oxygen glycine removed (mM) oxalic acid produced (mM) formic acid produced (mM) ammonium ions produced (mM)

8000

0.740 0.128 0.013 0.730

These concentrations were similar in the presence and in the absence of oxygen (Table 4). For all these experiments relative to the removal of glycine, when detected, the concentrations of nitrate ions and oxamic acid were always below 22 and 11 µM, respectively. OH° Induced Oxidation of Acetamide and Oxamic Acid. Acetamide. The H2O2/UV process was used for the study of the oxidation of acetamide by OH° radicals. Under the conditions of the experiments (1 mM acetamide) no direct photolysis of acetamide was observed by the end of the irradiation period (t ) 250 min) without hydrogen peroxide. Figure 6a,b shows the results of the H2O2/UV treatment of aqueous solutions of acetamide in the presence and in the absence of oxygen, respectively. In the presence of oxygen, the hydroxyl radicals resulting from the photolysis of hydrogen peroxide react with acetamide to form oxamic acid as the major byproduct with a ratio of 0.94 mole per mole of acetamide removed at the end of the experiment (Figure 6a). The rate of H2O2 removal was similar in the presence and in the absence of oxygen (2.23 × 10-4 s-1) corresponding to apparent quantum yields around 0.6 (Figure 7). Acetamide oxidation was more rapid in oxygen-free solutions (Figure 6b). 0.72 mole of oxamic acid was observed per mole of acetamide removed, and ammonium ions were produced, reaching 0.1 mole per mole of acetamide removed. Under both experimental conditions (with and without oxygen), the total organic carbon remained constant during irradiation showing no carbon mineralization. When present at the beginning of the irradiation, oxygen undergoes first a surprising short increase from 444 µM (t ) 0) to 553 µM (t ) 21 min) followed by a rapid decrease up to the value of 366 µM (t ) 246 min). Oxamic Acid. The reaction between oxamic acid and OH° radicals was studied in the presence of oxygen by using γ-irradiation as the OH° radicals source with continuous

FIGURE 7. Evolution of hydrogen peroxide during the removal of acetamide by the H2O2/UV process in the presence and in the absence of oxygen (experimental conditions of Figure 6).

FIGURE 8. Removal of oxamic acid (4) during γ-irradiation (N2O/ O2: 4/1); removed TOC (×); production of nitrate ions (o); [oxamic Ac.]o ) 1 mM; pH ) 8.0. introduction of N2O and O2 (4:1). Figure 8 indicates that high radiation doses are required for the removal of oxamic acid. This small reactivity is easily understandable since at pH 8 oxamic acid is present as a zwitterion and because of the absence of C-H bonds. However, results show that the OH° induced oxidation allows a complete mineralization of the oxamic acid into carbon dioxide and nitrates. The total organic carbon (TOC) concentration determined experimentally (Figure 8) is in good agreement, within experimental error, with that calculated from the residual oxamic acid concentrations during irradiation and is consistent with the absence of organic intermediates.

Discussion Glycine. Considering the rate constants for the reaction of OH° radicals with the zwitterion (k ) 1.7 × 107 M-1 s-1) and the anionic form (k ) 1.9 × 109 M-1 s-1) of glycine, at pH 8, the reaction with the anionic form represents 2.8 times that of the neutral form (pKa ) 9.6). The byproducts identified in this work give information about the fate of the short live species generated (Figure 9). As discussed in the Introduction, there are several different mechanisms by which hydroxyl radical may react with glycine. Assuming the OH° radical attack on the unprotonated amino group, the H°N-CH2-COO- radical can decompose into HNdCH2 and CO2°- (reaction 7). The hydrolysis of the intermediate HNdCH2 (and +H2NdCH2), proposed by Bonifacic et al. (16, 18), could account for the formic acid analyzed in the solutions. The radical CO2°- is rapidly oxidized into carbon dioxide by oxygen, by hydrogen peroxide (23) (and probably by N2O). The H°N-CH2-COO- radical can also be protonated (reaction 9) by reaction with phosphates (kH2PO4) 7.4 × 107 M-1 s-1, kHPO42- ) 2.5 × 105 M-1 s-1) or by the glycine zwitterion (k ) 3.9 × 105 M-1 s-1) (16). The protonated +H N°-CH -COO- radical has been reported to rapidly 2 2

FIGURE 9. Main pathways for the oxidation of glycine by hydroxyl radicals at pH 8. decarboxylate giving rise to the formation of the reducing H2N-C°H2 radical (reaction 11). However, the results obtained in this study and especially the formation of oxalic acid show that the intermediate species formed do not yield only decarboxylation reactions. Therefore, we cannot exclude the occurrence of carboncentered radicals that induce the N-C bond cleavage with NH4+ formation before carbon mineralization. H2N-C°HCOO- radicals could originate from the HN°-CH2-COOradicals by reactions with the glycine anions or by an intramolecular 1,2-H-atom shift even if these pathways turn out to be unfavored according to the rate constants (k ) 3 × 104 M-1 s-1 and 1200 s-1, respectively) determined by Bonifacic et al. (16). Moreover, the positive charge of +H3NCH2-COO- masks the protonated amino group from being oxidized by the attracking electrophilic OH° radical and makes noticeable the attack on the C-H bond (reaction 6b). Performing experiments with hydrogen peroxide using the γ-irradiation/H2O2 system, in the presence and in the absence of oxygen, helps support that this carbon-centered reducing radical reacts not only with oxygen (21) but also with hydrogen peroxide to yield the imino acetic acid (17) (reaction 12a). The production of oxalic acid detected in our work indicates that this route is likely favored against the competing reaction of this radical with the glycine zwitterion (reaction 10; k ) 3 × 103 M-1 s-1). Therefore, the reaction of hydrogen peroxide with the C-centered radical could be considered responsible for the rapid removal of the former in oxygen free solution, in addition to the reactions between H2O2 and solvated electrons (in absence of reaction of e-aq with O2) and CO2°radicals (23). The rapid hydrolysis of the imino acetic acid yields ammonia and glyoxylic acid rapidly oxidized into oxalic acid according to a well-known mechanism (25). However, opposite to the γ-irradiation/H2O2 process, oxalic acid was not detected in the absence of oxygen in the H2O2/UV process. Thus, this latter pathway could not be the exclusive one. VOL. 36, NO. 14, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3087

The lesson of this work is also that oxalic acid was found with the γ-irradiation/N2O system in the absence of oxygen. If we assume that reactions 12a and 12b are the main pathways leading to the C-C bond conservation, N2O appears to be involved as oxidant like H2O2 or O2. These results agree with the observations of von Sonntag et al. (26, 27) relative to electron transfer of reducing radicals to N2O contrary to the data of Bonifacic et al. (16) from γ-radiolysis experiments. We found that (i) with the γ-irradiation/N2O system, oxalic acid was detected in a higher amount in the presence of oxygen and (ii) in oxygen-free solutions, less oxalic acid was detected with the γ-irradiation/N2O system compared with γ-irradiation/H2O2. Therefore, the reactivity of N2O toward the C-centered radicals seems to be smaller than O2 or H2O2. Similarly, in addition to O2, the reaction of H2O2 and probably N2O with the reducing H2N-C°H2 radical to produce +H2Nd CH2 (reaction 14) could be expected (17). Both +H2NdCH2 and HNdCH2 are unstable and decompose into formaldehyde and NH4+ and NH3, respectively (reactions 13a and 13b). The oxidation of formaldehyde results in formic acid identified in this work. In the reactions of the C-centered radicals with oxygen, the superoxide release is estimated to be fast (21). However, the mechanism could involve the formation of intermediate peroxyl radicals (reactions 15 and 16). The peroxyl radical with two carbon atoms would competitively decompose into the imino acetic acid (20, 21, 28) or according to bimolecular reactions through several pathways as described by Von Sonntag and Schuchmann (29) for primary (RCH2O2°) and secondary (R2CHO2°) peroxyl radicals. These latter reactions would produce oxamic acid identified in small amounts, formamide, glyoxal, unstable iminoglycolic acid, and hydrogen peroxide. A minor pathway may exist that leads to the few unidentified nitrogen byproducts in the presence of oxygen. The reaction of O2 with the deprotonated (HN°-CH2-COO-) R-amino radical has been reported to form peroxyl radicals (°OONH-CH2-COO-) and then oxime, nitroparaffine, and hydroxylamine (30). However, the oxidation of these latter molecules by OH° radicals would subsequently produce nitrites and nitrates which were found in very small amounts under our experimental conditions compared to ammonia. The oxidation of the end-products NH3/NH4+ by the hydroxyl radicals is not expected under our experimental conditions. The unprotonated form ammonia is the only reactive species with a rate constant of 9 × 107 M-1 s-1 for the reaction with OH° radicals (31, 32). Thus at pH 8 and with a maximum ammonia concentration of 1 mM, no further oxidation occurs during our experimental conditions. Acetamide and Oxamic Acid. The formation of oxamic acid as the major intermediate from the oxidation of acetamide by OH° radicals and its subsequent decomposition into nitrate ions shows that the fate of the nitrogen group differs from the molecule of glycine. These results underline that the reactivity of the amino group toward OH° radicals is dependent on the R functional groups. The production of oxamic acid from the oxidation of acetamide in the presence of oxygen (0.94 mole of oxamic acid per mole of acetamide removed) indicates that the attack on the nitrogen of the acetamide molecule which is supposed to take place (33) does not seem to be favored and that the reaction retains the backbone of the amide. Unlike glycine, the initial attack of the OH° radicals on the acetamide molecule is directed only on the carbon atom (reaction 18) (34-36). The electronsattractive effect of the -CO-NH2 functional group increases the lability of the hydrogen atoms of the methyl group. It follows that the H abstraction is rapid and results in the formation of oxamic acid according to a mechanism 3088

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 14, 2002

probably similar to the oxidation of acetic acid into oxalic acid (25).

The smaller formation of oxamic acid in the absence of oxygen constitutes another similarity with the oxidation of acetic acid by the H2O2/UV process where oxalic acid was also found in lesser extent (25). Under these conditions, the presence of ammonium ions indicates that a minor pathway gives rise to the C-N bond cleavage. But, neither acetic acid nor oxalic acid was detected. Opposite to the glycine molecule, the major nitrogen inorganic endproduct of acetamide oxidation is the nitrate ion since we showed that the oxamic acid is slowly mineralized into nitrates (Figure 8). The deactivating effect of the carboxyl group (37) of the oxamic acid molecule directs the H-atom abstraction on the deprotonated amino group. This reaction is likely followed by the oxidation of the nitrogen before the C-N bond cleavage. However, in the pH range of 8, oxamic acid is mostly present as a zwitterion (pKa2 ) 11.8), and in a real water matrix trace concentrations will require for their oxidation a 100 times higher OH° dosage than most organic compounds. In water treatment practice, such high OH° radical dosages will never be achieved. Therefore, even if it was shown that OH°-based advanced oxidation processes can originate nitrates, the oxidation of simple amides will not constitute a significant source. More attention must be drawn to amines and complex molecules such as peptides and aminopolycarboxylic acids with one C-H group in the R-position since they will be subject like glycine to C-N bond cleavage and then will represent one potential source of ammonium ions during oxidation by OH°-based advanced oxidation processes. From the present study performed on simple amino acids, the importance of the pH was shown considering its influence on the protonation of the amino group and therefore on the site of OH° radical attack. Additionally, the role of not only oxygen but also hydrogen peroxide in the electron exchange with the reducing C-centered radicals must be taken into account in waters as they act on the formation of the imino intermediates before hydrolysis and consequently bond cleavage.

Acknowledgments The authors are indebted to Dr. Peter Gehringer from the Austrian Research Center of Seibersdorf for giving them the opportunity to perform experiments with the 60Co source.

Literature Cited (1) Anderson, H. A.; Bick, W.; Hepburn, A.; Stewart, M. In Humic Substances II- In search of structure; Hayes, M. H. B., McCarthy, P., Malcolm, R. L., Swift, R. S., Eds.; Wiley & Sons: 1989; Chapter 8, pp 223-253. (2) McCarty, P. L. Proceedings, Wastewater Reclamation and Reuse Workshop; June 25-27, 1970, Lake Tahoe, CA, pp 226-241. (3) Stanbro, W. D.; Smith, W. D. Environ. Sci. Technol. 1979, 13, 446. (4) Gould, J. P.; Richards, J. T.; Miles, M. G. Water Res. 1984, 18, 991. (5) De Leer, W. B.; Baggerman, T.; Van Schalk, P. Environ. Sci. Technol. 1986, 20, 1218. (6) Peters, J. B.; De Leer, W. B.; Galan, L. Environ. Sci. Technol. 1990, 24, 81. (7) Hoigne´, J.; Bader, H. Water Res. 1983, 17, 185. (8) Elmghari-Tabib, M.; Laplanche, A.; Venien, F.; Martin, G. Water Res. 1982, 16, 223. (9) Laplanche, A.; Le Cloirec, C.; Martin, G. Proceeding Wasser Berlin 1985, 69. (10) Duguet, J. P.; Jaulin, L.; Aurelle, Y.; Roques, H.; Le´geron, J. P. Ozone: Sci., Eng. 1980, 2, 105.

(11) Duguet, J. P.; Brodard, E.; Dussert, B.; Mallevialle, J. Ozone: Sci., Eng. 1985, 7, 241. (12) Le Lacheur, R. M.; Glaze, W. H. Environ. Sci. Technol. 1996, 30, 1072. (13) Berger, P.; Karpel Vel Leitner, N.; Dore´, M.; Legube, B. Water. Res. 1999, 33, 433. (14) Simic, M.; Neta, P.; Hayon, E. Int. J. Rad. Phys. Chem. 1971, 3, 309. (15) Neta, P.; Simic, M.; Hayon, E. J. Phys. Chem. 1970, 74, 1214. (16) Bonifacic, M.; Stefanic, I.; Hug, G. L.; Armstrong, D. A.; Asmus, K.-D. J. Am. Chem. Soc. 1998, 120, 9930. (17) Monig, J.; Chapman, R.; Asmus, K. D. J. Phys. Chem. 1985, 89, 3139. (18) Bonifacic, M.; Armstrong, D. A.; Carmichael, I.; Asmus, K.-D. J. Phys. Chem. 2000, 104, 643. (19) Weeks, B. M.; Garrison, W. M. Rad. Res. 1958, 9, 291. (20) Abramovitch, S.; Rabani, J. J. Phys. Chem. 1976, 80, 1562. (21) Ho¨bel, B.; von Sonntag, C. J. Chem. Soc., Perkin Trans. 2 1998, 509. (22) Buxton, G. V.; Greenstock, C.l.; Helman, W. P.; Ross, A. B. J. Phys. Chem. 1988, 17, 513. (23) Karpel Vel Leitner, N.; Dore´, M. J. Chim. Phys. 1994, 91, 503. (24) Gehringer, P.; Szinovatz, W.; Eschweiler, H.; Haberl, R. Ozone: Sci. Eng. 1997, 19, 157. (25) Karpel Vel Leitner, N.; Dore´, M. Water. Res. 1997, 31, 1383.

(26) Al-Scheikhly, M. I.; Schuchmann, H.-P.; von Sonntag, C. Int. J. Radiat. Biol. 1985, 47, 457. (27) Wang, W.-F.; Schuchmann, M. N.; Bachler, V.; Schuchmann, H.-P.; von Sonntag, C. J. Phys. Chem. 1996, 100, 15843. (28) Maxwell, C. R.; Peterson, D. C. J. Am. Chem. Soc. 1957, 79, 5110. (29) Von Sonntag, C.; Schuchmann, H. P. Angew. Chem., Int. Ed. Engl. 1991, 30, 1229. (30) Jayson, G. G.; Scholes, G.; Weiss, J. Chemical Action of Ionising Radiations in Solution-Part XIV 1955, 3248, 2594. (31) Hoigne´, J.; Bader, H. Environ. Sci. Technol. 1978, 12, 79. (32) Neta, P.; Maruthamuthu, P.; Carton, P. M.; Fessenden, R. W. J. Phys. Chem. 1978, 82, 1875. (33) Hayon, E.; Ibata, T.; Lichtin, N. N.; Simic, M. J. Am. Chem. Soc. 1970, 92, 3898. (34) Holian, J.; Garrison, W. M. J. Phys. Chem. 1968, 72, 4721. (35) Smith, P.; Wood, P. B. Can. J. Chem. 1966, 44, 3085. (36) Livingston, R.; Zeldes, H. J. Chem. Phys. 1967, 47, 4173. (37) Dixon, W. T.; Norman, R. O. C.; Buley, A. L. J. Chem. Soc. 1964, 3625.

Received for review April 25, 2001. Revised manuscript received February 21, 2002. Accepted April 22, 2002. ES0101173

VOL. 36, NO. 14, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3089