Variability of Nitrogen Isotope Fractionation during the Reduction of


Oct 15, 2008 - Fe(II)L2. 6-) over the pH range from 3 to 12 and variable reductant .... and 2,4,6-trinitrotoluene (2,4,6-TNT) by AH2QDS was com- plete...
0 downloads 0 Views 971KB Size


Environ. Sci. Technol. 2008, 42, 8352–8359

Variability of Nitrogen Isotope Fractionation during the Reduction of Nitroaromatic Compounds with Dissolved Reductants ´ E. HARTENBACH,† AKANE T H O M A S B . H O F S T E T T E R , * ,† MICHAEL AESCHBACHER,† MICHAEL SANDER,† DONGWOOK KIM,‡ TIMOTHY J. STRATHMANN,‡ WILLIAM A. ARNOLD,§ CHRISTOPHER J. CRAMER,| AND ´ P. SCHWARZENBACH† RENE Institute of Biogeochemistry and Pollutant Dynamics (IBP), ETH Zurich, Universita¨tstrasse 16, 8092 Zurich, Switzerland, Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign, 205 North Mathews, Urbana, Illinois 61801, Department of Civil Engineering, University of Minnesota, 500 Pillsbury Dr. SE, Minneapolis, Minnesota, and Department of Chemistry and Supercomputing Institute, University of Minnesota, 207 Pleasant St. SE, Minneapolis, Minnesota 55455

Received April 17, 2008. Revised manuscript received August 13, 2008. Accepted August 15, 2008.

Compound-specific nitrogen isotope analysis was shown to be a promising tool for the quantitative assessment of abiotic reduction of nitroaromatic contaminants (NACs) under anoxic conditions. To assess the magnitude and variability of 15N fractionation for reactions with dissolved reductants, we investigated the reduction of a series of NACs with a model quinone (anthrahydroquinone-2,6-disulfonate monophenolate; AHQDS-) and a Fe(II)-catechol complex (1:2 Fe(II)-tiron complex; Fe(II)L26-) over the pH range from 3 to 12 and variable reductant concentrations. Apparent kinetic isotope effects, AKIEN, for the reduction of four mononitroaromatic compounds by AHQDS- ranged from 1.039 ( 0.003 to 1.045 ( 0.002 (average(1σ),consistentwithpreviousresultsforvariousmineralbound reductants. 15N fractionation for reduction of 1,2dinitrobenzene and 2,4,6-trinitrotoluene by AHQDS- and that of 4-chloronitrobenzene by Fe(II)L26-, however, showed substantial variability in AKIEN-values which decreased from 1.043 to 1.010 with increasing pH. We hypothesize that the isotopesensitive and rate-limiting step of the overall NAC reduction can shift from the dehydration of substituted N,N-dihydroxyanilines (large 15N fractionation upon N-O bond cleavage) to protonation or reduction of nitroaromatic radical anions (small 15N isotope effect upon electron transfer) consistent with calculations of semiclassical 15N isotope effects. Our results

* Corresponding author phone: +41 44 632 83 28; fax: +41 44 633 11 22; e-mail: [email protected] † ETH Zurich. ‡ University of Illinois at Urbana-Champaign. § Department of Civil Engineering, University of Minnesota. | Department of Chemistry and Supercomputing Institute, University of Minnesota. 8352

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 22, 2008

imply that a quantitative assessment of NAC reduction using compound-specific isotope analysis (CSIA) might need to account for homogeneous and heterogeneous reactions separately.

Introduction Nitrogen isotope analysis has great potential for assessing transformation processes of N-containing organic contaminants including nitroaromatic compounds (NACs 1-4), an important class of soil and groundwater contaminants (5). We have recently shown for the abiotic reduction of NACs by mineral-bound Fe(II) species that the transformation of the aromatic nitro group to the nitroso intermediate is accompanied by large 15N fractionation (1-3). In all systems investigated, the corresponding apparent kinetic isotope effects, AKIEN, were found to be quite consistent between 1.03 and 1.04, indicating that light isotopologues (containing 14NO2) react 3-4% faster than the heavy isotopologues (containing 15 NO2). In contrast to the overall rates of reduction, 15N isotope effects were insensitive to aromatic substituents in reactions of the NACs with Fe(II) adsorbed to Fe(III) oxides or structural Fe(II) in reduced clay minerals (6, 7). Therefore, we hypothesized that the corresponding bulk 15N enrichment factors,  , are likely representative of the N abiotic reduction of NACs and can hence be used to quantify the extent of transformation for a wide variety of nitroaromatic contaminants in complex subsurface environments. Besides solid-bound reductants, however, NACs can also be transformed by a variety of homogeneous species including dissolved natural organic matter (8), naptho- and anthraquinones in the presence of H2S (9, 10), and Fe(II) complexes with catechol and organothiol ligands (11). While the overall rates of NAC reduction generally depend on the speciation of the reductant, reactions rates determined for homogeneous reductants are more sensitive to the electron-accepting properties of the NACs (8, 10, 12-15). With reduced naphthoquinone species, for example, changing pH, aromatic NAC substituent, or reductant concentration results in variation of reduction rate constants by orders of magnitude. The only available AKIEN-values for reduction of 2- and 4-chloronitrobenzene with homogeneous reductants (i.e., 8-hydroxy-1,4-naphthoquinone (juglone) in solutions of H2S (1)) were consistent with observations reported for heterogeneous reductants. This data set, however, is too limited for drawing general conclusions on the variability of 15N fractionation with homogeneous reductants. It remains unclear from our previous studies on 15N fractionation whether this strong variability of NAC transformation rates bears consequences for the variability of 15N isotope effects. The objective of this study was to provide AKIEN-values for NACs reduced by homogeneous reductants that allow for a more comprehensive evaluation of NAC reduction in anoxic environments with compound-specific 15N analysis. To assess the variability of 15N kinetic isotope effects we investigated the 15N fractionation over a wide range of experimental conditions. From our earlier studies, we proposed that the 15N fractionation occurs predominantly during the cleavage of the first N-O bond of substituted N,N-dihydroxyanilines after a series of e-- and H+-transfers (3). To explore whether these e-- and H+-transfers influence the overall 15N fractionation, in this study, we compared one- and two-electron transfer agents as homogeneous model reductants, that is Fe(II)-complexes with tiron (catechol-3,5-disulfonate) and 9,10-anthrahydroquinone10.1021/es801063u CCC: $40.75

 2008 American Chemical Society

Published on Web 10/15/2008

FIGURE 1. 15N fractionation during reduction of 2-nitrotoluene in solutions of AH2QDS at pH 7 using two different experimental setups vs fraction of remaining reactant (c/c0, panel a). Circles represent reactors in which 2-nitrotoluene reduction was followed in the presence of excess reductant concentrations compared to squares, which represent reactors in which the reductant concentration was limited to achieve fractional conversion of 2-nitrotoluene. The linearized 15N enrichment calculated according to eq 1 for the two experimental setups is shown in panel b (uncertainty of the slope represents ( σ); for 15N enrichment factors see entries 4 and 15 of Table 1. 2,6-disulfonate (AH2QDS), respectively. As was shown previously in studies on the kinetics of NAC reduction by Fe(II)-tiron (11) and AH2QDS (9), rates strongly correlated with the concentration of the Fe(II)L26- species (L4-)fully deprotonated tiron) and monophenolate species AHQDS-, respectively. By varying either reductant concentration or pH, that is by restricting the amount of available electrons or protons, we investigated the relevance of e-- and H+transfer kinetics in modulating 15N fractionation during NAC reduction. Similarly, substituent effects on 15N fractionation were investigated because the e- and H+accepting properties of NACs and their reaction intermediates differ with aromatic substituents, thus providing evidence for the origins of variable 15N fractionation.

Experimental Section For a complete list of chemicals, suppliers, and purities see the Supporting Information (SI). Hydroquinone Reductants. The reductions of NACs in solutions containing 9,10-anthrahydroquinonone-2,6-disulfonate acid (AH2QDS, reduced hydroquinone form) were conducted in an anoxic glovebox (N2 atmosphere at 25 ( 1 °C) over the pH range from 3 to 12 using 0.1 M buffer solutions of citrate, acetate, phosphate, borate, or carbonate in 0.1 M KCl. AH2QDS was generated via electrochemical reduction of the disodium salt of 9,10-anthraquinone-2,6-disulfonate (AQDS) following a procedure adapted from ref 16. Briefly, solid AQDS was transferred into the glovebox and suspended in the desired buffer (final AQDS concentrations were 10 to 28 mM). Dissolution was facilitated by immersion in an ultrasonic bath at temperature up to 70 °C outside the glovebox. Reduction of AQDS was carried out in a bulk electrolysis cell (Bioanalytical Systems Inc.) which consisted of a glass vessel closed with a Teflon cover including cathode (reticulated vitreous carbon electrode), anode (coiled platinum wire), and reference electrode (Ag/AgCl). The cathode compartment was separated from the anode compartment by a glass frit, thus preventing the reoxidation of AH2QDS. Aliquots of 90 mL AQDS solution were reduced by stepwise lowering the cathode potential using an Autolab PG 302 instrument (Eco Chemie B.V.). Final applied potentials ranged from -0.55 V (pH 3) to -0.8 V (pH 12). Total time of reduction was 2 and 6 h for solutions containing 10 mM and 30 mM AQDS, respectively. Complete reduction of AQDS to AH2QDS was verified by coulometry and absorption measurements at 328 nm.

Reduction of 2- and 4-nitrotoluene and 2- and 4-chloronitrobenzene in solutions of AH2QDS was performed in sealed 20 mL amber flasks, which were mixed with a magnetic stirrer. For the isotope ratio measurements, seven separate reactors were prepared containing 20 mL of AH2QDS at the desired pH and AH2QDS concentration. Reactions were initiated by addition of 200 µL of 40 mM methanolic NAC solution resulting in initial NAC concentrations of 400 µM. At given time intervals, the NAC reduction was quenched by addition of 0.5 M hexacyanoferrate. For each experiment, an AH2QDS-free buffer solution was spiked with the NAC and hexacyanoferrate. This blank reactor served as control and was used to determine the NAC initial concentration and the corresponding δ15N value. From each reactor, 0.5 mL was used for concentration measurements by HPLC, and the remaining solution was stored in the dark at 4 °C until δ15N analysis. Because the reduction of 1,2-dinitrobenzene (1,2-DNB) and 2,4,6-trinitrotoluene (2,4,6-TNT) by AH2QDS was complete within seconds, the degree of NAC conversion was controlled by limiting the amount of AH2QDS added to the reactor (Table 1, c(AH2QDS) is “variable” because AH2QDS is consumed during the reaction with the NAC and the AH2QDS concentration is, therefore, not constant). According to this procedure, seven reactors were prepared by diluting the required amount of AH2QDS with buffer to yield a final volume of 20 mL, and the reaction was initiated by addition of methanolic NAC stock solution. All other steps were identical to that of mononitroaromatic compounds. This approach did not lead to experimental artifacts on 15N fractionation. As shown in Figure 1, experiments conducted with 2-nitrotoluene at pH 7 according to this setup showed an almost identical 15N fractionation within one standard deviation (N ) -40 ( 1‰) as that measured from a kinetic experiment involving reaction with excess AH2QDS and quenching aliquots with 0.5 M hexacyanoferrate at different reaction times (N ) -38 ( 1‰, Figure 1). Catechol Ligand-Complexed FeII Reductant. 4-Chloronitrobenzene reduction by Fe(II)-tiron (catechol-3,5-disulfonate) complexes were conducted under strict anoxic conditions according to ref 11. In brief, reactions were performed in sealed 40 mL serum flasks with a total solution volume of 35 mL. Like 1,2-DNB and 2,4,6-TNT experiments with AH2QDS, the degree of 4-chloronitrobenzene conversion (initial concentration 150 µM) was controlled by varying the amount of Fe(II) (0-1 mM) added to a series of reactors VOL. 42, NO. 22, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8353

TABLE 1. Nitroaromatic Compounds (NACs), Pseudo-First-Order Reduction Rate Constants, kobs, Second-Order Rate Constant for NAC Reduction by 2,6-Disulfonate Anthrahydroquinone Monophenolate, kAHQDS, Bulk 15N Enrichment Factors, EN, and Apparent 15N Kinetic Isotope Effect, AKIEN AH2QDS entry no.

NACa

pH

c(AH2QDS)tot (mM)b

kobs (s-1)

kAHQDS (M-1 s-1)c

EN (‰)d

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

2-CH3-NB 2-CH3-NB 2-CH3-NB 2-CH3-NB 2-CH3-NB 2-CH3-NB 2-CH3-NB 2-CH3-NB 2-CH3-NB 2-CH3-NB 2-CH3-NB 4-CH3-NB 2-Cl-NB 4-Cl-NB 2-CH3-NB 1,2-DNB 1,2-DNB 1,2-DNB 2,4,6-TNT 2,4,6-TNT

3 5 6 7 8 9 10 11 12 5 5 5 5 5 7 5 6 7 5 7

3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 10.1 28.3 3.0 3.0 3.0 variable variable variable variable variable variable

1.06 · 10-6 5.63 · 10-5 5.61 · 10-4 2.83 · 10-3 9.22 · 10-3 9.42 · 10-3 8.70 · 10-3 2.69 · 10-3 4.29 · 10-4 2.69 · 10-4 1.51 · 10-3 6.27 · 10-4 2.44 · 10-3 9.20 · 10-3 n.d.e n.d.e n.d.e n.d.e n.d.e n.d.e

1.11 · 10-1 5.95 6.10 3.93 4.05 3.34 3.82 3.73 4.67 8.46 1.69 6.63 · 101 2.58 · 102 9.72 · 102 n.d.e 2.55 · 105f n.d.e n.d.e 1.60 · 105f n.d.e

-43.3 ( 0.3 -38.9 ( 0.8 -37.7 ( 1.0 -38.1 ( 0.8 -40.5 ( 0.7 -40.2 ( 0.4 -41.2 ( 0.7 -41.3 ( 0.5 -43.1 ( 0.8 -41.9 ( 0.3 -39.1 ( 0.8 -39.5 ( 0.5 -40.4 ( 0.2 -37.1 ( 0.4 -40.3 ( 0.6 -17.1 ( 0.3 -13.6 ( 0.1 -5.3 ( 0.5 -8.6 ( 0.4 -3.4 ( 0.2

Fe(II)-tiron entry no.

NACa

pH

c(tiron)tot (mM)

kobs (s-1)I

c(FeL26-) (M)j

EN (‰)d

4-Cl-NB 4-Cl-NB 4-Cl-NB 4-Cl-NB 4-Cl-NB 4-Cl-NB 4-Cl-NB 4-Cl-NB 4-Cl-NB

6.0 6.5 7.0 7.5 8.0 8.5 9.0 7.0 7.0

8.15 · 10-4 4.72 · 10-2 1.24 · 100 1.02 · 101 2.56 · 101 3.39 · 101 3.68 · 101 1.09 · 10-1 6.86 · 100

2.14 · 10-8 1.24 · 10-6 3.25 · 10-5 2.67 · 10-4 6.71 · 10-4 8.90 · 10-4 9.65 · 10-4 2.85 · 10-6 1.80 · 10-4

-34.6 ( 1.3 -30.3 ( 1.3 -28.1 ( 0.4 -20.7 ( 1.6 -13.3 ( 3.0 -12.1 ( 0.5 -13.7 ( 1.2 -41.2 ( 4.4 -17.5 ( 1.3

21 22 23 24 25 26 27 28 29

10 10 10 10 10 10 10 2 50

AKIEN (-)e 1.0453 ( 0.0003 1.0405 ( 0.0009 1.0392 ( 0.0011 1.0396 ( 0.0008 1.0423 ( 0.0008 1.0419 ( 0.0004 1.0429 ( 0.0007 1.0431 ( 0.0006 1.0450 ( 0.0008 1.0438 ( 0.0003 1.0407 ( 0.0008 1.0412 ( 0.0005 1.0421 ( 0.0002 1.0385 ( 0.0004 1.0420 ( 0.0006 1.0354 ( 0.0007g 1.0281 ( 0.0003g 1.0107 ( 0.0010g 1.0265 ( 0.0011h 1.0100 ( 0.0005h AKIEN (-)e 1.0358 ( 0.0013 1.0312 ( 0.0013 1.0289 ( 0.0004 1.0212 ( 0.0016 1.0135 ( 0.0030 1.0122 ( 0.0005 1.0139 ( 0.0012 1.0429 ( 0.0044 1.0178 ( 0.0013

a Abbreviations: 2-CH3-NB (2-nitrotoluene), 4-CH3-NB (4-nitrotoluene), 2-Cl-NB (2-chloronitrobenzene), 4-Cl-NB (4-chloronitrobenzene), 1,2-dinitrobenzene (1,2-DNB), 2,4,6-trinitrotoluene (2,4,6-TNT). b Total AQDS concentration in solution. c kAHQDS ) kobs/[AHQDS-]; for 2-CH3-NB kAHQDS equals 3.62 M-1 s-1 from linear regression of kobs vs [AHQDS-] for all pH-values. d Uncertainty represents ( 1 standard deviation. e n.d. ) Not determined. f Estimated from correlation of kAHQDS vs E1h/0.059, see Figure S2. g Calculated from eq 2 with n/x · z ) 2. h Calculated from eq 2 with n/x · z ) 3. I kobs ) kFeL2 [FeL2], kFeL2 ) 3.81 · 104 M-1s-1 (11). j Calculated concentration of the 1:2 Fe-tiron complex assuming [Fe2+]total ) 1 mM.

containing a constant total tiron concentration (2-50 mM), pH (6-9; 50 mM MES (2-(N-morpholino)ethanesulfonic acid), MOPS (3-(N-morpholino)ethane sulfonic acid), or TAPS (3-(tris(hydroxymethyl)methyl)-3-aminopropanesulfonicacid)), and ionic strength (0.25 M NaCl). Reactions were initiated by adding 350 µL of a 15 mM 4-chloronitrobenzene methanolic stock solution and were allowed to proceed to completion before analyzing for concentration and δ15N. For each isotopic experiment, a solution containing all constituents except Fe(II) was spiked with 4-chloronitrobenzene and served as a control for determining the initial concentration and δ15N value of the NAC. Analytical Procedures and Data Evaluation. Concentrations of NACs were determined by HPLC on a LC-18 reverse phase column with UV/vis detection as described previously (1). δ15N signatures of NACs were determined according to a method using solid-phase microextraction (SPME) coupled to a GC/C/IRMS (gas chromatography isotope-ratio mass spectrometry with combustion interface (17)). All δ15N values were derived from triplicate measurements with good precision (< (0.5‰, (1σ) and are reported relative to N2 in air as δ15NAir. To avoid uncertainty of δ15N due to instrument nonlinearity (18), aqueous samples were diluted to the level of the least concentrated NAC solution prior to analysis and measured at constant peak amplitudes between 0.8 and 2 V. 8354

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 22, 2008

Bulk 15N enrichment factors, N, of the NACs were derived from linear regression analysis of eq 1 (19) ln(δ15N + 1000) )

εN 1000

(

· ln(c) + ln

(δ15N0 + 1000) c0εN ⁄ 1000

)

(1)

where δ15N0 and δ15N are the initial 15N signatures of the NACs and its value during the reaction, and c0 and c are the substrate’s initial concentration and concentration following different extents of NAC reduction, respectively. Apparent kinetic isotope effects, AKIEN, were calculated according to eq 2 AKIEN )

(

1 1 + n ⁄ x · z · εN ⁄ 1000

)

(2)

where n, x, and z are correction factors accounting for isotopic dilution (n number of isotopic N atoms), number of reactive sites (x), and of reactive positions in intramolecular isotopic competition (z) (20). The overall correction factor (n/x · z) was unity for 2-/4-nitrotoluenes and 2-/4-chloronitrobenzenes and equaled 2 and 3 for 1,2-DNB and 2,4,6-TNT, respectively. Computational Methods. Calculations to determine kinetic isotope effects for the second electron transfer were performed with the Gaussian 03 electronic structure

FIGURE 2. Panel a: Logarithms of pseudo-first-order rate constants of 2-nitrotoluene reduction in solutions of AH2QDS (log kobs, empty squares) and logarithm of AHQDS- concentration (log [AHQDS-], filled squares) as a function of pH. Panel b: Apparent kinetic isotope effects, AKIEN, of 2-nitrotoluene reduction by AHQDS- as a function of the pH (circles) in solutions containing 3 mM total AQDS and in solutions of 10 and 28 mM total AQDS at pH 5 (diamonds and squares). Panel c, Logarithms of calculated pseudo-first-order rate constants 4-chloronitrobenzene reduction by Fe(II)-tiron complexes (log kobs, empty squares) at 10 mM total ligand concentration and logarithm of calculated FeL26- concentration (log [FeL26-], filled squares). Panel d, AKIEN of 4-chloronitrobenzene reduction by Fe(II)-tiron as function of the pH (diamonds); at pH 7, total tiron concentrations of 2 and 50 mM (circles and squares) were also considered. All uncertainties represent ( one standard deviation. program suite (21) as described in detail in the SI. Kinetic isotope effects were calculated following the formulation of Kavner et al. (22). Because the reaction is an electron transfer, activation free energies were determined using Marcus theory (23). The activation free energy ∆G* in Marcus theory is

(

l ∆G ∆G 2 + ∆G ) + 4 2 4λ /

)

(3)

where ∆G is the reaction driving force, and λ is the reorganization energy. The driving force (and thus the activation barrier and the KIE) is dependent on the reducing agent, so a series of ∆G values ranging from -0.6 to to 0.3 eV were surveyed. In addition, three different values of λ were considered, 100, 200, and 300 kJ/mol. Using eq 3, ∆G* values were determined for the 14N-containing species as a function of ∆G and λ. The activation free energy is sensitive to isotopic substitution because the 298 K driving force is sensitive to isotopic substitution. In principle, the difference in driving force for the two isotopes may be computed from isotopically sensitive molecular free energies determined by Gaussian 03 using standard statistical mechanical formulas associated with the ideal-gas, harmonic-oscillator, rigid-rotator approximation. Because of the output formatting employed by the program, however, an additional significant figure is available if one

instead employs the computed partition functions directly. In the latter case, the difference in driving forces may be computed as ∆G15N - ∆G14N ) -RT ln R

(4)

where R is the universal gas constant, T is temperature in K, and R is R)

Q15N

Q14N

Q14N

Q15N

dianion

dianion

radical anion

(5)

radical anion

and the Q values are the partition functions for the isotopologues of the product (dianion, Scheme 1 compound 3) and reactant (radical anion, Scheme 1, compound 1) as determined from the Gaussian 03 frequency calculations. Using eqs 3 and 4, activation energies were calculated for the 15N-containing species as a function of ∆G and λ. KIE values are then computed as 14N

KIE ) 15

k

N

k

)



MW14N MW15N

exp

(

∆G /15N - ∆G /14N RT

)

(6)

where the second term accounts for the change in activation free energy, while the first corrects for the change in collision frequency due to the change in molecular weight (MW). VOL. 42, NO. 22, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8355

FIGURE 3. (a) Linearized 15N enrichment behavior during the homogeneous reduction of 1,2-dinitrobezene (1,2-DNB) by AHQDS- at pH 5 (circles), pH 6 (squares), and pH 7 (triangles), lines represent the linear regression analysis used for the calculation of 15N enrichment factors, EN, according to eq 1. (b) Linearized 15N enrichment for 2,4,6-trinitrotoluene (2,4,6-TNT) reduction by AHQDS- at pH 5 (circles) and pH 7 (triangles) and corresponding linear regression analysis.

Results and Discussion NAC Reduction by Anthrahydroquinones. The pseudo-firstorder rate constant kobs of 2-nitrotoluene reduction in solutions of AH2QDS for the pH range 3 to 12 are shown in Figure 2a. Overall rates of NAC reduction varied over four orders of magnitude with pH, consistent with previous observations using other dissolved naphthoquinone reductants (10). Figure 2a shows that log kobs and the log of the anthrahydroquinone-2,6-monophenolate concentration (AHQDS-, Figure 2a (filled squares)) exhibited an almost identical dependence on solution pH. In fact, the rate constant, kobs, was strongly correlated with the AHQDSconcentration over the entire pH range (Figure S1), showing that the monophenolate is the predominant reductant of NACs in our experimental systems. Our data are consistent with previous observations in ref 9, albeit with a higher accuracy and over a wider pH range. Changes of pH modulate the concentration of the reactive species and thus indirectly affect the overall rate of reduction. Figure 2b shows the effect of the solution pH on the 15N fractionation. In contrast to the overall rates of NAC reduction, 15N fractionation of 2-nitrotoluene was constant over the entire pH range and insensitive of the total reduced anthrahydroquinone concentration (between 3 and 28 mM of AH2QDS). Measured bulk 15N enrichment factors, N, of 2-nitrotoluene were between -37.7 ( 1.0‰ and -43.3 ( 0.3‰ resulting in apparent kinetic isotope effects (AKIEN) between 1.0392 ( 0.0011 and 1.0453 ( 0.0003 (mean ( standard deviation). These values are very similar to what was found for 2-nitrotoluene reduction by mineral-bound Fe(II) species (1.0392 ( 0.0024 (3)). Note that for 2-nitrotoluene, 15N fractionation remained constant despite a change in H+ concentration of 9 orders of magnitude (Figure 2b) and variable total reductant concentration, and, thus, regardless of the corresponding change in the reduction potential of the system (24, 25). The constant AKIEN-values therefore suggest that for all chosen experimental conditions, NAC reduction proceeded by the same mechanism, that is via e- and H+ transfers from AHQDS- (see discussion below). Substituent Effects on Rates of NAC Reduction and 15N Fractionation. In AQDS solutions at pH 5, we measured the reduction rates and AKIEN-values of 2- and 4-nitrotoluene and 2- and 4-chloronitrobenzene. As found earlier for other dissolved reductants (NOM, juglone/H2S, Fe(II)-tiron (8, 10, 11, 15)), reduction rate constants (Table 1) were very sensitive to changes of the electron-accepting properties of the NACs (Figure S2). Note that reduction rates of 1,2-dinitrobenzene (1,2-DNB) and 2,4,6-trinitrotoluene (2,4,6-TNT) by AHQDS8356

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 22, 2008

were too fast to be experimentally accessible. While reduction rate constants of nitrotoluenes and chloronitrobenzenes varied by two orders of magnitude (Table 1, entries 2, 12-14), N-values were constant within a narrow range of a few per mil (i.e., -37.1 ( 0.4‰ to -40.4 ( 0.2‰). Corresponding AKIEN-values between 1.0385 ( 0.0004 and 1.0421 ( 0.0002 showed no substituent effects. This is in agreement with recent work in Fe(II)-mineral suspensions, where average AKIEN-values of 1.038 ( 0.001 for mono- and polynitroaromatic compounds were measured (2, 3). From the previously observed absence of substituent effects on 15N fractionation, we concluded that all NO2-groups in 1,2-DNB and 2,4,6-TNT are also equivalent with regard to 15N fractionation in homogeneous solution. Thus AKIEN-values of polynitroaromatic compounds (Table 1) were calculated by correcting the observed N-values (Figure 3) for isotopic dilution, the number of reactive sites, and intramolecular isotopic competition (eq 2). In contrast to mononitroaromatic compounds, AKIEN-values decreased for 1,2-DNB from 1.0354 ( 0.0007 at pH 5 to 1.0107 ( 0.0010 at pH 7 and for 2,4,6-TNT from 1.0265 ( 0.0011 (pH 5) to 1.0100 ( 0.0005 (pH 7). To evaluate whether decreasing 15N fractionation of 1,2DNB and 2,4,6-TNT with pH were caused by diffusion-limited reaction rates, second-order rate constants can be estimated from the correlation of log kAHQDS- vs the free energy changes of the first electron transfer to the NAC (E1h/0.059 V, Figure S2). This type of correlation has been used previously to quantify the sensitivity of reduction rate constants of a series of NACs toward different reductants (26). Resulting bimolecular rate constants of 1,2-DNB and TNT were 2.6 · 105 M-1s-1 and 1.6 · 105 M-1 s-1, respectively, that is clearly below diffusion-control (second-order rate constants (<109 M-1 s-1 (27)). Trends of decreasing AKIEN-values with increasing pH, moreover, also argue against any diffusion-control. The extent of masking of isotope fractionation (i.e, N-values tending toward less negative values and, finally, zero) by transportlimitation is expected to increase as the concentration of the reductant AHQDS- decreases due to larger diffusion distances at lower reductant concentrations. As illustrated by the data in Table 1 (entries no. 16-20), we observed the opposite trend. Smaller 15N fractionation correlates with increasing pH and, thus, with increasing concentration of AHQDS-species, indicating that decreasing AKIEN-values of 1,2-DNB and 2,4,6-TNT are likely due to other processes than diffusioncontrolled reactions (see further discussion below). 15N Fractionation during 4-Chloronitrobenzene Reduction by Fe(II)-Tiron. Figure 2c shows pseudo-first-order rate constants of 4-chloronitrobenzene reduction and concen-

SCHEME 1

trations of the reactive reductant species in solutions of Fe(II) and tiron (1:2 Fe(II)-tiron complex, Fe(II)L26- (11)) for pH values 6 to 9. Rate constants and FeL26--concentrations were calculated according to bimolecular rate constants and equilibrium constants derived previously (11). For the investigated pH range and the total ligand concentrations, 15N fractionation of 4-chloronitrobenzene varied between AKIEN-values of 1.0429 ( 0.0044 (2 mM total tiron concentration, pH 7) and 1.0122 ( 0.0005 (10 mM total tiron concentration, pH 8.5, Table 1). At constant ligand concentrations, AKIEN-values decreased from a maximum value of 1.04, which is in agreement with previously reported ones including those for AHQDS-, with increasing pH and leveled off at 1.01 for pH 8 and higher (Figure 2d). The decrease of 15N fractionation correlated with increasing rates of NAC reduction, achieved either via increasing pH or raising total ligand concentration. Masking of AKIEN-values by diffusion controlled reaction rates was ruled out on the basis of the bimolecular rate constant for 4-Cl-NB of 3.81 · 104 M-1s-1 (11). In contrast to reduced quinone species like AHQDS-, Fe(II)ligand complexes are one-electron reductants, such that the H+ required for NAC reduction has to originate from the solvent water. It is therefore likely that the kinetics of stepwise e-- and H+-transfers from different reductant molecules and water, respectively, contributed to the gradual decrease of the large isotope effects observed at lower pH when NAC reduction was slower. Effect of Electron and Proton Transfer Reactions on 15N Fractionation of NACs in Homogeneous Solution. On the basis of experimental data and theoretical considerations, we proposed recently that AKIEN-values for abiotic NAC reduction are about 1.04 if they are determined by the cleavage of an N-O bond of substituted N,N-dihydroxyanilines (5f6 in Scheme 1). Given that the maximum 15N fractionation measured for dissolved reductants agrees well with previous work, we hypothesize that the unprecedented variability of AKIEN-values reported here originates from the kinetics of e-- and H+-transfer pre-equilibria, which modulate the 15N kinetic isotope effects upon N-O bond cleavage. The decrease in AKIEN-values with increasing pH and increasing reductant concentration suggests that the availability of H+ and e- are key to changing the magnitude of 15N fractionation. On the one hand, protonation of the nitroaromatic radical anion (1 in Scheme 1) might have limited the overall NAC reduction and caused the AKIEN-values to decrease to values around 1.01. However, H+-transfer reactions to oxygen are usually near diffusion limit (28, 29), and it is unclear whether they could become rate-determining under the chosen experimental conditions. Isotope-sensitive changes in the N-O bonding upon protonation of oxygen atoms are likely too small to cause a significant isotope effects at the nitrogen atom but one could postulate that small 15N fractionation

reflects equilibrium isotope effects upon the first electron transfer. On the other hand, it has been found in the electrochemical reduction of NACs over a wide range of pHs (30-33) and in aprotic solvents (34) that the rate-limiting step of the NAC reduction to nitroso intermediates can shift depending on H+-availability. Although the protonation of the nitroaromatic radical anion (1 in Scheme 1) takes place prior to the second e--transfer (1f2f4, e-/ H+/e--pathway), in a reaction whose overall rate is limited by the dehydration of substituted N,N-dihydroxyanilines (5f6), the second e--transfer may become limiting when H+ are less available (i.e., either at high pH or in aprotic media, e-/e-/H+-pathway, 1f3f4). The latter has been invoked from the finding of more negative reduction potentials required to transfer a second e- prior to the first protonation (E2h, e-/e-/H+) compared to potentials for the second electron following transfer of a first proton (E2h, e-/H+/e-, refs 30-32, 34). A shift in the rate-limiting step of overall NAC reduction by Fe(II)-tiron complexes from the dehydration of 4-chloroN,N-dihydroxyaniline (5) to either protonation or reduction of 4-chloronitrobenzene radical anion (1) would both be consistent with the transition of AKIEN-values from 1.04 to 1.01 (Figure 2, Table 1). The AKIEN-values of 1.01 observed from pH 8 to 9 could thus represent first experimental evidence for the magnitude of 15N isotope effects upon electron transfer to aromatic NO2-groups, which are supposedly smaller than those involving a N-O bond cleavage. These numbers are in very good agreement with calculations of semiclassical 15N kinetic isotope effects for the second electron transfer 1f3 using Marcus theory, which were dependent on reaction driving force and reorganization energy, and resulted in values between 1.006 and 1.012 (Figure S3d). The decrease of 15N fractionation during NAC reduction by AHQDS- was observed exclusively for 1,2-DNB and 2,4,6TNT and exhibited the same trends as found in solutions of Fe(II)-tiron. We hypothesize that with AHQDS- as the reductant, similar shifts of rate-limiting step apply. In contrast to mono-NACs, radical anions of 1,2-DNB and 2,4,6-TNT are weaker bases (35, 36) thus making their protonation more difficult as pH increases. This tentative interpretation is corroborated (i) by the very similar slopes of correlations between AKIEN-values vs pH for the reduction of 4-chloronitrobenzene by Fe(II)-tiron (∆AKIEN/∆pH ) -0.011 ( 0.001) and 1,2-DNB/2,4,6-TNT reduction by AHQDS- (-0.012 ( 0.003), (ii) by very similar minimum AKIEN-values of 1.01 (Table 1), and (iii) by the fact that these minimum AKIENvalues match calculated semiclassical 15N kinetic isotope effects for the second electron transfer in the range of 1.005 to 1.012 (Figure S3e,f). The higher end of this range, which is in good quantitative agreement with measured values, is VOL. 42, NO. 22, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8357

associated with smaller reorganization energies and/or weaker driving forces. These conditions would be expected to prevail under our current experimental conditions. The calculated 15N isotope effects for all NACs investigated in this study (Figure S3) suggest that a decrease of 15N isotope effects to 1.01 might generally apply under experimental conditions where the overall reduction kinetics are limited by the electron transfer to the NACs. Our interpretation of low AKIEN-values for 1,2-DNB and 2,4,6-TNT also implies that AHQDS- would not supply the first H+ for NAC reduction if the reaction were sensitive to solution pH, in contrast to the lack of pH dependence for both rates and AKIEN-values of 2-nitrotoluene observed between pH 3 to 12. To verify contributions from H+ from solution to the reduction of polynitroaromatic compounds by AHQDS-, however, an alternative experimental setup for measurement of fast reactions is necessary for further study. Assessing Abiotic NAC Reduction in Anoxic Environments Using 15N CSIA. Knowledge on the extent and variability of 15N fractionation is pivotal for identifying and quantifying abiotic transformation of nitroaromatic contaminants with CSIA in anoxic soils and groundwaters. The combined evidence from this and earlier studies suggests that bulk 15N enrichment factors reflecting AKIEN-values of 1.04 at the reacting bonds are typical for NAC reduction provided that the reductants are associated with the solid phase (e.g., mineral-bound Fe(II) species, refs 26, 37-39). For reductions of NACs by dissolved species, however, 15N fractionations can vary with solution conditions. Our data illustrate that the differences in N-values can be significant in the typical environmental pH range, that is, 15N enrichment factors could be four times smaller than expected depending on the identity of the homogeneous reductant, solution conditions, and the type of nitroaromatic contaminant. Estimates of the extent of contaminant transformation assuming large 15N fractionation could in some cases significantly underestimate the amount of degraded contaminant. Further study with naturally occurring dissolved electron transfer mediators is necessary to evaluate the significance of varying isotope fractionation in the aqueous phase.

Acknowledgments

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

We are grateful for financial support from the Swiss NSF (200020-116447/1) to T.B.H. and from NSF (CHE0610183) to C.J.C. (18)

Supporting Information Available Chemicals, description of computational methods, and figures: (S1) correlation of pseudo-first-order rate constants of 2-nitrotoluene reduction with AHQDS- species concentration, (S2) correlation of second-order reduction rate constant with AHQDS- with E1h, and (S3) calculated 15N kinetic isotope effects for electron transfer to radical anions. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Hartenbach, A.; Hofstetter, T. B.; Berg, M.; Bolotin, J.; Schwarzenbach, R. P. Using nitrogen isotope fractionation to assess abiotic reduction of nitroaromatic compounds. Environ. Sci. Technol. 2006, 40, 7710–7716. (2) Tobler, N. B.; Hofstetter, T. B.; Schwarzenbach, R. P. Assessing iron-mediated oxidation of toluene and reduction of nitroaromatic contaminants in anoxic environments using compoundspecific isotope analysis. Environ. Sci. Technol. 2007, 41, 7773– 7780. (3) Hofstetter, T. B.; Neumann, A.; Arnold, W. A.; Hartenbach, A. E.; Bolotin, J.; Cramer, C. J.; Schwarzenbach, R. P. Substituent effects on nitrogen isotope fractionation during abiotic reduction of 8358

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 22, 2008

(19)

(20)

(21)

(22)

(23) (24)

nitroaromatic compounds. Environ. Sci. Technol. 2008, 42, 1997– 2003. Hofstetter, T. B.; Spain, J. C.; Nishino, S. F.; Bolotin, J.; Schwarzenbach, R. P. Identifying competing aerobic nitrobenzene biodegradation pathways using compound-specific isotope analysis. Environ. Sci. Technol. 2008, 42, 4764–4770. Spain, J. C.; Hughes, J. B.; Knackmuss, H.-J. Biodegradation of Nitroaromatic Compounds and Explosives; Lewis Publishers: Boca Raton, 2000; p 434. Hofstetter, T. B.; Neumann, A.; Schwarzenbach, R. P. Reduction of nitroaromatic compounds by Fe(II) species associated with iron-rich smectites. Environ. Sci. Technol. 2006, 40, 235–242. Hofstetter, T. B.; Schwarzenbach, R. P.; Haderlein, S. B. Reactivity of Fe(II) species associated with clay minerals. Environ. Sci. Technol. 2003, 37, 519–528. Dunnivant, F. M.; Schwarzenbach, R. P.; Macalady, D. L. Reduction of substituted nitrobenzenes in aqueous solutions containing natural organic matter. Environ. Sci. Technol. 1992, 26, 2133–2141. Tratnyek, P. G.; Macalady, D. L. Abiotic reduction of nitroaromatic pesticides in anaerobic laboratory systems. J. Agric. Food Chem. 1989, 37, 248–254. Schwarzenbach, R. P.; Stierli, R.; Lanz, K.; Zeyer, J. Quinone and iron porphyrin mediated reduction of nitroaromatic compounds in homogeneous aqueous solution. Environ. Sci. Technol. 1990, 24, 1566–1574. Naka, D.; Kim, D.; Strathmann, T. J. Abiotic reduction of nitroaromatic compounds by aqueous iron(II)-catechol complexes. Environ. Sci. Technol. 2006, 40, 3006–3012. Tobler, N. B.; Hofstetter, T. B.; Straub, K. L.; Fontana, D.; Schwarzenbach, R. P. Iron-mediated microbial oxidation and abiotic reduction of organic contaminants under anoxic conditions. Environ. Sci. Technol. 2007, 41, 7765–7772. Klausen, J.; Tro¨ber, S. P.; Haderlein, S. B.; Schwarzenbach, R. P. Reduction of substituted nitrobenzenes by Fe(II) in aqueous mineral suspensions. Environ. Sci. Technol. 1995, 29, 2396– 2404. Heijman, C. G.; Grieder, E.; Holliger, C.; Schwarzenbach, R. P. Reduction of nitroaromatic compounds coupled to microbial iron reduction in laboratory aquifer columns. Environ. Sci. Technol. 1995, 29, 775–783. Hofstetter, T. B.; Heijman, C. G.; Haderlein, S. B.; Holliger, C.; Schwarzenbach, R. P. Complete reduction of TNT and other (poly)nitroaromatic compounds under iron reducing subsurface conditions. Environ. Sci. Technol. 1999, 33, 1479– 1487. Kappler, A.; Haderlein, S. B. Natural organic matter as reductant for chlorinated aliphatic pollutants. Environ. Sci. Technol. 2003, 37, 2714–2719. Berg, M.; Bolotin, J.; Hofstetter, T. B. Compound-specific nitrogen and carbon isotope analysis of nitroaromatic compounds in aqueous samples using solid-phase microextraction coupled to GC/IRMS. Anal. Chem. 2007, 79, 2386– 2393. Sherwood Lollar, B.; Hirschorn, S. K.; Chartrand, M. M. G.; Lacrampe-Couloume, G. An approach for assessing total instrumental uncertainty in compound-specific carbon isotope analysis: Implications for environmental remediation studies. Anal. Chem. 2007, 79, 3469–3475. Scott, K. M.; Lu, X.; Cavanaugh, C. M.; Liu, J. S. Optimal methods for estimating kinetic isotope effects from different forms of the Rayleigh distillation equation. Geochim. Cosmochim. Acta 2004, 68, 433–442. Elsner, M.; Zwank, L.; Hunkeler, D.; Schwarzenbach, R. P. A new concept linking observable stable isotope fractionation to transformation pathways of organic pollutants. Environ. Sci. Technol. 2005, 39, 6896–6916. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A., et al. Gaussian 03, Revision E.01; Gaussian, Inc.: Wallingford, 2004. Kavner, A.; Bonet, F.; Shahar, A.; Simon, J.; Young, E. The isotopic effects of electron transfer: An explanation for Fe isotope fractionation in nature. Geochim. Cosmochim. Acta 2005, 69, 2971–2979. Marcus, R. A. Theory of oxidation-reduction reactions involving electron transfer. I. J. Chem. Phys. 1956, 24, 966–978. Gamage, R.; McQuillen, A. J.; Peake, B. M. Ultraviolet-visible and electron paramagnetic resonance spectroelectrochemical studies of the reduction products of some anthraquinone sulphonates in aqueous solutions. J. Chem. Soc., Faraday Trans. 1991, 87, 3653–3660.

(25) Bailey, S. I.; Ritchie, I. M.; Hewgill, F. R. The construction and use of potential-pH diagrams in organic oxidation-reduction reactions. J. Chem. Soc., Perkin Trans. 2 1983, 645–652. (26) Ru ¨gge, K.; Hofstetter, T. B.; Haderlein, S. B.; Bjerg, P. L.; Knudsen, S.; Zraunig, C.; Mosbæk, H.; Christensen, T. H. Characterization of predominant reductants in an anaerobic leachate-affected aquifer by nitroaromatic probe compounds. Environ. Sci. Technol. 1998, 32, 23–31. (27) Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental Organic Chemistry, 2nd ed.; John Wiley & Sons: New York, 2003; p 1311. (28) Eigen, M. Proton transfer, acid-base catalysis, and enzymatic hydrolysis. Part I: Elementary processes. Angew. Chem. Int. Ed. 1964, 3, 1–19. (29) Guthrie, J. P. Predicting the rates of proton transfer reactions: a simple model using equilibrium constants and distortion energies. J. Phys. Org. Chem. 1998, 11, 632–641. (30) Laviron, E.; Roullier, L. The reduction mechanism of aromatic nitro compounds in aqueous medium Part 1. Reduction to dihydroxylamines between pH 0 and 5. J. Electroanal. Chem. 1990, 288, 165–175. (31) Laviron, E.; Meunier-Prest, R.; Vallat, A.; Roullier, L.; Lacasse, R. The reduction mechanism of aromatic nitro compounds in aqueous medium Part 2. The reduction of 4-nitropyridine between H0)-6 and pH 9.6. J. Electroanal. Chem. 1992, 341, 227–255. (32) Lacasse, R.; Meunier-Prest, R.; Laviron, E.; Vallat, A. The reduction mechanism of aromatic nitro compounds In aqueous medium Part 3. The reduction of 4-nitropyridine-N-oxide

(33)

(34) (35) (36)

(37)

(38)

(39)

between H0)-6 And pH 9. J. Electroanal. Chem. 1993, 359, 223–239. Laviron, E.; Meunier-Prest, R.; Lacasse, R. The reductionmechanism of aromatic nitro-compounds in aqueous-medium Part 4. The reduction of p-nitrobenzophenone between H0)-5 and pH 14. J. Electroanal. Chem. 1994, 375, 263–274. Jensen, B. S.; Parker, V. D. Reversible anion radical-dianion redox equilibria involving ions of simple aromatic compounds. J. Chem. Soc., Chem. Commun. 1974, 367, 367–368. Lund, H. Cathodic reduction of nitro and related compounds. In Organic Electrochemistry, 4th ed.; Lund, H., Hammerich, O., Eds.; Marcel Dekker: New York, 2001; pp 379-409. Gru ¨ nbein, W.; Fojtik, A.; Henglein, A. Pulse radiolysis of absorption spectra and dissociation constants of short-lived semireduced aromatic nitro compounds. Z. Naturforsch., B: Chem. Sci. 1969, B 24, 1336–1338. Christensen, T. H.; Kjeldsen, P.; Bjerg, P. L.; Jensen, D. L.; Christensen, J. B.; Baun, A.; Albrechtsen, H. J.; Heron, C. Biogeochemistry of landfill leachate plumes. Appl. Geochem. 2001, 16, 659–718. Christensen, T. H.; Bjerg, P. L.; Banwart, S. A.; Jakobsen, R.; Heron, G.; Albrechtsen, H.-J. Characterization of redox conditions in groundwater contaminant plumes. J. Contam. Hydrol. 2000, 45, 165–241. Kenneke, J. F.; Weber, E. J. Reductive dehalogenation of halomethanes in iron- and sulfate- reducing sediments. 1. Reactivity pattern analysis. Environ. Sci. Technol. 2003, 37, 713–720.

ES801063U

VOL. 42, NO. 22, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8359