Environ. Sci. Technol. 2008, 42, 8381–8387
Assessing the Redox Reactivity of Structural Iron in Smectites Using Nitroaromatic Compounds As Kinetic Probes ANKE NEUMANN,† T H O M A S B . H O F S T E T T E R , * ,† ¨ SSI,† OLAF A. CIRPKA,‡ MAJA LU SABINE PETIT,§ AND ´ P. SCHWARZENBACH† RENE Institute of Biogeochemistry and Pollutant Dynamics (IBP), ETH Zurich, Switzerland, Swiss Federal Institute of Aquatic Science and Technology, CH 8600 Dubendorf, Switzerland, and HydrASA, UMR CNRS 6532, Universite´ de Poitiers, 86022 Poitiers Cedex, France
Received July 3, 2008. Revised manuscript received September 2, 2008. Accepted September 3, 2008.
Structural Fe(II) in clay minerals is an important source of electron equivalents for the reductive transformation of contaminants in anoxic environments. We investigated which factors control the reactivity of Fe(II) in smectites including total Fe content, Fe(II)/total Fe ratio, and excess negative charge localization using 10 nitroaromatic compounds (NACs) as reactive probe molecules. Based on evidence from this work and previous spectroscopic studies on Fe redox reactions in ironrich smectites, we propose a kinetic model for quantifying the reactivity, abundance, and interconversion rates of two distinct Fe(II) sites in the minerals’ octahedral sheet. Excellent agreement between observed biphasic NAC reduction kinetics and model fits points toward existence of two types of Fe(II) sites exhibiting reactivities that differ by 3 orders of magnitude in iron-rich ferruginous smectite (SWa-1) and O¨lberg montmorillonite. Low structural Fe content, as found in Wyoming montmorillonite (SWy-2), impedes the formation of highly reactive Fe sites and results in pseudo-first order kinetics of NAC reduction that originate from the presence of a single type of Fe(II) species of even lower reactivity. Similar correlations of one-electron reduction potentials of the NACs vs their second order reduction rate constants for all smectite suspensions suggest that contaminant-Fe(II) interactions were identical in all smectite minerals.
Introduction In anoxic freshwater environments, biogeochemical processes provide a variety of mineral-bound Fe(II) species, which are of great relevance for the reductive transformation of organic pollutants including solvents, pesticides, and explosive residues (1-4). Such abiotic reduction of organic contaminants by Fe(II) generally leads to compounds that can be further biodegraded. However, several cases have * Corresponding author phone: +41 44 632 83 28; fax: +41 44 633 11 22; e-mail:
[email protected]. † Institute of Biogeochemistry and Pollutant Dynamics (IBP). ‡ Swiss Federal Institute of Aquatic Science and Technology. § Universite´ de Poitiers. 10.1021/es801840x CCC: $40.75
Published on Web 10/22/2008
2008 American Chemical Society
been reported in which products exhibit equal or even greater (eco-)toxicity than the parent compound. Thus, understanding the reactivity of mineral-bound Fe(II) species is crucial for assessing the risk of water resource contamination from organic pollutants. Fe(II) species associated with clay minerals are of particular interest because clay minerals are ubiquitously present in subsurface environments, most clay minerals contain some iron, and structural Fe(II) can be formed from microbial Fe(III) reduction (5). Finally, in contrast to Fe(III)oxy(hydr)oxides (6), Fe(III) in clay minerals is less susceptible to reductive dissolution (7, 8) thus providing a renewable source of reduction equivalents for contaminant transformation (9). Previous studies found that structural Fe(II) species in the octahedral sheet of smectites were the predominant reductants of contaminants such as nitroaromatic compounds (NACs 10, 11), technetium, Tc(VII) (12), or uranium, U(VI) (13), whereas Fe(II) surface hydroxyl complexes did not contribute significantly to the overall transformation. In earlier work, we showed that NACs not only represent an important class of contaminants that can be transformed by Fe(II) of clay minerals but can also be used as reactive probe molecules for investigating the redox activity of structural iron. While the reduction kinetics of NACs followed pseudo-first order behavior in suspensions of smectites containing small amounts of Fe, NAC reaction kinetics were biphasic in suspension of iron-rich smectites. This observation suggests that additional processes can contribute to the overall rate of transformation and/or that sites of different Fe(II) reactivity may exist. We hypothesized that electron transfer processes within the octahedral sheets of the smectite and the associated Fe migration affected the observed rates of NAC disappearance. However, only two NAC isomers (2- and 4-acetylnitrobenzene) were investigated, and it is unclear to what extent the rates of contaminant reduction were determined by the compound’s susceptibility for reduction, its affinity for reactive mineral surfaces, or by intrinsic properties of different reactive sites of Fe(II) bearing smectites. Thus, for a given clay mineral, extrapolations of reduction rates from one compound to a structurally similar one are hampered. In addition, it is not known how rates of reduction of one contaminant compare for different clay minerals exhibiting, for example, different structural Fe contents or variable Fe(II) to total Fe ratios. Spectroscopic studies evaluating the mechanism of structural Fe(III) reduction and Fe(II) reoxidation in ironrich smectites indeed proposed the formation of different Fe(II)/Fe(III) clusters in the octahedral sheet upon changes of Fe redox state (14-20). Investigation of absorption bands for Fe(II)-Fe(III) intervalence electron transfer led to the conclusion that Fe(III) is reduced in a way that Fe ions must occupy adjacent octahedral sites and must be of different valence (15, 17). Fe(II)-O-Fe(II) entities were postulated to form as soon as all Fe(III)-O-Fe(III) groups are partly reduced to Fe(II)-O-Fe(III). This process is accompanied by the migration of Fe(II) ions within the octahedral sheet yielding trioctahedral Fe(II) clusters which are separated by domains of vacancies (21). Additionally, the mechanism of Fe(III) reduction depends on the distribution of cations over transand cis-octahedral sites (19). Iron redox activity can thus be expected to vary for different smectites (e.g., nontronites, montmorillonites) as a consequence of mineral properties such as total Fe content, distribution of cations within the structure, or location of the excess charge. Therefore, quantifying contaminant reduction kinetics by structural Fe(II) in smectites should take into account (i) that at least two types of Fe(II) sites exhibiting different reactivity exist, (ii) that their relative abundance depends on the Fe(II)/Fe(III) VOL. 42, NO. 22, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8381
TABLE 1. Names, Acronyms, And Total Iron Contents of the Smectites Used in This Studya location of negative excess charge
total Fe content [wt %]
Wyoming montmorillonite (SWy-2) ferruginous smectite (SWa-1)
octahedral tetrahedral
2.8 b 12.6 b
¨ lberg montmorillonite O
octahedral
12.2 d
smectite
structural Fe(II) concentration [mol L-1] Na+ Na+ K+ Na+ K+
(3.34-3.87) × 10-3 c 6.33 × 10-3 6.16 × 10-3 3.22 × 10-3 5.40 × 10-3
Fe(II)/Fe(tot) [%] Na+ Na+ K+ Na+ K+
75-90 c 78 70 42 52
a Fe(II) concentrations and Fe(II)/Fe(tot) ratios refer to the conditions in the experimental reactors of Na+- and K+-homoionized minerals. b Determined as described in ref 11. c Range of Fe(II) concentrations in SWy-2 from different stock suspensions. d Determined according to a modified 1,10-phenanthroline method (25, 26).
ratio and total Fe content of the smectite, (iii) that structural rearrangements might induce interconversion of different Fe(II) sites during (re)oxidation, and (iv) that besides the degree of Fe(II) reduction and total Fe content other smectite properties might influence structural Fe(II) reactivity. The objective of this study was to evaluate contaminant reduction kinetics with regard to the above-mentioned factors determining octahedral Fe(II) reactivity in smectites using a quantitative kinetic model. Using NACs as reactive probe compounds, we compared the Fe(II) reactivity of chemically reduced, iron-rich montmorillonite (montmorillonite from ¨ lberg (22)) and ferruginous smectite (SWa-1), both conO taining approximately the same amount of total Fe (13 wt %, Table 1), with that of a montmorillonite of average Fe content (Wyoming montmorillonite, SWy-2, 3 wt % Fe). The two ironrich smectites differ predominantly in the position of the excess charge, which is located in the tetrahedral sheet in the case of ferruginous smectite (23) and in the octahedral ¨ lberg montmorillonite (22). These sheet in the case of O minerals were used to study the two types of Fe(II) sites exhibiting different reactivity and the effects of clay mineral properties such as excess charge localization and degree of Fe reduction on the rate of electron transfer from Fe(II) to the NACs. Experiments with reduced Wyoming montmorillonite, in contrast, provided a reference case for randomly distributed structural Fe(II), which, owing to the lower total Fe content, is not subject to Fe clustering and formation of high and low reactivity sites. To assess the contribution of compound-specific properties to the rates of organic contaminant reduction by Fe(II) in different smectites, we quantified the reaction kinetics of a set of 10 NACs exhibiting different intrinsic reactivities with regard to reduction, as indicated by their first electron potentials (E1h) varying from -590 to -260 mV. The chosen set of organic compounds and smectites allows for a quantitative comparative evaluation of structural Fe(II) reactivity and contaminant transformation rates between different smectites.
Materials and Methods Kinetic Experiments. For a complete list of chemicals used in this study see the Supporting Information (SI). Suspensions containing reduced Wyoming montmorillonite (SWy-2), ¨ lberg reduced ferruginous smectite (SWa-1), or reduced O ¨ lberg (22)) montmorillonite (iron-rich montmorillonite from O were prepared under anoxic conditions as described previously (11). Suspensions of clay mineral were reduced according to a modified citrate-bicarbonate-dithionite method (24) followed by cation exchange with 1 M KCl or NaCl. Homoionized smectite stock solutions were diluted to yield ¨ lberg) g L-1 of reduced 3.8 (SWa-1), 8.1 (SWy-2), or 4.0 (O mineral to ensure approximately the same Fe(II) concentration, using anoxic solutions of 10 mM MOPS (morpholinopropane sulfonic acid) buffered at pH 7.5 ( 0.1. Reactors contained 20 mL smectite suspension each and were kept in the dark on a roller apparatus at 25 ( 1 °C during the 8382
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 22, 2008
experiments. All experiments were performed in triplicates. Kinetic experiments were initiated by adding methanolic NAC stock solution to the anoxic reactors resulting in initial NAC concentration of 50 µM (methanol content
