Identification of Active Agents for Tetrachloroethylene Degradation in

In DS/S, organic pollutants, such as tetrachloroethylene (PCE), are destroyed while inorganic contaminants are immobilized (5). PCE has been observed ...
0 downloads 0 Views 785KB Size
Environ. Sci. Technol. 2007, 41, 5824-5832

Identification of Active Agents for Tetrachloroethylene Degradation in Portland Cement Slurry Containing Ferrous Iron S A E B O M K O * ,† A N D B I L L B A T C H E L O R ‡ Department of Environmental Health, East Tennessee State University, Johnson City, Tennessee 37614; Department of Civil Engineering, Texas A&M University, College Station, Texas 77843

Experimental studies were designed to identify the active agents in Fe(II)-based degradative solidification/stabilization (Fe(II)-DS/S) that are responsible for the degradation of tetrachloroethylene (PCE) as well as the conditions that enhance the formation of these active agents. First, the conditions that lead to maximizing production of the active agents were identified by measuring the ability of various chemical mixtures to degrade PCE. Results showed that Fe(II), Fe(III), and Cl were the elements most closely associated with high degradation rates. In addition to elemental composition, unknown factors associated with the formation of solid phases could also be important in determining the extent of formation of active reducing agents. Second, instrumental analysis techniques (XRD, SEM, SEM-EDS) were used to identify compounds in chemical mixtures that were observed to have high activities for PCE degradation. SEM-EDS analysis indicated that Fe was associated with hexagonal particles, which is the typical shape of several AFm phases in hydrated Portland cement that are composed of calcium, aluminum/iron, hydroxide, and possibly other anions. No Fe-containing solid phases could be identified. Therefore, it appears that AFm phases are the most likely active agents for PCE degradation in mixtures containing Portland cement or its acid extract. Mixtures without cement did not form the same solid phases but were observed to form ferrous hydroxide as a major solid phase.

Introduction

destroyed while inorganic contaminants are immobilized (5). PCE has been observed to be degraded by pseudo-firstorder kinetics in slurries of Portland cement that contain Fe(II), and it has been observed to primarily follow a reductive elimination pathway, which is enhanced at high pH (5). When Portland cement is mixed with water, its four major phases of tricalcium silicate (C3S), dicalcium silicate (C2S), tricalcium aluminate (C3A), and tetracalcium aluminoferrite (C4AF) are completely changed as it sets and hardens (13, 14). However, cement hydration is more complicated than simply conversion of anhydrous cement phases to their corresponding hydrates. The hydration of C3A and C4AF forms mostly aluminite-ferrite-tri (AFt) and aluminate-ferritemono (AFm) phases. AFm phases are examples of compounds called layered double hydroxides (LDHs) (15). LDH are compounds that have sheets of metal hydroxides with anions in the interlayer. Divalent and trivalent metal cations are present in the sheet and, to some degree, the divalent cations can substitute for trivalent cations. The divalent and trivalent cations are randomly distributed in an edge-sharing octahedral sheet, forming hydroxylated M(OH)2 sheets similar to those of brucite, Mg(OH)2. The excess positive charge created by isomorphous substitution is balanced by the presence of anions in the interlayers. A general formula for a LDH is as follows: II mx+ [M1-x MIII [Xx/m ‚nH2O]xx (OH)2]

(1)

where MII is a divalent cation, MIII is a trivalent cation, and X is an anion (16). The general formula for an AFm is [Ca2(Al, Fe)(OH)6]+x-‚ mH2O and Friedel’s salt, one of AFm phases, can be written in the form of an LDH as [Ca2Al(OH)6]+[(Cl-)‚2H2O] (13, 17). Other AFm phases in hydrated cement use Ca2+ as the divalent cation, Al3+ and/or Fe3+ as the trivalent cation and OH-, Cl-, SO42-, and CO32- as interlayer anions (13). Due to the complexity of Portland cement constituents and cement hydration chemistry, the active agents for reductive dechlorination of PCE in Fe(II)-DS/S system have not been identified. Identifying the active agents could lead to the development of more effective methods of producing them without the need for extraneous reagents. This would benefit field applications of Fe(II)-DS/S by reducing costs. It could also extend the technology to other applications such as water treatment where addition of Portland cement is not practical. The principal goal of this research was to identify the active agents for PCE degradation as well as the conditions that enhance the formation of the active agents in Fe(II)-DS/S in laboratory batch tests. To achieve this goal, two objectives were pursued. First, conditions that lead to maximizing production of the active agents were identified by measuring the ability of synthesized solids under various chemical mixtures to degrade PCE. The abilities of these solids to act as reductants for PCE were compared to mixtures of Fe(II) and Portland cement, such as have been used in Fe(II)DS/S. Second, X-ray diffraction (XRD), scanning electron microscopy (SEM), and SEM with electron-dispersive spectrometry (SEM-EDS) were used to identify compounds in chemical mixtures that had high activities for degradation of PCE.

PCE and other chlorinated aliphatic compounds can undergo abiotic reductive dechlorination under anoxic conditions that usually involve inorganic Fe(II) or sulfide (1-6). Intensive research related to abiotic reductive dechlorination of chlorinated aliphatic compounds has included investigations on the reactivity of reductants such as Fe(II) in Portland cement slurries (5, 6), zerovalent iron (7-9), and iron and sulfide minerals (3, 4, 10-12). Fe(II)-based degradative solidification/stabilization (Fe(II)-DS/S) has been developed (5, 6) as a modification of conventional solidification/stabilization (S/S). In DS/S, organic pollutants, such as tetrachloroethylene (PCE), are

Experimental Section

* Corresponding author phone: (423)439-5249; fax: (423)439-5243; e-mail: [email protected]. † East Tennessee State University. ‡ Texas A&M University.

Materials. PCE (99.9+%, HPLC grade, Aldrich) was used as the target organic compound. Portland cement (ASTM C-150, type I) and FeCl2 (99+%, tetrahydrate, Aldrich) were used as DS/S reagents. Synthetic cement extract (SCX) solutions,

5824

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 16, 2007

10.1021/es070361f CCC: $37.00

 2007 American Chemical Society Published on Web 06/26/2007

Friedel’s salt, and tetracalcium aluminate hydrate (C4AH13) were made using the following chemicals (ACS or higher grade): AlCl3 (hexahydrate, 98+%, Sigma), H3BO3 (Matheson), BaCl2 (dehydrate, 100.3%, Fisher), BeSO4(tetrahydrate, 99+%, Fluka), CaCl2 (dihyrate, 99.5-105.0%, ACS grade, EM), FeCl3 (hexahydrate, 98+%, Sigma), CuCl2 (dehydrate, 99+%, Aldrich), MgCl2 (hexahydrate, 99+%, EM), MgSO4 (heptahydrate, 98+%, EM), MnSO4 (monohydrate, 98.6+%, Fisher), NiSO4 (hexahydrate, Aldrich), SrCl2 (hexahydrate, 99+%, Fluka), ZnCl2 (anhydrous, 98+%, EM), NaSiO3 (ACS grade, Fisher), NaAlO2 (anhydrous, EM), and Ca(OH)2 (Fisher). NaCl (100.8%, ACS grade, Mallinckrodt), and Na2SO4 (99.9%, ACS grade, Sigma) were used to adjust chloride and sulfate concentrations in synthetic cement extract (SCX). Deaerated deionized water was prepared by taking water purified by a Barnstead Nanopure system and purging it with nitrogen for at least 12 h in an anaerobic chamber (Coy Laboratory Product) that contained 5% hydrogen and 95% nitrogen. Preparation of 10% (w/v) Portland Cement Extract (PCX). Portland cement was dissolved by mixing it with strong acid (2.2 N HCl) on a shaking table for at least 24 h. After 24 h, the mixture of Portland cement and acid was transferred to several 250 mL plastic centrifuge bottles and centrifuged at 6000 rpm (6650 g) for 5 min (Beckman, model J-6M centrifuge, JS-7.5 rotor). Supernatant was filtered with filter paper (2 µm, VWR) to remove the visible suspension and solids at the bottom of bottle were discarded. The filtered solution was called a Portland cement extract (PCX). To remove oxygen, PCX was purged with nitrogen for at least 24 h in the anaerobic chamber. Solid Synthesis. Syntheses of all solids were performed in an anaerobic chamber. The potential active agents were prepared by adding Fe(II) (39.2 mM) and Ca(OH)2 (1.25 M) to PCX. This resulted in pH near 12, which was the optimum pH for PCE degradation by Fe(II)-DS/S. After adding reagents, the solutions were mixed for several hours in the anaerobic chamber. The mixtures were centrifuged at 6000 rpm (6650 g) for 5 min and the light blue colored solids at the top layer were taken for further experiments (Fe(II)PCX solids hereafter). Solids in cement slurry systems were prepared by mixing Portland cement (solid/solution mass ratio ) 0.1) with Fe(II) (39.2 mM). The pH of the mixture was adjusted to 12 using 5 N HCl. After shaking for 5 days, the mixtures were centrifuged at 6000 rpm (6650g) for 5 min and the solids at the top layer were taken for instrumental analyses. Friedel’s salt (18) and C4AH13 (18) were synthesized in the lab as examples of cement hydration products (CHPs). Each CHP was mixed with either a solution of 39.2 mM of Fe(II) or with a mixture of 39.2 mM Fe(II) and 47.8 mM Fe(III). The Fe(III) concentration was the same as the sum of Fe(III) and Al concentrations in 10% PCX. SCX was prepared to examine the effect of an individual cement element on the formation of the active agents. The concentrations of the elements in SCX were the same as those in 10% (w/v) PCX (See Supporting Information Table S-1). Three different sets of SCX were prepared. One solution contained all elements of PCX. The second set of solutions each excluded a single element of PCX. The third solution excluded all major PCX elements (Ca, Mg, Al, and SO4). PCE degradation tests were conducted after a 3 day mixing period. The pH was adjusted to 12.0 with 1.25 M Ca(OH)2 for SCX containing full PCX elements and with 5 N NaOH for SCX that excludes individual elements. To examine the effects of major elements in the cement extract, Fe(III), Mg, Al, SO4, and/or SiO3 were added during the synthesis of another set of solids at concentrations that were the same as those in 10% PCX. Fe(II) (39.2 mM) and Cl (2.2 M) were always added to the mixtures, because they were presumed to be the critical elements for the formation of the active agent in the preliminary experiment. PCE degradation tests were performed with the solids

produced after mixing for 3 days in the anaerobic chamber. These solids were synthesized in two ways. One way simply mixed the necessary elements and the other applied a method used to synthesize green rust (GR) (19) in which air oxidation is used to produce Fe(III). Experimental Procedures. Activity tests of all synthesized solids were conducted as one-point kinetic experiments in an anaerobic chamber. The reaction pH was near 12. The reactor system, the preparation of PCE stock solution and the PCE extraction procedures followed the method of Hwang and Batchelor (5). Analytical Methods. PCE was analyzed by gas chromatography (Hewlett-Packard 6890 GC with a combination of DB-VRX column (60 m × 0.25 mm i.d. × 1.4 µm film thickness, J & W Scientific) and an electron capture detector (ECD). 1,2-dibromopropane (1,2-DBP, 97%, Aldrich) was used as an internal standard. Solid phases were characterized by XRD and SEM-EDS. Rigaku automated diffractometer using Cu KR radiation (λ ) 1.5406 Å) was used to obtain the powder X-ray patterns (Department of Geology and Texas Transportation Institute at Texas A&M University). The XRD analysis was conducted by scanning between 5 and 60° 2θ with scan speed of 3° 2θ/minute. A JEOL 6400 scanning microscope (Microscopy and Image center at Texas A&M University) was used to analyze morphology and qualitative chemical compositions of the sample. Inductively coupled plasma (Perkin-Elmer Optima DV 3300 dual-view optical emission ICP spectrometer) analysis was conducted to examine chemical composition of Fe(II)-PCX by a technician in the Department of Oceanography at Texas A&M University. Chloride ion was analyzed by ionic chromatography (Dionex 500) equipped with AS9-HC column (250 mm × 4 mm i.d., Dionex) and conductivity detector. The Ferrozine method (20) was used to analyze Fe(II) and total iron using a UV-vis spectrophotometer (HewlettPackard G1103A). The ethylene glycol monoethylether (EGME) procedure (21) was used for the surface area measurement of Fe(II)-PCX. Evaluation of Kinetic Data. PCE degradation by Fe(II)DS/S previously showed pseudo-first-order kinetics and apparent kinetic constants were determined that considered partitioning of PCE to gas, liquid, and solid phases (5). Apparent first-order rate constants were calculated from concentrations measured in one-point solid activity tests using kapp ) (ln C0 - ln C)/t. Apparent first-order rate constants, kapp, were normalized with concentrations of Fe(II) (kFe(II)), Fe(III) (kFe(III)), or solids (ksolid) to specifically correlate the relationships between reductant concentrations and rate constants. The 95% confidence intervals for the rate constants of one-point solid activity tests were calculated using an equation of the confidence interval on the slope for simple linear regression. A total of five data points (two values for C0 and three for Ct,) were used to calculate 95% confidence intervals for k (k ( σ/xsxxt[n - 2, 0.025], n ) 5.) The error in the solids-normalized rate constants (ksolid) was calculated using a Taylor Series.

Results and Discussion Fe(II)-PCX Solids Activity. Values of kFe(II), kFe(III), and ksolid for PCX solids at pH 11.8 are shown in Table 1. The values of kFe(II) measured in these experiments were 1 order of magnitude higher than those measured in cement slurries with Fe(II), at pH 12.1 (5). These higher rate constants might be the result of active reductants being formed more readily from soluble components of PCX instead of the solid forms found in cement. Effect of CHPs on Solid Activities. Table 1 shows the results of PCE reduction experiments using mixtures of Fe(II) and Fe(III) with PCX and CHPs (C4AH13 and Friedel’s VOL. 41, NO. 16, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5825

TABLE 1. Pseudo First-Order Rate Constants for PCE Reduction by Fe(II)-PCX and Cement Hydration Products (tetracalcium Aluminate and Friedel’s Salt) by Themselves, with Fe(II) and with Both Fe(II) and Fe(III)a expt

solid

1 2 3 4 5 6 7 8 9

Fe(II)-PCX C4AHx

Friedel’s

2 h Fe(II) 2 h Fe(II)+Fe(III) 24 h Fe(II) 24 h Fe(II)+Fe(III) 2 h Fe(II) 2 h Fe(II)+Fe(III) 24 h Fe(II) 24 h Fe(II)+Fe(III)

pH

ksolidb L/(g × d)

kFe(II)c (mM × d)-1

kFe(III)d (mM × d)-1

11.8 11.8 11.8 11.8 11.8 11.8 11.8 11.8 11.8

6.1 × 10-03 ((19%) 1.8 × 10-04 ((352%) N/A 5.5 × 10-04 ((53%) 1.5 × 10-04 ((463%) 1.8 × 10-04 ((203%) 2.1 × 10-04 ((62%) 4.5 × 10-04 ((119%) 3.1 × 10-04 ((135%)

1.1 × 10-02 ((4.9%) 1.1 × 10-04 ((352%) N/A 3.3 × 10-04 ((53%) 1.2 × 10-04 ((187%) 8.9 × 10-05 ((203%) 1.3 × 10-04 ((62%) 2.3 × 10-04 ((119%) 2.0 × 10-04 ((135%)

1.1 × 10-01 ((4.9%) 2.1 × 10-03 ((352%) N/A 6.2 × 10-02 ((53%) 1.1 × 10-04 ((187%) 7.3 × 10-04 ((203%) 2.5 × 10-05 ((62%) 7.2 × 10-04 ((119%) 6.6 × 10-05 ((135%)

a N/A, not able to estimate kinetic constants because no reduction of PCE was observed; Initial PCE concentration was 0.242 mM; Sampling times for individual experiment: 4 days for expts 2-5; and 3 days for expts 6-9; Uncertainties represent 95% confidence limits expressed in % relative to estimate k. b ksolid) kapp/solid conc. c kFe(II) ) kapp/Fe(II) conc. d kFe(III) ) kapp/Fe(III) conc.; kapp ) pseudo-first-order rate constant, unit is day-1.

TABLE 2. Pseudo First-Order Rate Constants of PCE Reduction by Solids Composed of Full, Minor, and Major Elements of Cement Extracta expt 10 11 12 13 14 15 16 17 18 19 20 1

solid

pH

ksolid L/(g × d)

kFe(II) (mM × d)-1

kFe(III) (mM × d)-1

FSCXFeb MSCXFec Fe(II)(III)Cld Fe(II)(III)ClMgd Fe(II)(III)ClAld Fe(II)(III)ClMgAld Fe(II)(III)ClSO4d Fe(II)(III)ClSO4Ald Fe(II)(III)ClSiO3d Fe(II)(III)Cle_GR7 Fe(II)(III)Clf_GR12 Fe(II)-PCX

12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0 11.8

9.7 × 10-04 ((26%) 1.2 × 10-02 ((15%) 3.4 × 10-03 ((40%) 4.6 × 10-03 ((26%) 2.6 × 10-03 ((7.9%) 1.9 × 10-03 ((31%) 5.2 × 10-03 ((67%) 2.6 × 10-03 ((26%) 4.4 × 10-03 ((33%) 1.1 × 10-02 ((32%) 4.4 × 10-03 ((50%) 6.1 × 10-03 ((19%)

6.1 × 10-03 ((26%) 3.8 × 10-03 ((15%) 8.2 × 10-04 ((40%) 1.5 × 10-03 ((25%) 1.2 × 10-03 ((7.7%) 1.3 × 10-03 ((31%) 1.7 × 10-03 ((67%) 1.1 × 10-03 ((25%) 1.3 × 10-03 ((33%) 2.2 × 10-03 ((32%) 1.5 × 10-03 ((50%) 1.1 × 10-02 ((4.9%)

1.2 × 10-02 ((26%) 4.1 × 10-02 ((15%) 1.6 × 10-02 ((40%) 6.1 × 10-03 ((25%) 1.1 × 10-02 ((7.7%) 1.2 × 10-02 ((31%) 2.0 × 10-02 ((67%) 1.7 × 10-02 ((25%) 1.6 × 10-02 ((33%) 7.7 × 10-02 ((32%) 1.4 × 10-02 ((50%) 1.1 × 10-01 ((4.9%)

TCEf mM 0.02

a Initial PCE concentration was 0.242 mM; A sampling time for individual experiment was 4.5 days for expt 10; 6.9 days for expt 11; 8.5 days for expt 12; 7 days for expts 13-15, 18; 5.6 days for expts 16 and 17; 6.9 days for expt 19; and 7 days for expt 20; Uncertainties represent 95% confidence limits expressed in % relative to estimate k. b All PCX elements are added, full element synthetic cement extract (FSCX). c All PCX elements are added other than Ca, Mg, and Al, minor element synthetic cement extract (MSCX). d ,e,f Each solid was named after its chemical compositions. d Solids were synthesized by simple mixing method. e ,fSolids were synthesized by the green rust synthesis method (19); Initial Fe(III) concentration of solid was 8.7 mM. e Initial NaOH concentration is 70 mM NaOH. f Initial NaOH concentration is 110 mM NaOH.

salt). When both Fe(II) and Fe(III) were added to C4AH13 suspensions and mixed for 2 h in the anaerobic chamber, PCE was not degraded (expt 3). The values of kFe(II) in CHP suspensions (expts 2-9) were 1-3 orders of magnitude lower than those in experiments with Fe(II)-PCX solids (expt 1). Addition of 48 mM Fe(III), which is the same concentration as sum of Fe(III) and Al in PCX, did not improve the degradative activities of the solids (compare expts 3 to 2, 5 to 4, 7 to 6, and 9 to 8). Furthermore, the time allowed for formation of the solids (2 or 24 h) did not affect the solid activities (compare expts 4 to 2, 5 to 3, 8 to 6, and 9 to 7). Effect of SCX on Solid Activities. Table 2 presents results of experiments with SCXs. Full synthetic cement extract with Fe(II) addition (FSCXFe) (expt 10) solids showed 50% less activity than Fe(II)-PCX solids (expt 1) as shown by the values of kFe(II). Although the composition of FSCXFe simulated PCX as close as possible, the same activities were not obtained. When major constituents of PCX (Ca, Al, and Mg) were not added to the synthetic extract (MSCXFe), the solids produced (expt 11) had similar activity as those produced with FSCXFe (expt 10). However, significant amounts of TCE were detected, with concentrations as high as 10% of the initial PCE concentration. This indicates that the PCE degradation pathway of MSCXFe solids was different from that of solids produced by Fe(II)-DS/S, where no TCE was detected (5). All of the elements removed from SCX had similar effects on PCE degradation rates, because their Fe(II)normalized rate constants were the same order of magnitude (see Supporting Information Table S-2). Adding Mg, Al, SO4, or SiO3 (expts 13-25) to Fe(II)(III)Cl slightly improved solid activities (80%) compared to Fe(II)5826

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 16, 2007

(III)Cl (expt 12). Fe(II)(III)Cl solids had values of kFe(II) that were 1 order of magnitude smaller than the values observed with FSCXFe solids (expt 10) and 2 orders of magnitude smaller than those observed with Fe(II)-PCX solids (expt 1). The GR synthesis method (19) did not increase activities of solids compared to those formed with simple mixing. Fe(II), Fe(III), and Cl might be the most important elements in forming active agents due to the observation that there was not a significant enhancement to activity of solids when other elements were introduced to cement extract. Instrumental Analyses of Solids. Cement Systems (10% PCS and 10% PCX). Figure 1(top) shows the XRD pattern for solids from a 10% Portland cement slurry with and without Fe(II). The major solid phases identified by XRD in solids formed in experiments with Fe(II) were found to be calcium chloroaluminate hydrates (Friedel’s salt, JCPD 42-558), calcium aluminate hydrates (JCPD 33-255) and calcium aluminum silicate hydrates (JCPD 18-274). Peaks circled in Figure 1 show the distinctive differences between solids containing Fe(II) (Fe(II)-PCS) and solids not containing Fe(II) (PCS). The intensity of the peak at a d-spacing of 2.87 Å is higher in Fe(II)-PCS. Ettringite (JCPD 41-1451) was identified in the solids from the 10% cement slurry that did not contain Fe(II) (PCS). Most Ettringite peaks disappeared in Fe(II)-PCS. In general, Ettringite can be formed within 30 min when cement is mixed with water (13, 14). Introducing Fe(II) into the cement slurry system might facilitate the formation of Friedel’s salts rather than Ettringite, so that the formation of Ettringite was either inhibited or decelerated. Adding Fe(II) did not make any new solids that could be identified by XRD.

FIGURE 1. X-ray patterns of solids formed in cement systems (PCS and Fe(II)-PCS solids at the top and PCX and Fe(II)-PCX solids at the bottom). Note: Unit of d-spacing is Å. Figure 1(bottom) shows the XRD patterns of solids formed in 10% PCX with and without addition of Fe(II). Portlandite (JCPD 87-673) and Friedel’s salts were the major solids identified in 10% PCX solids to which Fe(II) was added. Peaks of Friedel’s salts were more clearly observed when Fe(II) was added to 10% PCX, which was the same thing that occurred with Fe(II) was added to slurries with 10% cement. The pH in these slurries was increased by addition of 1.25 M Ca(OH)2, which resulted in considerable amounts of Ca(OH)2 in the solids being analyzed. The highest intensity peak in 10% Fe(II)-PCX solids was observed at d-spacing of 2.35 Å, which is associated with Friedel’s salt, the fourth most intense peak of the β form of Friedel’s salt (JCPD 35-105). The highest intensity peak in 10% PCX solids was found at a d-spacing of about 2.65 Å, which is associated with Portlandite. The peak with d-spacing of 2.77 Å might be associated with calcium aluminum silicate hydrates, but the first and second

most intense peaks (3.1 and 5.8 Å) of calcium aluminum silicate hydrates were not detected. Intensity and d-spacing values of PCX solids were not exactly matched with references. This might be the result of different chemical composition and atomic arrangement of these solids compared to the referenced solids. Portlandite and calcium aluminum silicate hydrate probably were the dominant solid phases formed in 10% PCX solids. Peaks for Friedel’s salt were not observed in 10% PCX solids. XRD patterns of 10% Fe(II)-PCX solids were similar to those of 10% Fe(II)-PCS solids (Figure 1). Peaks from either calcium aluminum silicate hydrate or calcium aluminate hydrate in 10% Fe(II)-PCX solids were not observed as strongly as they were for 10% Fe(II)-PCS solids. However, both cement slurries and cement extracts showed the presence of Friedel’s salts when Fe(II) was also added. This supports the hypothesis that it is a compound that could be active in reducing PCE. VOL. 41, NO. 16, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5827

FIGURE 2. SEM image and EDS of solids formed in cement systems (Fe(II)-PCS at the top and two different shapes of Fe(II)-PCX solids in the middle and at the bottom). Figure 2 at top shows SEM images and EDS spectra of solids from experiments with Fe(II)-PCS and PCX with Fe(II). EDS spectra were taken from a hexagonal particle. Needle-like crystals of Ettringite and hexagonal plates of Friedel’s salt were found in solids formed in cement slurries without Fe(II) (See Supporting Information Figure S-2). Disappearance of needle shaped particles of Ettringite in samples from systems with Fe(II) addition was observed in SEM images, which supports the results from XRD. The morphology of Portlandite and hydrated calcium aluminate (C4AHx) is hexagonal as well. Moreover, Friedel’s salt and C4AH13, which are LDHs, are difficult to distinguish due to 5828

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 16, 2007

their similar shapes and positions of their maximum intensity peaks. However, they can be distinguished by the particle size. The sizes of particles of Friedel’s salt are typically from 2 to 3 µm while particles of C4AHx are usually less than 1 µm. Particles of Portlandite are much larger than those of Friedel’s salt, with a maximum length of about 100 µm (14, 22). The SEM images showed that the particle sizes of solids from cement slurries with and without Fe(II) were about 5 and 11 µm, respectively. When Fe(II) was added to the cement slurry, the particle sizes were reduced. Particle size could affect degradation kinetics because smaller particles would have larger specific surface areas. The importance of particle

FIGURE 3. XRD patterns (top) and SEM image and EDS (bottom) of FSCXFe solids. Note: Unit of d-spacing is Å. surface area is supported by observations that kinetics of degradation of target compounds in Fe(II)-DS/S were described by a saturation model, which is consistent with a reaction mechanism controlled by reactive surface sites (6). Reducing particle size would, thus, generate more active sites on the surface of Fe(II) solids and increase rates of reductive dechlorination of PCE. Figure 2 in the middle shows SEM images and EDS spectra of Fe(II)-PCX solids. Although SEM images of 10% PCX of Figure 2 did not show the perfect hexagonal shapes observed in Fe(II)-PCS (top), stacked thin plate particles were observed in Fe(II)-PCX solids. Particle sizes in 10% Fe(II)-PCX were 1-3 µm, which were substantially smaller than those in 10% Fe(II)-PCS. EGME specific surface area of Fe(II)-PCX was 406 m2/g which is much larger than BET specific surface area of hydrated cement pastes, 10-246 m2/g (13).Although some ethylene glycol monoethylether (EGME) remained as a liquid in the pores of the particle rather than being adsorbed onto a surface, resulting in overestimating the surface area (21), one advantage of the EGME method is that it keeps the sample from being oxidized because it can be conducted within an anaerobic chamber. The change of particle sizes could be the result of different particle nucleation rates in the two systems. Nuclei of hexagonal particles could be formed more readily in Fe(II)-PCX, which would tend to cause the resultant suspensions to have a greater number of

smaller particles. This phenomenon could explain the faster PCE degradation rate in Fe(II)-PCX solids than one in Fe(II)-PCS solids. The EDS spectra in Figure 2 in the middle shows that the presence of Mg could result in a lower degree of substitution of Mg for Ca and that the presence of Si could result in substitution of Si for Al. Another explanation for the presence of Si is that it exists as an interlayer anion in the LDH that appear as thin hexagonal plates in the SEM. Based on ICP analysis, Fe(II)-PCX is composed of 44% Ca and 31% Cl on a molar basis. High contents of Cl could come from addition of 2.2 N HCl that was used to digest cement. Fe, Al, and Mg were present at lower levels (see Supporting Information Table S-3). Interestingly, no Fe was detected in solids prepared from 10% Fe(II)-PCX (Figure 2 at the bottom), where hexagonal shape particles were not observed. These pictures probably showed type II calcium silicate hydrate (C-S-H) (13). Figure 2 at the bottom supports the observation that Fe is associated with hexagonal particles. XRD, SEM, and EDS analyses support the supposition that AFm phases such as Friedel’s salt, calcium aluminate hydrates, and calcium aluminum silicate hydrates, are responsible for reduction of PCE. This explanation is not supported by results of PCE degradation tests with C4AH13 and Friedel’s salts (Table 1) that show lower degradation rates for these compounds than were observed for Fe(II)PCS and Fe(II)-PCX solids. This discrepancy could be due VOL. 41, NO. 16, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5829

FIGURE 4. SEM image and EDS of Fe(II)(III)Cl_GR12 (top) and Fe(II)(III)Cl_GRN (bottom). to differences in the AFm phases formed in mixtures of cement and pure phases formed in defined media. Since AFm phases readily changed their constituents depending on their environment, AFm phases formed in cement mixtures can be expected to be heterogeneous and complex in composition because of the heterogeneity and complexity of the environment of cement pore water (23). Such complexity would not be expected in pure phases of C4AH13 and Friedel’s salts used in kinetic experiments. AFm phases with diverse chemical composition due to isomorphous substitution of LDHs would be formed in cement system and one or more substituted elements could play an important role in increasing the ability of the solids to dechlorinate. Furthermore, AFm phases formed in cement systems would also have a smaller particle size, have increased specific surface area, and potentially more active sites for dechlorination. Conditions during formation of AFm phases in cement mixtures could also lead to differences in reactivity. When cement is mixed with water, reactants are gradually dissolved from cement components and dissolved ions react to form cement hydration products. Changing concentrations of dissolved species over time could result in different solid phases being formed in cement mixtures compared to more defined systems used to produce pure AFm phases. Rates of nucleation of solid phases could also be different in cement mixtures compared to more defined media and this could result in different particle size distributions and reactivities. Therefore, the characteristics of active solids could depend on characteristics of systems in which they were formed (complex cement mixtures or defined solutions). Characterizing the effects of system composition on particle formation mechanisms and the reactivity of resulting solids phases 5830

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 16, 2007

would be interesting future research areas that would help to identify active agents and conditions to promote the formation of active agents in Fe(II)-DS/S. Non-Cement Systems. XRD analysis demonstrated that Portlandite was the major solid formed in FSCXFe solids and Friedel’s salt was not formed (Figure 3 at the top). The majority of Portlandite identified in XRD analysis came from the Ca(OH)2 added for pH adjustment. The rest of the peaks match reasonably well with calcium silicate hydrate (JCPDS 23125), except for the peak at 8.03 Å. EDS spectra (Figure 3 at the bottom) showed that all FSCXFe solids were composed of Ca, Al, Si, Cl, S, and Fe. Fe(II) was not incorporated into calcium silicate hydrate based on SEM and EDS spectra taken in Fe(II)-PCX solids (Figure 2 at the bottom). Therefore, calcium silicate hydrates identified in Fe(II)/FSCX system are probably not active reducing agents. In addition, a mixture of Fe(II) and Portlandite showed approximately 2 orders of magnitude lower kFe(II) than mixture of Fe(II) and cement (5). Portlandite is probably not an active reducing agent, either. The peak at 8.03 Å matches that expected from GR_Cl as does the peak at 3.95 Å. However, the third most intense peak of GR_Cl (2.7 Å) is not found. It could be possible that the intensity of the peak at 2.7 Å was too low to be distinguished from noise due to very high intensities of peaks from Portlandite. Green rusts (GRs) are usually observed as corrosion products and as natural minerals in anaerobic soils and sediments at neutral pH (24). According to Glasser (25), GRs formed during steel corrosion are stable up to about pH 13. Therefore, the formation of GR at pH 12.0 in Fe(II)/FSCX is possible. Fe(II) was selectively incorporated with particles with hexagonal shapes as seen in Figure 2 at the bottom, so Fe found in EDS spectra of FSCXFe solids could be due to

GRs. The sizes of hexagonal particles circled in Figure 3 were less than 0.5 µm, which is much smaller than is typical for Portlandite. Particle sizes of green rust have been reported to vary in size from 0.02 to about 1 µm (19) and to be larger than 2 µm (26). Since the ability of GR to act as a reductant is dramatically improved when metals, such as Ag, Au, and Cu, were added (11), active reductants formed in Fe(II)/FSCX system could be modified GRs that include trace metals from cement extract elements and/or substituted di- and trivalent metals, such as Ca and Al. Solids formed in the mixture of FeCl2, FeCl3, and NaCl (Fe(II)(III)Cl) and minor synthetic cement extract (MSCX) were ferrous hydroxide (see Supporting information Figure S-3) regardless of chemical composition, synthesis method, and pH. A large amount (2.2 M) of NaCl was added to keep the same concentration of chloride as in the cement extract, and this extra NaCl might remain in the solids being analyzed. During drying, white solids were observed along with colored solids, which could have been halite formed from excess NaCl. Green rust chloride (GR_Cl) was not formed in the neutral pH system. This might be due to low concentration (0.4 mM) of ferric iron. GR_Cl has a ratio of Fe(II) to Fe(III) of 3, but the ratio in this system was around 10. In the absence of major elements of cement, only iron could participate in the formation of solids. Figures 4 shows the SEM images and EDS spectra of Fe(II)(III)Cl solids synthesized by the coprecipitation method (19) at different pH. EDS spectra were taken from a single hexagonal particle. The particle sizes were much smaller when they had been formed at pH 12, (0.2 and 0.5 µm) than at neutral pH (4-5 µm). Chemical compositions in solution and pH during solid formation were the important factors in determining the ability of different solids to act as reductants. This research could not clearly demonstrate whether LDH such as Friedel’s salt, calcium aluminum hydrate and/or green rust chloride are the active agents in Fe(II)-DS/S system. Evidence supporting their role as active reductants included increased levels of Fe on particles with the characteristic hexagonal shape of LDH. Evidence opposed to their role is the lower rates of reduction observed for Friedel’s salt (expts 6-9) and calcium aluminum hydrate (expts 2-5) compared to Fe(II)-PCX (expt 1) and Fe(II)-PCS solids (5). A unique characteristic of LDHs is the diversity of their chemical compositions by isomorphous substitutions on the sheet of metal hydroxide and/or anion exchanges in the interlayer (27). Since cements are very complex substances composed of various kinds of elements, it is possible that the LDHs identified in this research could be modified from their pure forms and these modifications could enhance their activities for PCE degradation. It might be also possible that active agents are not a single phase but a complex mixture of different LDHs. To more precisely identify the compounds responsible for reduction in Fe(II)-DS/S it would be necessary to develop techniques to separate solid phases from each other so that their individual activities as reducing agents could be determined.

Acknowledgments This material is based in part upon work supported by the Texas Advanced Research Program under grant no. 0005120066-2001.

Supporting Information Available Table of composition of SCX and pseudo first-order rate constants of PCE reduction by ferrous iron containing SCX solids which excluded only one element of PCX. Figure of SEM and EDS of PCS solids and XRD patterns of MSCX and mixture of Fe(II), Fe(III), and Cl with different solid synthesis

method as well as pH. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Macalady, D. L. Abiotic reduction reactions of anthropogenic organic chemicals in anaerobic systems a critical review. J. Contam. Hydrol. 1986, 1, 1-28. (2) Tsukano, Y. Transformations of selected pesticides in flooded rice-field soilsA review. J. Contam. Hydrol. 1986, 1, 4763. (3) Kriegman-King, M. R.; Reinhard, M. Transformation of carbon tetrachloride by pyrite in aqueous solution. Environ. Sci. Technol. 1994, 28, 692-700. (4) Kriegman-King, M. R.; Reinhard, M. Transformation of carbon tetrachloride in the presence of sulfide, biotite, and vermiculite. Environ. Sci. Technol. 1992, 26, 2198-2206. (5) Hwang, I.; Batchelor, B. Reductive dechlorination of tetrachloroethylene by Fe(II) in cement slurries. Environ. Sci. Technol. 2000, 34, 5017-5022. (6) Hwang, I.; Batchelor, B. Reductive dechlorination of chlorinated methanes in cement slurries containing Fe(II). Chemosphere 2002, 48, 1019-1027. (7) Arnold, W. A.; Roberts, A. L. Pathways and kinetics of chlorinated ethylene and chlorinated acetylene reaction with Fe(0) particles. Environ. Sci. Technol. 2000, 34, 1794-1805. (8) Johnson, T. J.; Scherer, M. M.; Tratnyek, P. G. Kinetics of halogenated organic compound degradation by iron metal. Environ. Sci. Technol. 1996, 30, 2634-2640. (9) Roberts, A. L.; Totten, L. A.; Arnold, W. A.; Burris, D. R.; Campbell, T. J. Reductive elimination of chlorinated ethylenes by zerovalent metals. Environ. Sci. Technol. 1996, 30, 26542659. (10) Lee, W.; Batchelor, B. Abiotic reductive dechlorination of chlorinated ethylenes by iron-bearing soil minerals. 2. Green rust. Environ. Sci. Technol. 2002, 36, 5348-5354. (11) O’Loughin, E. J.; Kemner, K. M.; Burris, D. R. Effects of AgI, AgIII, and CuII on the reductive dechlorination of carbon tetrachloride by green rust. Envoron. Sci. Technol. 2003, 37, 29052912. (12) Butler, E. C.; Hayes, K. F. Kinetics of the transformation of trichloroethylene and tetrachloroethylene by iron sulfide. Environ. Sci. Technol. 2000, 34, 422-429. (13) Talyor, H. F. W. Cement Chemistry, 2nd ed; Thomas Telford: London, 1997 (14) St John, D. A.; Poole, A. W.; Sims, I. Concrete Petrography; Arnold: London, 1998 (15) Renaudin, G.; Kubel, F.; Rivera, J. P.; Francois, M. Structural phase transition and high temperature phase structure of Friedel’s salt, 3CaO‚Al2O3‚CaCl2‚10H2O. Cem. Concr. Res. 1999, 29, 1937-1942. (16) You, Y.; Vance, G. F.; Zhao, H. Surfactant-enhanced adsorption of organic compounds by layered double hydroxides. Colloids Surf., A 2002, 205, 161-172. (17) Birnin-Yauri, U. A.; Glasser, F. P. Friedel’s salt, Ca2Al(OH)6(Cl, OH)‚2H2O: Its solid solutions and their role in chloride binding. Cem. Concr. Res. 1998, 28, 1713-1723. (18) Chudek, J. A.; Hunter, G.; Jone, M. R.; Scrimgeour, S. N.; Hewlett, P. C.; Kudryavtsev, A. B. Aluminium-27 solid state NMR spectroscopic studies of chloride binding in Portland cement and blends. J. Mater. Sci. 2000, 35, 4275-4288. (19) Gehin, A.; Ruby, C.; Abdelmoula, M.; Benali, O.; Ghanbaja, J.; Refait, P.; Genin, J. M. R. Synthesis of Fe(II-III) hydroxysulfate green rust by coprecipitation. Solid State Sci. 2002, 4, 6166. (20) Gibbs, C. R. Characterization and application of FerroZine iron reagent as a ferrous iron indicator. Anal. Chem. 1976, 48, 1197120. (21) Klute, A. Eds. Methods of Soil Analysis, Part 1. Physical and Mineralogical Methods, 2nd ed; Agronomy: Madison, WI, 1986. (22) Stutzman, P. E. Scanning Electron Microscopy in Concrete Petrography, In Materials Science of Concrete Special Volume: Calcium Hydroxide in Concrete; Skalny, J. P., Gebauer, J., Odler, I., Eds.; The American Ceramic Society: Columbus, OH, 2001; pp 59-72. (23) Glasser, F. P.; Kindness, A.; Strounach, S. A. Stability and solubility relationship in AFm phases part I. chloride, sulfate and hydroxide. Cem. Concr. Res. 1999, 29, 861-866. (24) Refait, Ph.; Genin, J. M. R. The oxidation of ferrous hydroxide in chloride-containing aqueous media and pourbaix diagrams of green rust one. Corros. Sci. 1993, 34, 797-819. VOL. 41, NO. 16, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5831

(25) Sagoe-Crentsil, K. K.; Glasser, F. P. Constitution of green rust and its significance to the corrosion of steel in Portland cement. Corrosion 1993, 49, 457-463. (26) Peulon, S.; Legrand, L.; Antony, H.; Chausse, A. Electrochemical deposition of thin film of green rusts 1 and 2 on inert gold substrate. Electrochem. Commun. 2003, 5, 208-213. (27) Roy, A. D.; Forano, C.; Malki, K. E.; Bess, J. P. Anionic Clays: Trends in Pillaring Chemistry. In Expanded Clays and Other

5832

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 16, 2007

Microporous Solids; Occelli, M. L., Robson, H. E., Eds; Van Nostrand Reinhold: New York, 1992; pp 108-169.

Received for review February 12, 2007. Revised manuscript received May 20, 2007. Accepted May 22, 2007. ES070361F