Combined Removal of Chlorinated Ethenes and Heavy Metals by

Combined Removal of Chlorinated Ethenes and Heavy Metals by Zerovalent Iron in Batch ... reduction by ZVI (100 g L-1) due to catalytic hydrodechlorina...
0 downloads 0 Views 161KB Size
Environ. Sci. Technol. 2005, 39, 8460-8465

Chromate undergoes a reductive reaction in ZVI systems where chromium serves as the oxidizing agent (eq 2) (5).

Combined Removal of Chlorinated Ethenes and Heavy Metals by Zerovalent Iron in Batch and Continuous Flow Column Systems

Fe0 + CrO42- + 4 H2O f Fe3+ + Cr3+ + 8 OH- (2)

J A N D R I E S , †,‡ L E E N B A S T I A E N S , * ,† D I R K S P R I N G A E L , †,§ SPIROS N. AGATHOS,‡ AND LUDO DIELS† Department of Environmental Technology, Flemish Institute for Technological Research (Vito), Boeretang 200, 2400 Mol, Belgium, and Unit of Bioengineering, Catholic University of Louvain, Croix du Sud 2, boıˆte 19, 1348 Louvain-la-Neuve, Belgium

The combined removal of chlorinated ethenes and heavy metals from a simulated groundwater matrix by zerovalent iron (ZVI) was investigated. In batch, Ni (5-100 mg L-1) enhanced trichloroethene (TCE, 10 mg L-1) reduction by ZVI (100 g L-1) due to catalytic hydrodechlorination by bimetallic Fe0/Ni0. Cr(VI) or Zn (5-100 mg L-1) lowered TCE degradation rates by a factor of 2 to 13. Cr(VI) (100 mg L-1) in combination with Zn or Ni (50-100 mg L-1) inhibited TCE degradation. Addition of 20% H2(g) in the headspace, or of Zn (50-100 mg L-1), enhanced TCE removal in the presence of Ni and Cr(VI). Sorption of Zn to ZVI alleviated the Cr(VI) induced inhibition of bimetallic Fe0/ Ni0 apparently due to release of protons necessary for TCE hydrodechlorination. In continuous ZVI columns treating tetrachloroethene (PCE, 1-2 mg L-1) and TCE (10 mg L-1), and a mixture of the metals Cr(VI), Zn(II), and Ni(II) (5 mg L-1), the PCE removal efficiency decreased from 100% to 90% in columns operated without heavy metals. The PCE degradation efficiency remained above 99% in columns receiving heavy metals as long as Ni was present. The findings of this study indicate the feasibility and limitations of the combined treatment of mixtures of organic and inorganic pollutants by ZVI.

Introduction In-situ zerovalent iron (ZVI) permeable barriers are successfully implemented for the treatment of chlorinated solvents or heavy metal contaminated groundwaters. The ZVI technology is very effective for the removal of contaminants with a variety of chemical characteristics due to the occurrence of multiple interactions such as reduction, (co)precipitation, and sorption (e.g., 1-3). Chlorinated aliphatics are reductively dechlorinated by a direct electron-transfer mechanism from the corroding ZVI (Fe0) surface (eq 1) (4).

Fe0 + RCl + H+ f Fe2+ + RH + Cl-

(1)

* Corresponding author tel: +32 14.33.51.79; fax: +32 14.58.05.23; e-mail: [email protected]. † Flemish Institute for Technological Research (Vito). ‡ Catholic University of Louvain. § Present address: Laboratory for Soil and Water Management, Catholic University of Leuven, Kasteelpark Arenberg 20, 3001 Heverlee, Belgium. 8460

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005

The removal of the reduced chromium species Cr(III) occurs through precipitation of the sparingly soluble Cr(OH)3 or precipitation of mixed iron(III)-chromium(III) (oxy)hydroxide solids (6, 7). Nickel participates in a cementation reaction, which is a spontaneous electrochemical process, also referred to as reductive deposition (8). It involves the reduction of the more electropositive species such as Ni2+ (E° ) -0.25 V) by the more electronegative ZVI (E° ) -0.44 V) (9). The cementation reaction results in a bimetallic Fe0/Ni0 system. Zinc (E° ) -0.76 V), which is more electronegative than ZVI, does not react with ZVI to form reactive zerovalent Zn0, but takes part in surface complexation reactions with the iron oxides present on the ZVI surface (10-12). The basis for these sorption reactions is the corrosion of ZVI (eq 3), forming amorphous ferrous hydroxide and mixed valent iron salts, known as green rusts (10, 12).

Fe0 + 2 H+ f Fe2+ + H2

(3)

The subsequent oxidation of these corrosion products can lead to the formation of iron oxides such as magnetite, maghemite, and lepidocrocite, forming a coating layer on the surface of ZVI. In the presence of water, the iron oxides are generally covered with surface hydroxyl groups (tSs OH) (13, 14). Bivalent cationic metals (Me2+) in solution can participate in several surface complexation reactions with these functional groups (eqs 4-6).

tSsOH + Me2+ T tSsOMe+ + H+

(4)

2 tSsOH + Me2+ T (tSsO)2Me + 2 H+

(5)

tSsOH + Me2+ + H2O T tSsOMeOH + 2 H+ (6) Zinc can form unidentate (eq 4) or bidentate surface complexes (eq 5). Metal-ligand complexes such as hydrolyzed zinc can also be adsorbed (eq 6) (13, 14). There is limited expertise available about the treatability of mixtures of organic and inorganic groundwater pollutants by ZVI. The effect of the presence of the single metals Cr(VI) or Ni on the reduction of chlorinated aliphatics by ZVI has been reported in previous studies. Schlicker et al. (15) showed that Cr(VI) severely hindered chloroethene degradation. In contrast, several authors reported on the enhancement of chlorinated ethene degradation in bimetallic Fe0/Ni0 systems (e.g., 16-18). In zerovalent bimetals, ZVI is used in conjunction with another more electropositive metal such as Pd, Pt, Cu, or Ni, forming electronic contact (18-22). The second metal may (i) catalyze hydrodechlorination or hydrogenation reactions whereby dechlorination proceeds via hydrogen reduction rather than via electron transfer, (ii) enhance the corrosion rate of ZVI, or (iii) prevent the formation of an oxide film on ZVI (17, 23). Chlorinated ethenes and methanes were hydrodechlorinated in minutes to hours by bimetals, while ZVI alone showed dechlorination rates that were more than 2 to 3 orders of magnitude lower. The order of reactivity of bimetals was observed to be Fe0/Pd0 > Fe0/Ni0 > Fe0/Cu0 (22). In addition to the enhanced rate of reduction by bimetals, hydrodechlorination also prevents the accumulation of less chlorinated reaction products (8, 23, 24). 10.1021/es050251d CCC: $30.25

 2005 American Chemical Society Published on Web 09/27/2005

TABLE 1. Nominal Input Concentrations of Chlorinated Ethenes and Heavy Metals, and Average Performance of the Continuous Flow Column Systems with (+M) and without (-M) Input of Heavy Metals during Four Consecutive Operational Periods average removal nominal input concentration (mg L-1) period (days) 1 2 3 4

0-376 377-526 527-586 587-769

PCE

TCE

Znb

1 1 2 2

10 10 10 10

5 5 5 5

PCE (%)

TCE (%)

productsa (%)

Nib

Crb

-M

+M

-M

+M

-M

+M

5 0 0 5

5 5 5 5

99 97 94 90

100 100 98 99

100 100 100 100

100 100 100 100

3.2 0.9 0.2 0.06

0.6 0 0 0

a Average molar concentration ratio of the sum of the chlorinated ethene reaction products 1,2-cis-DCE, 1,1-DCE, and VC in the effluent, to the sum of PCE and TCE in the influent of the column systems. b Only in the continuous flow column systems with (+M) input of heavy metals.

Jeong and Hayes (25) described the effect of various transition metals, including Zn and Ni, on the reductive dechlorination rate of hexachloroethane by the iron sulfide mackinawite, but, to our knowledge, the effect of Zn on the reductive dechlorination of chlorinated aliphatics by ZVI has not been described before. Moreover, little information is available on the impact of mixtures of these metals on the reactivity of ZVI systems. We set up a study to (i) assess the effect of individual metals and heavy metal mixtures containing Cr(VI), Zn, and Ni on the degradation of trichloroethene (TCE) in batch ZVI systems, and (ii) investigate the combined removal of the chlorinated ethenes tetrachloroethene (PCE) and TCE, and a mixture of the heavy metals from a simulated groundwater matrix by ZVI in continuous flow columns.

Experimental Section Media. All experiments were performed using a dilute simulated groundwater at neutral pH. The simulated groundwater consisted of NaHCO3, KHCO3, CaCl2‚2H2O, and MgCl2.6H2O at a concentration of 0.5 mM each in MilliQ water. The ZVI was cast iron supplied by Gotthart Maier Metallpulver (Germany). The iron filings had a size ranging from 0.25 to 2 mm and a specific surface area of 0.745 ( 0.007 m2 g-1 (as determined by N2 single point BET analysis). PCE (99.9% purity) and TCE (>99.5%) were supplied by Merck (Germany). The metals Zn (as ZnCl2; >98%) and Ni (as NiCl2‚ 6H2O; >98%) were purchased from Merck (Germany), and Cr(VI) (as K2CrO4; >99%) was supplied by JT Baker Chemicals (The Netherlands). Setup of the Batch Experiments. Batch tests were set up to investigate effects of heavy metals on the reduction kinetics of TCE by ZVI. In a first series of tests, 26-mL amber glass serum bottles were supplied with 100 g L-1 of ZVI, and were filled with simulated groundwater, leaving no headspace. TCE was added from a saturated aqueous solution to obtain an initial concentration of 10 mg L-1. Each series included a control set without ZVI, and a reference set with ZVI without heavy metals. If appropriate, the metals Zn, Ni, and Cr(VI) were added as (i) individual metals in a concentration range of 5-100 mg L-1, (ii) combinations of two metals (at 50 and 100 mg L-1), and (iii) a combination of the three metals (at 50 and 100 mg L-1). In a second series of batch experiments, the setup consisted of 118-mL glass serum bottles supplied with 100 g L-1 of ZVI. The flasks were filled with simulated groundwater, leaving a 38-mL headspace. The bottles were crimp-sealed with thick Viton stoppers. The headspace was subsequently exchanged to an atmosphere of 100% N2 under pressure (0.2 atm) after 8 cycles of evacuation and pressurization. A first set of bottles was first exposed during one week to simulated groundwater containing Ni, Cr(VI), or a mixture of the three metals (at 100 mg L-1 each) without TCE. After one week, an aqueous solution containing TCE (10 mg L-1) and either Ni (100 mg L-1), Cr, or a mixture of metals (Ni+Zn or Cr+Zn),

was added to the bottles. A second set of bottles, containing ZVI (100 g L-1), TCE (10 mg L-1), and combinations of either Ni and Cr or Ni, Cr, and Zn (100 mg L-1 each), were incubated in the presence or absence of hydrogen gas, i.e., 20% H2(g), in the headspace. Each series of bottles also included a control without ZVI, and a reference with ZVI without heavy metals. Refer to the Supporting Information for a complete description of the batch experimental setup. Continuous Flow Column setup. Two parallel continuous flow column systems were set up, treating a continuous input of simulated groundwater contaminated by a mixture of PCE (1 mg L-1) and TCE (10 mg L-1). The first column system was the reference system without input of heavy metals, whereas the second column system was continuously supplied with the heavy metals Cr(VI), Zn, and Ni, all at 5 mg L-1. For a description of the laboratory setup, refer to the text and Figure S1 in the Supporting Information. The continuous flow column test was operated for over 860 days during which a number of changes were applied. Table 1 shows the main characteristics of the first 4 operational periods of the test. After about 1 year of operation, Ni was no longer added to the influent solution of the second column system (with heavy metals). The input concentration of PCE was increased from 1 to 2 mg L-1 in the third operational period. In the fourth period, Ni (5 mg L-1) was again added to the influent of the second column system. The ZVI column system fed with heavy metals was terminated after 770 days of operation, at the end of the fourth operational period. One of the two remaining ZVI columns originally operated without input of heavy metals was then fed with simulated groundwater containing Zn (5 mg L-1). Reactivity and Analysis of Reacted ZVI. To examine the reactivity of reacted ZVI, appropriate columns were dismantled in an anaerobic glovebox, and the reacted ZVI filings recovered from the lower and upper halves of the columns were collected separately. The filings were freeze-dried, and stored in closed serum bottles under a N2 headspace. The ZVI filings recovered from the heavy metal loaded column were analyzed for the total metal content after microwave destruction with HCl/HNO3/HF/H3BO3. Batch tests were set up to investigate the sorption and degradation of TCE (at 10 mg L-1) and PCE (at 5 mg L-1) by reacted ZVI from the two different heights in each column, in comparison with fresh ZVI. The procedure followed for the setup of the batch reactivity tests was as described above (26-mL vials set up without headspace). Zerovalent Iron Extraction Protocol. To determine the sorption isotherms for TCE and PCE in fresh and reacted ZVI systems, we determined the evolution of both their aqueous and sorbed concentrations (26, 27). The protocol, modified from Allen-King et al. (28), is described in the Supporting Information section. Analyses. pH, ORP, concentration of chlorinated ethenes and metals, and the morphology and elemental composition VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8461

FIGURE 1. Removal of TCE (10 mg L-1) in batch zerovalent iron systems in the presence of increasing concentrations (5-100 mg L-1) of Zn, Ni, or Cr(VI) (control, no ZVI present; reference, ZVI present, without addition of heavy metals; error bars represent one standard deviation for triplicate systems; note that the concentration profiles in the presence of Ni at 20, 50, and 100 mg L-1 are superimposed). of the fresh and reacted ZVI filings were measured as described in the Supporting Information. Modeling. The Supporting Information provides a detailed description of (i) the first-order model applied to describe the reductive dechlorination of a parent compound by ZVI (29), and (ii) the Freundlich sorption isotherm expression describing the equilibrium distribution of organic compounds between the aqueous and solid phase in ZVI systems (14).

Results and Discussion Effect of Individual Metals on TCE Degradation in Batch ZVI Systems. In the absence of metals, TCE was degraded according to first-order kinetics, with a first-order rate constant of 0.021 ( 0.001 h-1 (n ) 18, R2 ) 0.97). The main chlorinated ethene reaction product was 1,2-cis-dichloroethene (cisDCE), peaking at a concentration of 25-30 mol % of the initial TCE concentration (data not shown). Figure 1 shows the influence of increasing concentrations, from 5 to 100 mg L-1, of the individual heavy metals on the dechlorination kinetics of TCE. TCE was completely removed in the sets supplied with Ni alone in the first 24 h of the test (Figure 1), while chlorinated reaction products were never observed (data not shown). In comparison to the reference without heavy metal addition, TCE reduction was 2-4 times slower in the presence of increasing concentrations of Zn, and 3-13 times slower in the presence of Cr(VI) at concentrations higher than 5 mg L-1. The strongly enhanced TCE dechlorination activity in the presence of Ni, and the reduced TCE degradation rates in the presence of Cr(VI) and Zn, can be explained as a result of the different heavy metal removal mechanisms in ZVI systems. Schlicker et al. (15) described the delayed onset of trichloroethene (TCE) reduction in continuous ZVI columns when Cr(VI) was present as co-contaminant. The reductive transformation of the strong oxidant Cr(VI) competed with the reductive dechlorination of TCE. Chromate reduction contributes to ZVI passivation by formation of Fe(III)-Cr(III) oxides at anodic sites (30, 31). In contrast, TCE half-lives were 50-180 times lower in ZVI systems in the presence of Ni due to the formation of a highly reactive Fe0/Ni0 system (17). Fennelly and Roberts (32) reported significantly faster 1,1,1-trichloroethane reduction rates with nickel-plated ZVI than with ZVI alone. The adverse effect of the presence of Zn2+ on the reduction of chlorinated ethenes by ZVI has not been reported previously. Iron oxides have a strong affinity for zinc and hydrolyzed zinc (11) and these sorption reactions may have a weak corrosion inhibition effect and prevent 8462

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005

FIGURE 2. Removal of TCE (10 mg L-1) in batch zerovalent iron systems in the presence of Zn, Ni, and Cr(VI) supplied in different ratios (legend, initial [Zn]/[Ni]/[Cr] ratio in mg L-1; control, no ZVI present; reference, ZVI present, without addition of heavy metals; error bars represent one standard deviation for triplicate systems). access to the reactive sites at the ZVI surface (12, 13). The presence of individual metals had distinct effects on the bulk pH in the experimental sets. In the reference sets without heavy metals, the pH increased from near-neutral to approximately 9.5, due to proton consumption by reductive dechlorination (eq 1) and anaerobic corrosion (eq 3). The pH value in the flasks supplied with Cr(VI) increased to almost 11, more than 2.5 units higher than the initial pH, corresponding to the reaction mechanism for chromate removal (eq 2). In contrast, a quick pH drop of about 1- 2 units was observed in the flasks with Zn, potentially resulting from acid production by the sorption reaction (eqs 4-6). Effect of Binary and Ternary Metal Mixtures on TCE Degradation in Batch ZVI Systems. In experiments where we investigated the effect of combinations of two metals, the following patterns were observed (Figure 2). The degradation rate of TCE was 3-7 times lower in the presence of Zn combined with Cr. TCE was completely removed within 24 h in the presence of 100 mg L-1 Ni in combination with 50 mg L-1 Cr or 100 mg L-1 Zn. In contrast, TCE reduction was inhibited in the presence of 100 mg L-1 Cr in any combination with Ni. The TCE degradation pattern in the presence of Ni and Cr, both at 50 mg L-1, consisted of an inhibition phase during the first 24 h, followed by an accelerated decay phase. TCE was completely removed within the first 24 h in the presence of a mixture of the three heavy metals present at 50 mg L-1 each. The TCE concentration profile in the presence of a mixture of the three heavy metals present at 100 mg L-1 each was very peculiar, first showing a slow decrease, followed (after one week) by a rapid and complete dissappearance. This phenomenon was recorded in 3 independent experiments. Interestingly, the TCE removal pattern observed in the presence of the three metals (at 100 mg L-1 each), was not seen in the presence of Ni and Cr(VI) (each at 100 mg L-1) in the absence of Zn. The difference between the TCE removal pattern in the presence of the three metals at 50 and 100 mg Ll-1 each is probably caused by differences in the rate of metal removal at the different concentrations, although more detailed investigation is required to elucidate metal concentration effects on the removal rate of metals and TCE. Concerning the pH, we observed a significant pH increase of more than 2 units in sets supplied with 100 mg L-1 Cr(VI) in combination with Zn (at all concentrations) or Ni (at 50 mg L-1). The pH remained neutral in flasks receiving Ni and Cr(VI), both at 100 mg L-1. In all other sets, a pH decrease of 1-2 units was recorded (data not shown). With respect to the removal of heavy metals, we observed the following qualitative trends (Figure S2, Supporting Information). First, the removal of Zn (100 mg L-1) was only hampered significantly in the presence of a mixture of Ni or Cr(VI). Second, the rate of Ni (100 mg L-1) removal decreased

FIGURE 3. Removal of TCE (10 mg L-1) in batch zerovalent iron systems in the presence or absence of heavy metals, after 7 days incubation without TCE in the presence of Ni, Cr, or Zn+Ni+Cr (a) (legend, metals present initially > metals added after 7 days; 0, no addition), and TCE removal in the presence (+H2) or absence (-H2) of 20% H2(g) in the headspace of batch ZVI systems in the presence or absence of heavy metals (b) (note that some of the concentration profiles are superimposed). to some extent in the presence of Cr(VI), but decreased significantly in the presence of Zn alone or in combination with Cr(VI). Third, the removal of Cr(VI) (100 mg L-1) was slightly impacted in the presence of Ni, and was slightly accelerated in the presence of Zn alone or in combination with Ni. Additional batch experiments were set up to explore the mechanisms that may explain the observed metal interaction effects. ZVI was first exposed to a solution containing either Ni, Cr(VI), or a mixture of the three metals (at 100 mg L-1 each). After one week, TCE alone or in combination with one or two metals was added to the flasks. In the systems initially exposed to Ni, TCE was completely removed within 24 h after addition of Cr(VI) alone or in combination with Zn (Figure 3a). The performance of the Fe0/Ni0 system was thus not affected significantly by the subsequent addition of Cr(VI) alone. In the systems initially exposed to Cr(VI), TCE degradation was strongly inhibited only in the batch receiving no subsequent addition. Addition of Ni, alone or in combination with Zn, enhanced the removal of TCE, indicating that Ni alleviated the inhibition caused by Cr precipitates. In a subsequent experiment, we investigated the effect of the supply of extra hydrogen, in the form of hydrogen gas (20% H2(g) in the headspace), on the heavy metal effects. The lag time for TCE degradation decreased from more than 1 and 2 days, in the absence of H2, to less than 24 h in the presence of H2 in Zn/Ni/Cr (100 mg L-1 each) and Ni/Cr (100 mg L-1 each) systems, respectively (Figure 3b). The presence of hydrogen did not influence TCE removal in the reference batches without addition of heavy metals. The presence of Zn enhanced the TCE degradation in Ni/Cr systems by decreasing the lag time by 1 day. The results reported in Figures 2 and 3 refer to different experimental conditions (see the Experimental Section), which may explain the differences in lag times observed between the removal profiles in the 2 figures. However, since the impact of these

experimental conditions is unknown, the reported removal profiles were not compared quantitatively. Mechanisms. The most noticeable phenomenon in the various batch experiments was the peculiar effect of Zn. We observed that Zn did not negatively impact the enhanced TCE reduction in the presence of Ni (Figure 2). More so, the presence of Zn seemed responsible for the fast TCE reduction observed in batches supplied with the three metals together (Figures 2 and 3). These observations contrast with the negative impact of Zn alone on TCE degradation by ZVI, caused by sorption reaction reactions at the ZVI surface (see above and Figure 1). These findings indicate that the effect of metal mixtures on TCE degradation is significantly more complex than the effect of the individual metals. The reductive deposition of Ni by ZVI results in the bimetallic Fe0/Ni0 system that degrades chlorinated ethenes by a mechanism different from that of ZVI alone. The results from the present study suggest that the impact of the addition of Zn and/or Cr(VI) differs between ZVI and bimetallic Fe0/Ni0, due to these differences in reaction mechanism. The coating of zerovalent Ni acts as a hydrodechlorination catalyst (24). As ZVI corrodes, protons derived from water are reduced to atomic hydrogen that chemisorbs to the catalytic Ni0 surface, and then serves as the reducing agent (22, 23, 33). The fact that different reaction mechanisms are at play is clearly illustrated by the distinct impact of molecular hydrogen (H2(g)). The addition of H2(g) only enhanced TCE degradation in the bimetallic Fe0/Ni0 system, and not in the ZVI system, where TCE degradation results from direct electron transfer (Figure 3b). Molecular hydrogen dissociates at the Ni0 surface into atomic hydrogen, which is the reducing agent (22). Besides passivating ZVI, the reductive precipitation of 100 mg L-1 Cr(VI) by ZVI also consumes 15.4 mM H+ (eq 2). The resulting pH increase may have impacted the Ni0 catalyzed hydrodechlorination, which depends on the input of protons. On the other hand, depending on the type of surface complex formed, sorption of 100 mg L-1 Zn species by the iron oxide coating on the ZVI releases 1.53-3.06 mM H+ (eqs 4-6). This release of protons may partially alleviate the negative effect of Cr(VI) by delivering protons necessary for the two competing proton consuming reactions, i.e., (i) the hydrodechlorination of TCE by the Fe0/Ni0 bimetallic system, and (ii) the reduction of chromate by ZVI. To confirm this hypothesis on a more quantitative basis, more detailed experiments would be necessary. Effect of Mixed Metals on Chloroethene Degradation in Continuous Flow Column Systems. The average pH in the effluent of the ZVI columns, i.e., 6.8 ( 0.7, was similar to the average pH of the influent, i.e., 6.9 ( 0.5. The pH increase resulting from ZVI corrosion and reductive dechlorination typically recorded in continuous flow column experiments (e.g., 34), was not observed, probably due to the low contact time in the ZVI columns (about 1 day), the continuous elution of corrosion products such as hydroxyl ions, and the presence of the bicarbonate buffer (1 mM) in the simulated groundwater. The average ORP, i.e., -60 ( 97 mV vs SHE, was significantly lower in the ZVI columns than in the influent, i.e., 387 ( 37 mV vs SHE. Table 1 summarizes the performance of both column systems, and the removal of PCE and TCE in both systems is compared in Figure S3 (Supporting Information). During a first experimental period, the influent heavy metal concentration in the system fed with heavy metals was 5 mg L-1 for each of the metals. The average heavy metal removal ranged from 97 to 100% (data not shown). PCE and TCE were almost completely removed in both systems (Table 1). The main chlorinated reaction product observed in the effluent of the ZVI columns was 1,2-cis-DCE, but traces of vinyl chloride (VC) and 1,1dichloroethene (1,1-DCE) were also recorded. In the first period, the concentration of chlorinated reaction products VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8463

TABLE 2. Heavy Metal Content, Batch First-Order Reaction Kinetics, and Freundlich Sorption Isotherm Parameters for Fresh (Reference) and Reacted ZVI Recovered from Column Systems Operated With (+M) and Without (-M) Input of Heavy Metalsa metal content (mg kg-1) ZVI

Zn

Ni

Cr

PCE sorption isotherm

reaction k/krefb

reference 60 ( 32 590 ( 94 1300 ( 419 1 ( 0.2 bottom -M n.d. n.d. n.d. 0.77 ( 0.04 top -M n.d. n.d. n.d. 0.60 ( 0.02 bottom +M 1300 ( 53 1700 ( 128 2980 ( 85 2.1 ( 0.3 top +M 80 ( 46 800 ( 118 3200 ( 811 1.12 ( 0.04

KF

N

A/Arefc

44 ( 1 19 ( 1 24 ( 1 31 ( 1 17 ( 1

0.45 ( 0.02 0.46 ( 0.05 0.48 ( 0.02 0.23 ( 0.02 0.54 ( 0.03

1 0.43 0.55 0.60 0.43

reaction k/krefb

TCE sorption isotherm

KF

N

1 ( 0.06 10 ( 1 0.66 ( 0.02 1.28 ( 0.06 4 ( 1 0.44 ( 0.04 0.9 ( 0.2 5 ( 1 0.61 ( 0.07 3.5 ( 0.9 4 ( 1 0.60 ( 0.05 1.7 ( 0.1 5 ( 1 0.53 ( 0.04

A/Arefc 1 0.24 0.41 0.34 0.33

aPCE and TCE supplied at 5 and 10 mg L-1, respectively; k, surface area normalized first-order rate constant (h-1); K , Freundlich sorption capacity F (µmol kg-1 µM-N); N, degree of isotherm linearity; n.d., not determined. b kref: first-order rate constant for fresh ZVI (h-1) normalized to a surface 2 -1 c area of 1 m mL [PCE, 0.16 ( 0.03; TCE, 0.27 ( 0.02]. Comparison of the area below the isotherm for a certain set (A) and for the reference case (Aref), calculated by integration of the fitted Freundlich expression within the experimental range (0-5 µM for PCE and 0-20 µM for TCE).

was about 5 times higher in the effluent of the column system operated without input of heavy metals (Table 1). The findings of the batch experiments help explain the results obtained in the column test. The better performance of the column system receiving heavy metals was most likely caused by the enhanced dechlorination activity of the bimetallic Fe0/Ni0 reductant resulting from cementation of Ni on the ZVI surface. The reactivity of ZVI was significantly enhanced in the presence of 5 mg L-1 Ni in batch, while the presence of Cr(VI) and Zn at the same concentration did not significantly affect the degradation kinetics of TCE (Figure 1). In the second and third operational periods of the column test, we investigated the longevity of the Fe0/Ni0 system in the presence of heavy metals without Ni (Table 1). The average PCE removal in the ZVI systems fed with heavy metals gradually decreased, from 100 to 98%, but was still significantly higher than in the system operated without input of metals, indicating that the Fe0/Ni0 reductant was only partially being passivated. TCE removal was not affected. The PCE removal efficiency rapidly increased when Ni was added again during the fourth operational period (Table 1), confirming the role of Ni as the performance enhancing factor. In the system without input of heavy metals, the average PCE removal efficiency gradually decreased from almost 100% to less than 90% at the end of the test (Table 1). The average efficiency of TCE degradation remained higher than 99% throughout. In the fifth operational period of the test, we investigated the effect of Zn (at 5 mg L-1) on the efficiency of PCE and TCE removal by the column system that was previously operated without input of heavy metals. No significant effect was observed (data not shown). Degradation of TCE and PCE by Reacted ZVI. Table 2 shows the total metal content of reacted ZVI recovered from the column system that was fed with heavy metals, in comparison with fresh ZVI. The bottom half of the iron column contained 22, 3, and 2.3 times more Zn, Ni, and Cr, respectively, than fresh ZVI. The top half was enriched in Cr, Ni, and Zn by a factor of 2.5, 1.4, and 1.3, respectively (Table 2). No attempt was made to make mass balances on the column system in view of the high standard deviation of the reported metal content values (Table 2), potentially resulting from the difficulty to obtain a homogeneous sample of reacted ZVI particles containing a surface coating enriched in heavy metals. The reactivity of the reacted ZVI filings collected from the top and the bottom halves of one column from both systems was tested in a batch assay, and compared to the reactivity of fresh ZVI (Table 2). The surface area normalized firstorder rate constant estimated for fresh ZVI was 0.16 ( 0.03 h-1 for PCE (n ) 9; R2 ) 0.84) and 0.27 ( 0.02 h-1 for TCE (n ) 12; R2 ) 0.97). These kinetic parameters were based on total concentrations (i.e., aqueous + sorbed) to exclude the effect of sorption (26). The degradation of TCE by reacted 8464

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005

ZVI recovered from the bottom and the top of the column fed with heavy metals was, respectively, 3.5 and 1.7 times faster than by fresh ZVI. In contrast, the rate of TCE degradation by reacted ZVI collected from the column operated without input of metals was not significantly different from the rate recorded with fresh ZVI (Table 2). PCE degradation was more than 2 times faster in batch systems with reacted ZVI recovered from the bottom of the column fed with heavy metals. Reacted ZVI collected from the top of the metal-fed column reduced PCE as fast as fresh ZVI, while reacted ZVI sampled from the column operated without input of metals was 1.3-1.7 times less reactive toward PCE (Table 2). The enhanced reactivity of ZVI recovered from the bottom of the metal-fed column is most likely a result of the significant enrichment in Ni as zerovalent Ni0 in an Fe0/Ni0 system. The batch experiments with fresh ZVI (see Figure 1) indicate that reductively deposited Ni on the ZVI surface promotes TCE degradation at concentrations as low as 50 mg kg-1. In reacted ZVI systems collected from the top part of the metal-fed column, the potential positive effect of the enrichment in Ni, as zerovalent Ni0, may have been compensated by the negative impact of the more pronounced enrichment in Cr (see above) (15). SEM micrographs of reacted ZVI filings show the presence of an oxide film on the surface (Figure S4). In reacted ZVI systems recovered from the column operated without input of metals, the presence of this coating did not interfere with the reduction of TCE but probably hindered the degradation of the less reactive PCE. With regard to PCE degradation, the decreased reactivity of reacted vs fresh ZVI agrees with the gradual decline in the removal efficiency observed in the column system operated without input of metals (Table 1). For TCE, the reactivities of fresh and reacted ZVI were essentially the same, corresponding to the high average removal efficiency observed throughout the continuous flow column test. Sorption of TCE and PCE to Reacted ZVI. Table 2 also shows the estimated Freundlich sorption isotherm coefficients for PCE and TCE. The data were best described by a nonlinear Freundlich sorption isotherm (R2 > 0.96). The sorption data should be considered as quasi-equilibrium sorption isotherms since chemical degradation reactions are also taking place (26). The area below the sorption isotherm, computed by integration of the fitted Freundlich expression, was used as a measure for the magnitude of sorption. TCE sorption was 2.4-4.2 times lower in reacted ZVI systems than in fresh ZVI systems. PCE sorption was 1.7-2.3 times lower in reacted ZVI systems (Table 2). The drop in sorption magnitude in reacted ZVI systems might be explained by the presence of the oxide film on the surface of the ZVI filings (Figure S4), which makes that part of the sorption sites, i.e., the embedded graphite in cast iron (35), are no longer available for sorption. Our results indicate the potential of reactive iron barriers to treat mixtures of organic and inorganic groundwater

contaminants. Individual heavy metals and mixtures of metals had distinctive effects on the reductive dechlorination of chlorinated aliphatics by ZVI. The results of the present study confirm the performance enhancing effect in ZVI systems of low input concentrations of nickel present in a heavy metal mixture. The presence of Ni not only affected the reaction kinetics, but also the mechanism by which chlorinated compounds are degraded.

(16) (17) (18) (19)

Acknowledgments We thank Miranda Maesen, Queenie Simons, Vale`re Corthouts, Gust Nuyts, Stefaan Kuypers, and Raymond Kemps for assistance during this work. This research was funded in part by a Vito Ph.D. grant, and by EC project QLK3CT-2000-00163.

(20)

Supporting Information Available

(22)

Information on the materials and methods, the laboratory continuous flow column setup, and additional results of the batch and column experiments. This material is available free of charge via the Internet at http://pubs.acs.org.

(23)

(21)

(24)

Literature Cited (1) Gillham, R. W.; O’Hannesin, S. F. Enhanced degradation of halogenated aliphatics by zerovalent iron. Ground Water 1994, 32, 958-967. (2) Puls, R. W.; Paul C. J.; Powell, R. M. The application of in situ permeable reactive (zerovalent iron) barrier technology for the remediation of chromate contaminated groundwater: a field test. Appl. Geochem. 1999, 14, 989-1000. (3) Shokes, T. E.; Mo¨ller, G. Removal of dissolved heavy metals from acid rock drainage using iron metal. Environ. Sci. Technol. 1999, 33, 282-287. (4) Matheson, L. J.; Tratnyek, P. G. Reductive dehalogenation of chlorinated methanes by iron metal. Environ. Sci. Technol. 1994, 28, 2045-2053. (5) Powell, R. M.; Puls, R. W.; Hightower, S. K.; Sabatini, D. A. Coupled iron corrosion and chromate reduction: mechanisms for subsurface remediation. Environ. Sci. Technol. 1995, 29, 1913-1922. (6) Blowes, D. W.; Ptacek, C. J.; Jambor, J. L. In-situ remediation of Cr(VI)-contaminated groundwater using permeable reactive walls: laboratory studies. Environ. Sci. Technol. 1997, 31, 33483357. (7) Pratt, A. R.; Blowes, D. W.; Ptacek, C. J. Products of chromate reduction on proposed subsurface remediation material. Environ. Sci. Technol. 1997, 31, 2492-2498. (8) Zhang, W. X.; Wang, C. B.; Lien, H. L. Treatment of organic contaminants with nanoscale bimetallic particles. Catal. Today 1998, 40, 387-395. (9) Khudenko, B. M.; Gould, J. P. Specifics of cementation processes for metals removal. Water Sci. Technol. 1991, 24, 235-246. (10) Odziemkowski, M. S.; Schumacher, T. T.; Gillham, R. W.; Reardon, E. J. Mechanism of oxide film formation on iron in simulating groundwater solutions: Raman spectroscopic studies. Corros. Sci. 1998, 40, 371-389. (11) Lehman, M.; Zouboulis A. I.; Matis K. A. Removal of metal ions from dilute aqueous solutions: a comparative study of inorganic sorbent materials. Chemosphere 1999, 39, 881-892. (12) Roh, Y.; Lee, S. Y.; Elless, M. P. Characterization of corrosion products in the permeable reactive barriers. Environ. Geol. 2000, 40, 184-194. (13) Cornell, R. M.; Schwertmann, U. The Iron Oxides: Structure, Properties, Reactions, Occurrence and Uses; VCH: New York, 1996. (14) Stumm, W.; Morgan, W. Aquatic Chemistry; Wiley: New York, 1996. (15) Schlicker, O.; Erbert, M.; Fruth, M.; Weidner, M.; Wu¨st, W.; Dahmke, A. Degradation of TCE with iron: the role of competing

(25) (26)

(27)

(28)

(29) (30)

(31)

(32) (33)

(34) (35)

chromate and nitrate reduction. Ground Water 2000, 38, 403409. Appleton, E. L. A nickel-iron wall against contaminated groundwater. Environ. Sci. Technol. 1996, 30, 536A-539A. Cheng, S. F.; Wu, S. C. The enhancement methods for the degradation of TCE by zerovalent metals. Chemosphere 2000, 41, 1263-1270. Cheng, S. F.; Wu, S. C. Feasibility of using metals to remediate water containing TCE. Chemosphere 2001, 43, 1023-1028. Lien, H. L.; Zhang, W. X. Transformation of chlorinated ethenes by nanoscale iron particles. J. Environ. Eng. 1999, 125, 10421047. Lien, H. L.; Zhang, W. X. Nanoscale iron particles for complete reduction of chlorinated ethenes. Colloids Surf., A 2001, 191, 97-105. Wan, C.; Chen Y. H.; Wei, R. Dechlorination of chloromethanes on iron and palladium-iron bimetallic surfaces in aqueous systems. Environ. Toxicol. Technol. 1999, 18, 1091-1096. Kim, Y. H.; Carraway, E. R. Reductive dechlorination of TCE by zerovalent bimetals. Environ. Technol. 2003, 24, 69-75. Schrick, B.; Blough, J. L.; Jones, A. D.; Mallouk, T. E. Hydrodechlorination of trichloroethylene to hydrocarbons using bimetallic nickel-iron nanoparticles. Chem. Mater. 2002, 14, 5140-5147. Wang, J.; Blowes, P.; Farrell, J. Understanding reduction of carbon tetrachloride at nickel surfaces. Environ. Sci. Technol. 2004, 38, 1576-1581. Jeong, H. Y.; Hayes, K. F. Impact of transition metals on reductive dechlorination rate of hexachloroethane by mackinawite. Environ. Sci. Technol. 2003, 37, 4650-4655. Burris, D. R.; Campbell, T. J.; Manoranjan, V. S. Sorption of trichloroethylene and tetrachloroethylene in a batch reactive metallic iron-water system. Environ. Sci. Technol. 1995, 29, 2850-2855. Dries, J.; Bastiaens, L.; Springael, D.; Agathos, S. N.; Diels, L. Competition for sorption and degradation reactions in batch zerovalent iron systems. Environ. Sci. Technol. 2004, 38, 28792884. Allen-King, R. M.; Halket, R. M.; Burris, D. R. Reductive transformation and sorption of cis- and trans-1,2-dichloroethene in a metallic iron-water system. Environ. Toxicol. Chem. 1997, 16, 424-429. Johnson, T. L.; Scherer, M. M.; Tratnyek, P. G. Kinetics of halogenated organic compound degradation by iron metal. Environ. Sci. Technol. 1996, 30, 2634-2640. Melitas, N.; Chuffe-Moscoso, O.; Farrell, J. Kinetics of soluble chromium removal from contaminated water by zerovalent iron media: corrosion inhibition and passive oxides effects. Environ. Sci. Technol. 2001, 35, 3948-3953. Melitas, N.; Farrell, J. Understanding chromate reaction kinetics with corroding iron media using Tafel analysis and electrochemical impedance spectroscopy. Environ. Sci. Technol. 2002, 36, 5476-5482. Fennelly, J. P.; Roberts, A. L. Reaction of 1,1,1-trichloroethane with zerovalent metals and bimetallic reductants. Environ. Sci. Technol. 1998, 32, 1980-1988. Odziemkowski, M. S.; Gui, L.; Gillham, R. W. Reduction of N-nitrosodimethylamine with granular iron and nickelenhanced iron. 2. Mechanistic studies. Environ. Sci. Technol. 2000, 34, 3495-3500. Orth, W. S.; Gillham, R. W. Dechlorination of trichloroethene in aqueous solution using Fe0. Environ. Sci. Technol. 1996, 30, 66-71. Burris, D. R.; Allen-King, R. M.; Manoranjan, V. S.; Campbell, T. J.; Loraine, G. A.; Deng, B. Chlorinated ethene reduction by cast iron: sorption and mass transfer. J. Environ. Eng. ASCE 1998, 124, 1012-1019.

Received for review February 7, 2005. Revised manuscript received August 2, 2005. Accepted August 3, 2005. ES050251D

VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8465