Reduction of Chlorinated Ethenes by Ag- and Cu-Amended Green Rust

: Chlorinated ethenes have been used extensively as solvents, degreasers, and dry-cleaning agents in a range of commercial and industrial applications. This has created a legacy of contaminated soils and groundwater, particularly with respect to perchloroethylene (PCE; a.k.a. tetrachloroethene— C 2 Cl 4 ), and trichloroethylene (TCE; a.k.a. trichloroethene—C 2 HCl 3 ), prompting the development of a wide array of treatment technologies for remediation of chlorinated ethene-contaminated environments. Green rusts are highly redox-active layered Fe(II)-Fe(III) hydroxides that have been shown to be facile reductants for a wide range of organic and inorganic pollutants. The reduction of chlorinated ethenes [vinyl chloride (VC); 1,1-dichloroethene(11DCE), cis -1,2-dichloroethene (c12DCE), trans -1,2-dichloroethene (t12DCE), TCE, and PCE] was examined in aqueous suspensions of green rust, alone as well as with the addition of Ag(I) (AgGR) or Cu(II) (CuGR). Green rust alone was ineffective as a reductant for the reductive dechlorination for all of the chlorinated ethenes. Near-complete removal of PCE was observed in the presence of AgGR, but all other chlorinated ethenes were essentially non-reactive. Partial removal of chlorinated ethenes was observed in the presence of CuGR, particularly 11DCE (34%), t12DCE (51%), and VC (66%). Signiﬁcant differences were observed in the product distributions of chlorinated ethene reduction by AgGR and CuGR. The effectiveness of Ag(I)- and Cu(II)-amended green rusts for removal of chlorinated ethenes may be improved under different conditions (e.g., pH and interlayer anion) and warrants further investigation.

No previous studies of chlorinated ethene reduction by green rusts (native or modified) have examined the entire chlorinated ethene series. Therefore, the objective of this study is to examine the potential of native and metal-amended [Ag(I) and Cu(II)] green rust for reductive dechlorination of PCE, TCE, 1,1-dichloroethene (11DCE: C 2 H 2 Cl 2 ), c12DCE, t12DCE, and chloroethene (a.k.a. vinyl chloride (VC)-C 2 H 3 Cl).

Reduction of Chlorinated Ethenes by GRSO4
Plots of concentration versus time for the chlorinated ethenes in aqueous suspens with GRSO4 are shown in Figure 2 and mass recoveries of the reactants and volatile p ucts (chlorinated ethenes, acetylene, ethene, and ethane) are given in Table 1. No rea products were detected, although water-soluble reactions products such as acetate acetaldehyde (for which our headspace analysis would have had low sensitivity) been observed or postulated as minor products of chlorinated ethene reduction [86 and may have been present. The absence of volatile reaction products and >95% reco of the chlorinated ethenes (except for 86% PCE recovery) at the end of the experim indicate that GRSO4 was not an effective reductant for any of the chlorinated ethenes u our experimental conditions. The incomplete mass recovery for PCE is not likely du chemical transformation of PCE, rather, minor amounts of PCE may have been lo sorption to the septum or to green rust [53].  [85]). The blue arrows designate β-elimination reactions (RCl = RCl + 2e − → R ≡ R + 2Cl − ), red indicates α-elimination (R = RCl 2 + 2e − → R ≡ R + 2Cl − ), green designates hydrogenation (R ≡ R + 2e − + 2H + → HR = RH or R = R + 2e − + 2H + → HR-RH), and black designates hydrogenolysis (R = RCl + 2e − + H + → R = RH + Cl − ). PCE = perchloroethene; TCE = trichloroethene; 11DCE = 1,1-dichloroethene; c12DCE = cis-1,2-dichloroethene; and trans-1,2-DCE = trans-1,2-dichloroethene.

Reduction of Chlorinated Ethenes by GR SO4
Plots of concentration versus time for the chlorinated ethenes in aqueous suspensions with GR SO4 are shown in Figure 2 and mass recoveries of the reactants and volatile products (chlorinated ethenes, acetylene, ethene, and ethane) are given in Table 1. No reaction products were detected, although water-soluble reactions products such as acetate and acetaldehyde (for which our headspace analysis would have had low sensitivity) have been observed or postulated as minor products of chlorinated ethene reduction [86,87], and may have been present. The absence of volatile reaction products and >95% recovery of the chlorinated ethenes (except for 86% PCE recovery) at the end of the experiment indicate that GR SO4 was not an effective reductant for any of the chlorinated ethenes under our experimental conditions. The incomplete mass recovery for PCE is not likely due to chemical transformation of PCE, rather, minor amounts of PCE may have been lost to sorption to the septum or to green rust [53].  Table S1. Table 1. The extents of chlorinated ethene removal, final product distributions, and carbon recoveries in aqueous suspensions of unamended green rust (GR), and green rust amended with 10 −4 M of either Ag(I) or Cu(II) (AgGR and CuGR, respectively).  99.8% a PCE = perchloroethene; TCE = trichloroethene; 11DCE = 1,1-dichloroethene; c12DCE = cis-1,2-dichloroethene; t12DCE = trans-1,2-dichloroethene; VC = vinyl chloride; AC = acetylene; EE = ethene; EA = ethane; and ND = no products detected. b Carbon molar balance of identified products and any remaining parent compound relative to the initial moles of parent compound.  Table S1. Our results are consistent with previous studies showing no reduction of polychlorinated ethenes by green rusts (GR SO4 , GR CO3 , GR Cl , and GR F ) [75,77,78]; however, others have reported significant, if not complete, reduction of polychlorinated ethenes and vinyl chloride by green rusts [53,72,74,76,88]. The variability in the reactivity of green rust with chlorinated ethenes has been attributed to multiple factors including the type of interlayer anion, surface area, pH, Fe(II)/Fe(III) ratio, and amount of Fe(II) sorbed on the green rust, and artifacts from GR preparation and handling [53,72,76,77,88]. Interlayer anion composition has been shown to affect the rate of reduction of nitrate, chromate [Cr(VI)], and U(VI) by green rust [43,48,52], and Liang et al. [76] observed faster rates of PCE and TCE reduction with GR Cl relative to GR SO4 , which they attributed to the greater surface area of the GR Cl used in their study. Han et al. [88] observed a decrease in the reduction rates of c12DCE and VC with increasing pH, which they ascribed to increasing amounts of Fe(II) on the green rust surface with increasing pH. Conversely, both Lee and Batchelor [53] and Maithreepala and Doong [72] reported a decrease in the rates of TCE and PCE reduction with increasing pH. Lee and Batchelor suggested their results could be explained by an increase in the overall thermodynamic driving force of the reaction at high pH due to the lower concentration hydrogen ions, a product of the dechlorination reaction, or to the increase in formation of hydroxide functional groups on green rust, which are presumed to be more reactive reductants. The lack of a consistent pattern in the reactivity of chlorinated ethenes with green rusts suggests that the observed differences in reactivity are perhaps due to multiple, often interrelated, factors.

Reduction of Chlorinated Ethenes by Ag-Amended GR SO4
As with the un-amended GR, there was essentially no dechlorination of VC and the DCEs by AgGR (Figure 2), as reaction products accounted for only 0.1-1.1% of the carbon mass balance (Table 1). TCE concentration decreased by~12% over 501 h in the presence of AgGR, but only 3% was recovered as reaction products (acetylene and ethene). However, within 189 h, PCE was nearly completely transformed to TCE (87.4%), t12DCE (7.3%), and acetylene (7.4%). The distribution of the reaction products suggests that PCE is primarily transformed to TCE via hydrogenolysis (Figure 1), with TCE being essentially a terminal product as reduction of TCE by AgGR was minimal ( Figure 2). However, some transformation of PCE to dichloroacetylene via β-elimination ( Figure 1) is suggested by the formation of t12DCE and acetylene. Dichloroacetylene was not observed as an intermediate; however, this is not surprising given that it is rapidly dechlorinated to chloroacetylene via hydrogenolysis and to t12DCE via hydrogenation [89]. Given that t12DCE is essentially unreactive with AgGR, it is unlikely that β-elimination of t12DCE contributed significantly to the observed production of acetylene during PCE reduction ( Figure 1). Rather, acetylene was likely the result of hydrogenolysis of chloroacetylene. As with dichloroacetylene, chloroacetylene was not observed as an intermediate; however, like dichloroacetylene, it is easily dechlorinated via hydrogenolysis, resulting in acetylene [89,90]. The formation of acetylene from the dechlorination of TCE by AgGR and the non-reactivity of lesser chlorinated ethenes suggests transformation of TCE to chloroacetylene via β-elimination, with subsequent reduction to acetylene (Figure 1). Reductive dechlorination of PCE and TCE by transition metal species often results in the formation of non-chlorinated C 3 -C 6 alkanes and alkenes [85,86,[90][91][92][93] attributed to radical coupling reactions; however, no such products were detected resulting from PCE reductive dechlorination by AgGR.

Reduction of Chlorinated Ethenes by Cu-Amended GR SO4
With the exception of PCE, chlorinated ethene reduction by CuGR was enhanced relative to AgGR or unamended GR (Figure 2). Over 622 h, 65.6% of VC was reduced to ethene (58.2%) and ethane (7.4%), consistent with dechlorination of VC via hydrogenolysis followed by hydrogenation of ethene to ethane (Figure 1). Approximately 50% of t12DCE was transformed to ethene (42.6%) and ethane (4.5%). These products could result from t12DCE dechlorination via β-elimination resulting in acetylene, and hydrogenation of acetylene to ethene or via sequential hydrogenolysis reactions forming VC then ethene; however, in either case, the initial transformation product (acetylene or VC) never accumu-lated at detectable concentrations. CuGR was less effective at dechlorination of c12DCE, with >80% remaining after 625 h. The product distribution (VC (1.0%), ethene (14.9%), and ethane (1.55) is consistent with c12DCE dechlorination via sequential hydrogenolysis reactions, although the potential for dechlorination by the β-elimination pathway cannot be discounted. Within 624 h,~44% of 11DCE was dechlorinated in the presence of CuGR, resulting in ethene (30.2%) and ethane (2.85). Ethene could be produced directly from 11DCE via an α-elimination reaction or via serial hydrogenolysis (Figure 1).
Although greater than in the presence of AuGR or unamended GR, the dechlorination of TCE in the CuGR system was limited, with~80% remaining after 625 h. No lesser chlorinated intermediates were detected, but the formation of ethene (11.1%) and ethane (1.1%) suggest the potential for multiple reaction pathways as shown in Figure 1. PCE concentrations decreased by~32% within 627 h accompanied by the formation of TCE (3.6%), ethene (3.3%) and, ethane (0.3%). Similar to TCE, the observed products of PCE reduction may result from multiple reaction pathways. As with AgGR, there was no evidence of the formation of radical coupling products.

Comparrison with Other Studies of Chlorinated Ethenes by Metal-Amended Green Rust
This is the first study to examine the dechlorination of the entire series of chlorinated ethenes by metal-amended green rusts; however, other studies have focused on PCE and TCE. No prior studies have examined chlorinated ethene reduction by Ag-amended green rust; however, Choi and Lee observed enhanced reductive dechlorination of PCE by green rusts amended with Pt(IV) [75]. PCE was essentially completely transformed to acetylene by chloride, fluoride, carbonate, and sulfate green rusts amended with Pt(IV), unlike our system with AgGR, in which TCE was the dominant product, with lesser amounts of t12DCE and acetylene (Table 2). Maithreepala and Doong examined the reduction of PCE and TCE in the presence of Cu-amended chloride green rust [72]. Within 35 days, >85% of PCE was removed, with TCE (16.6%) and ethene (31.95) as the major transformation products. Over the same period in TCE-amended systems,~50% of TCE was removed, with ethene (11%) and ethane (1%) as products. Although the overall extents of reaction were greater, the products formed were similar to those observed in our study.  [79] is from the PtGR SO4 system.
As discussed in Section 3.1, differences in the reactivity of unamended green rusts with respect to dechlorination of chlorinated ethenes has been linked to several factors, several of which have been investigated in relation to the efficiency of chlorinated ethene reduction by metal-amended green rusts. As with unamended green rust, pH was found to affect PCE removal efficiency, and the relationship between pH and PCE removal efficiency was not consistent among experimental systems. Maithreepala and Doong [72] reported that the rates of PCE removal by Cu(II)-amended GR Cl increased from pH 5.5 to 7.2, but decreased at pH 9, while Choi et al. reported the rates of PCE removal by C(II)-amended GR F highest at pH 11, with decreasing rates at pH 9 and 7.5 [79]. It is possible that the differences in pH effect on PCE removal rates between these two studies is related to the type of interlayer anion in the green rusts used in each, as differences in PCE removal rates have been reported for Cu(II)-and Pt(IV)-amended green rusts (F − >> CO 3 2− >> SO 4 2− > Cl − and F − > Cl − > SO 4 2− > CO 3 2− , respectively) [75,79]. Maithreepala and Doong reported that both the removal efficiency and the rate of PCE removal by Cu(II)-amended GR Cl increased with increasing amounts of green rust at a fixed Cu(II) concentration of 0.5 mM [72]. Similarly, the rates of PCE and TCE removal by Cu(II)-and Pt(IV)-amended green rusts increased with increasing metal loadings [72,75,79], although in many cases the rates decreased at the highest metal loadings examined, which may have been due to corresponding decreases in Fe(II) levels under these conditions [72,79]. Choi et al. [79] observed an initial increase followed by a leveling off of PCE removal rates by Cu(II)-amended fluoride green rust consistent with saturation of PCE sorption sites at the higher PCE concentrations.
The product distributions of chlorinated ethene reduction by metal-amended green rusts varied based on the metal amendment (Tables 1 and 2), as has been reported for the reduction of chlorinated methanes and ethanes [57,73]. Enhanced reduction of chlorinated hydrocarbons has been reported for Ag(I)-, Au(III)-, Cu(II)-, and Pt(IV)-amended green rust, and in each system the metals are reduced by green rust to their zero valent (ZV) forms (i.e, Ag(0), Au(0), Cu(0), and Pt(0)) [58,72,75]. The enhanced reduction of chlorinated hydrocarbons by green rust-ZVAg/Au/Cu/Pt systems may be comparable to the enhanced chlorinated hydrocarbon reduction observed with bimetallic reductants [94][95][96][97][98][99] consisting of a ZV noble metal (e.g., Pd(0), Pt(0), or Ni(0)) with a more active metal (e.g., Fe(0) or Zn(0)). The specific mechanism(s) leading to the enhanced effectiveness of these bimetallic systems have not been fully characterized [100]. However, it is known that coupling a noble metal with a more reactive metal accelerates the oxidation of the more reactive metal, forming a galvanic cell wherein the active metal serves as the anode and the noble metal acts as the cathode (where electrons are transferred to the oxidant (e.g., chlorinated hydrocarbons). Such a process has been proposed for the catalytic activity of bimetallic reductants [101], and a similar process may be occurring in the green rust-ZVAg/Au/Cu/Pt systems. Furthermore, Liu et al. have reported that the kinetics and product distribution of the electrolytic reduction of carbon tetrachloride varies with the metal used for the cathode [102], which is consistent with the differences in the rates/extents and product distributions observed with different green rust-ZV metal systems.

Potential Utility of Metal-Amended Green Rusts for Remediation of Chlorinated Ethenes
Ag(I)-and Cu(II)-amended green rusts have been shown to be effective reductants for the dechlorination of many chlorinated methanes and ethanes [57,72,73]. Under our experimental conditions, these materials (with a couple of notable exceptions) did not demonstrate a level of reactivity that would make them suitable for use in remediation of chlorinated ethene contamination. However, under different conditions (e.g., pH and interlayer anion), metal-amended green rusts have demonstrated much higher levels of PCE and TCE removal [72,75,79] than observed in this study, suggesting that the effectiveness of Ag(I)-and Cu(II)-amended green rusts for removal of chlorinated ethenes could be improved and warrants further investigation. Given the potential ecotoxicology issues involved with the use of Ag and Cu, as well as the costs associated with precious metals like Ag, Au, and Pt, the use of metal-amended green rusts for remediation of chlorinated hydrocarbons is probably best suited to ex situ applications, as this would limit the release of metals to the environment and allow for easier recovery of the Ag, Au, or Pt.