Electrochemical Reduction and Oxidation of Chlorinated Aromatic Compounds Enhanced by the Fe-ZSM-5 Catalyst: Kinetics and Mechanisms.

Devising cost-effective electrochemical catalyst system for the efficient degradation of chlorinated aromatic compounds is urgently needed for environmental pollution control. Herein, a Fe-ZSM-5 zeolite was used as a suspended catalyst to facilitate the degradation of lindane as a model chlorinated pesticide in an electrochemical system consisting of the commercial DSA (Ti/RuO2-IrO2) anode and graphite cathode. It was found that the Fe-ZSM-5 zeolite greatly accelerated the degradation of lindane, with the degradation rate constant more than 8 times higher than that without Fe-ZSM-5. In addition, the Fe-ZSM-5 zeolite widened the working pH range from 3 to 11, while efficient degradation of lindane in the absence of Fe-ZSM-5 was only obtained at pH ≤ 5. The degradation of lindane was primarily due to reductive dechlorination mediated by atomic H* followed by •OH oxidation. Fe-ZSM-5 zeolite could enrich lindane, H*, and •OH on its surface, thus provided a suitable local environment for lindane degradation. The Fe-ZSM-5 zeolite exhibited high stability and reusability, and reduced the energy consumption. This research provides a potential reduction-oxidation strategy for removing organochlorine compounds through a cost-efficient Fe-ZSM-5 catalytic electrochemical system.

Introduction

Chlorinated aromatic compounds such as insecticides, herbicides, and pharmaceuticals have been widely detected in the environment as they have been widely used over the past decades.[1,2] These chemicals usually have strong biotoxicity, bioaccumulation, and teratogenic mutagenicity.[3] Lindane (γ-hexachlorocyclohexane), a representative organochlorine pesticide, has been widely used in agricultural, veterinary, and human applications between the 1940s and 1990s.[4,5] Although lindane has been banned in many countries, it is still widely detected in surface waters, groundwaters and soils because of a large number of applications in the past.[6,7] Due to its refractoriness and potential hazardous risk, lindane was included in the Stockholm Convention as a persistent organic pollutant and challenging to remove from the environment. Therefore, highly effective remediation technologies are urgently required for this kind of chlorinated aromatic pollutants.

Although the intrinsic characteristics of the molecule, several techniques including incineration,[8] microbial dehalogenation,[9] and chemical oxidation[10] have been used for chlorinated aromatic compound degradation in waters or soils. Chemical oxidation with radical pathways, especially hydroxyl radicals (•OH), is frequently reported for the oxidation of lindane.[10]•OH could be generated by ozonation, photocatalytic reaction, Fenton reaction, and plasma oxidation.[11] Fenton processes caused decomposition of H2O2 to •OH catalyzed by Fe2+/Fe3+ but were strongly dependent on pH. The photo-Fenton process was another way to generate •OH but also required strict pH control.[11] The electro-Fenton process is one of the most powerful methods for H2O2 generation.[12,13] The process still requires acidic pH conditions (i.e., pH 2.8–3.5).[14] Moreover, it generates a lot of iron-containing sludge.[15] Heterogeneous electro-Fenton processes can widen the operational pH range.[16] H2O2 would be transformed to •OH on the surface of the catalyst in the heterogeneous electro-Fenton process. The catalysts include iron minerals, iron-supporting materials, nanoparticles, and modified cathodic materials.[17] Among them, iron- and iron oxide-loaded materials such as layered double hydroxide containing Cu and Fe, Y–Fe,[18] zeolite catalyst,[19] Fe-clay,[20] and magnetic single-walled carbon nanotubes[21] have been proven to be efficient catalysts.[17]

Chemical reduction of chlorinated aromatic compounds has also been explored largely by reducing agents such as zero-valent iron,[22,23] bimetallic Fe-based nanoparticles,[24] and natural reducing agents.[25] The atomic H* can realize the dechlorination of lindane[26−29] and can be generated on noble metals in an electrochemical system to dechlorinate chlorinated organic compounds.[30−32] Palladium was the most favorable catalyst for its excellent capability in retaining H*.[30,33] However, the poor dispersibility and high cost of Pd metals limit its application.[31] Although H* could initiate dechlorination, the reaction could not continue readily for the high energy required.[26]

It was found that the coexistence of H* and •OH could decrease energy consumption and increase the degradation efficiency of pollutants.[26,34] Such synchronous reduction–oxidation systems have been constructed using a Pd-loaded Cu/Cu2O/CuO electrode,[26] a CoNi anchored on the ionic liquid functionalized graphene electrode,[34] and a Pd/graphene oxide hybrid fabricated on carbon fiber paper.[29] H* and •OH are generated through the hydrogen evolution reaction (HER), oxygen evolution reaction (OER), and oxygen reduction reaction (ORR) in electrochemical systems.[35,36] Boron-doped diamond (BDD) as a promising electrode material, has excellent properties for generation H* and •OH.[37,38] However, noble metals and BDD electrodes are more expensive than ruthenium oxide and/or iridium oxide layers based on dimensionally stable anodes (DSAs) and graphite electrodes.[39] DSAs and graphite electrodes are widely explored to remove organic compounds and generally are used in industrial electrochemical processes.[40−44] Thus, there will be a great need to use these commercial electrodes for industrial applications if H* and •OH could be simultaneously generated in such electrochemical systems.

Electrocatalysts have been proven to affect the rate of chemical reactions and promote the generation of active species.[45] Powder-activated carbon loaded with metallic cobalt (Co) could catalytically generate both H* and •OH in a three-dimensional electrocatalytic system.[29,46] Pd/Al2O3 particles provided a large surface area for atomic H* production, which was responsible for Fe3+ reduction to Fe2+.[47] Nowadays, transition metals such as Fe, Co, Ni, Cu, and Mn exhibit efficient catalytic activity for electrochemical reactions which are cheaper and easier to obtain than noble metals.[26,46,48] Metal oxide-based materials such as TiO2, Ti2O3, and MnO2 also served as efficient HER electrocatalysts associated with H* adsorption/desorption via either Volmer–Heyrovsky or Volmer–Tafel mechanisms.[49] Fe-ZSM-5 zeolites have been widely explored in the heterogeneous Fenton process due to their highly catalytic ability,[50−56] which have a potential for industrialization.[57] Thus, it is significant to explore cost-efficient materials as electrocatalytic particles to degrade organochlorine compounds.

The main objective of this paper is to promote the usage of atomic H* and •OH for the effective degradation of lindane by Fe-ZSM-5 particles. Lindane was selected as a model of chlorinated aromatic compounds. The degradation of lindane was performed in a heterogeneous catalytic electrochemical system over the Fe-ZSM-5 zeolite catalyst using commercially available electrodes. The effects of current density, solution pH, and amounts of Fe-ZSM-5 catalysts on the lindane degradation were studied. The energy consumption and catalyst reusability were studied as well.

Materials and Methods

Chemicals and Materials

ZSM-5 zeolites with Si/Al ratios of 25 were supplied by Tianjin Nankai University Catalyst Co., Ltd. Lindane (97%), 5,5-dimethyl-1-pyrroline N-oxide (DMPO), isopropyl alcohol (IPA), tert-butanol (TBA), and 1,4-benzoquinone (PBQ) were obtained from Sigma-Aldrich. Other chemicals were purchased from Sinopharm.

Preparation and Characterization of Fe-ZSM-5 Zeolites

The Fe-ZSM-5 catalyst was synthesized according to the previous work with a minor modification.[58] In brief, ZSM-5 zeolites (200 g) were impregnated with Fe(NO3)3 aqueous solution (17.8 g Fe(NO3)3·9H2O), and the mixture was stirred for 4 h and then evaporated at 70 °C. Then, the samples were calcined for 4 h at 550 °C in a muffle furnace to obtain Fe-ZSM-5 zeolites. The Fe elements on the surface of Fe-ZSM-5 were characterized by using an ESCALAB 250Xi spectrometer (Thermo Scientific, USA). After Fe-ZSM-5 was degassed at 120 °C for 12 h, the surface area and porosity of Fe-ZSM-5 zeolites were characterized by N2 adsorption–desorption isotherms measured on an ASAP2010 Accelerated Surface Area and Porosimetry Adsorption Analyzer (Micromeritics). The surface morphology and elements of Fe-ZSM-5 zeolites were observed using a LEO 1530 VP scanning electron microscope at the Nanjing Institute of Geology and Paleontology. The XRD patterns (SmartLab 9, Rigaku Corporation in Japan) of Fe-ZSM-5 were detected at a scan rate of 2° min–1 under 40 kV and 150 mA. UV–vis diffuse reflectance spectroscopy (UV–vis DRS) of Fe-ZSM-5 was obtained using a UV–vis–NIR 3600 Plus spectrophotometer (Shimadzu Corporation, Kyoto, Japan).

Electrochemical Experiments

Electrochemical experiments were performed in a cylindrical glass reactor (150 mL) with screw caps working with magnetic stirring. A graphite plate cathode (2.5 cm × 2 cm × 1 mm) was purchased from Qingdao Baofeng Graphite Corporation, and a DSA (mixed metal oxide, Ti/RuO2–IrO2) anode (2.5 cm × 2 cm × 1 mm) was purchased from Yiwanlin Electronics Technology Corporation. A sodium sulfate aqueous solution (3 g/L) was used as the supporting electrolyte according to the previous work.[59]

The effects of current density (from 6 to 30 mA/cm2), initial pH (from 3 to 11 adjusted by HCl or NaOH), and doses of Fe-ZSM-5 zeolites (from 0.1 to 1.0 g/L) on the lindane degradation process were studied. 0.5 mL samples were taken off from the reaction solution with magnetic stirring at specific times and mixed with 1 mL n-hexane, and the extract was measured by gas chromatography with an electron capture detector (GC-ECD). Scavenging experiments were conducted by different scavengers (IPA, TBA, and PBQ) to estimate the role of reactive radicals. The degradation intermediate solution was collected at 5 min, 10 min, 30 min, 1 h, 2 h, 4 h, and 8 h and mixed together. The mixture was extracted with n-hexane and was ready for gas chromatography/mass spectrometry (GC–MS). To study the dechlorination of lindane, the initial concentration of 5 mg/L was selected, and Cl– concentrations were detected using Thermo Scientific Dionex Aquion Ion. The adsorption–desorption of lindane onto Fe-ZSM-5 was also studied, and a detailed procedure is shown in the Supporting Information. All the experiments were performed in triplicate.

The degradation reactions of lindane were modeled using the first-order kinetic equation.[60−64]where C0 and C are the initial and residual concentrations of lindane at a specific time t, respectively. k is the rate constant of the degradation reaction calculated from the slope of ln(C/C0) versus time t.

Analytical Methods

Lindane concentration was measured by an Agilent 7890A GC-ECD equipped with a CP-Sil 8 CB column (30 m × 0.25 mm × 0.25 μm), and intermediates of lindane degradation were analyzed by GC–MS (TQ-8050, Shimadzu). Operating conditions are shown in the Supporting Information. Possible products were identified by retrieving the mass spectra in the NIST standard library. Total organic carbon (TOC) in the solution was detected by Analyzer multi N/C UV (Analytic Jena, Germany) according to the 800 °C combustion catalytic oxidation method. Electrochemical energy consumption (EEC) for TOC removal during electrolysis at a constant current was computed from experimental TOC versus time according to the following equations (eq ).[13,25]where U is the average cell voltage (V), I is the applied current (A), t is the electrolysis time (h), V is the solution volume (L), and ΔTOC is the experimental TOC decay (mg/L) reached at time t.

The electron paramagnetic resonance (EPR) signals of radicals trapped by DMPO were detected using an E500-9.5/12 EPR spectrometer. Detailed information is shown in the Supporting Information.

Results and Discussion

Characterization of Fe-ZSM-5

Nitrogen sorption–desorption isotherms of Fe-ZSM-5 are shown in Figure S1A. Fe-ZSM-5 showed type I adsorption isotherms and type H4 hysteresis loops, suggesting a microporous property (Figure S1A), and the pore size distribution of Fe-ZSM-5 by applying the Horvath–Kawazoe model ranged from 0.50 to 0.98 nm (Figure S1B). The BET specific surface area and total pore volume of Fe-ZSM-5 were 403.6 m2/g and 0.215 cm3/g, respectively. The main pore size was 0.545 nm (Table S1). The XRD pattern of Fe-ZSM-5 showed typical diffraction peaks attributed to traditional ZSM-5 MFI topology (JCPDS: no. 49-0657), such as the peaks at 7.9, 8.8, 23.0, 23.2°, and so forth. Weak diffraction peaks (33.2, 40.9, and 49.5°) attributed to Fe2O3 (JCPDS: no. 033-0664) were also observed (Figure A). The XPS spectrum of Fe-ZSM-5 (Figure S1C) revealed the composition and chemical states of Fe on its surface. The binding energy peaks located around 711 and 724.7 eV in Fe-ZSM-5 corresponded to the Fe 2p3/2 and 2p1/2 states of iron in Fe2O3.[65,66] The spectrum of Fe 2p3/2 could be deconvoluted into two peaks at 710.9 and 713.6 eV.[65] The weak peak at 718.5 eV was related to the Fe3+ oxidation state.[65,66] Chemical states of Fe on the ZSM-5 surface were further confirmed by UV–vis diffuse reflectance spectroscopy of Fe-ZSM-5 (Figure B). An adsorption band at 545 nm appeared, showing the existence of Fe2O3 particles on the ZSM-5 surface,[67,68] and an apparent adsorption band at about 280 nm showing isolated octahedral Fe3+ complexes.[68,69] A weak band at about 400 nm was observed, showing an aggregation of Fe2O3 clusters.[68,70] The SEM morphology of ZSM-5 zeolite showed cube-like crystals with a smooth surface (Figure C), and the SEM morphology of Fe-ZSM-5 zeolite showed few Fe2O3 particles on the surface of ZSM-5 (Figure D). The EDS elemental percentage of ZSM-5 and Fe-ZSM-5 are shown in Table S2, and the appearance of Fe was observed on the Fe-ZSM-5 catalyst. Thus, Fe2O3 was formed on the surface of ZSM-5.

Figure 1

Characterization of ZSM-5 and Fe-ZSM-5. (A) XRD patterns of Fe-ZSM-5; (B) UV–vis diffuse reflectance spectroscopy of Fe-ZSM-5; (C) SEM morphology of ZSM-5; (D) SEM morphology of Fe-ZSM-5.

Lindane Degradation Kinetics

The degradation of lindane was conducted at different current densities, Fe-ZSM-5 doses, and initial pH values (Figure ). Lindane degradation efficiencies of 58.2, 70.0, 77.4, 84.6, and 86.7% with free Fe-ZSM-5, 0.1, 0.3, 0.5, and 1.0 g/L Fe-ZSM-5, respectively, were achieved in 2 h (Figure ). All the data can be well fitted with first-order kinetic model (Tables S3 and S4). The degradation of lindane increased with the increase of Fe-ZSM-5 dose (Figure A). When the Fe-ZSM-5 zeolite dose was increased from 0.1 to 1.0 g/L, the degradation rate constants increased from 0.6494 to 0.8343 h–1 (Table S3). When the Fe-ZSM-5 dose was over 0.5 g/L, no further apparent increase of the degradation rate constant was observed. The degradation rate constant with 0.5 g/L Fe-ZSM-5 (0.8284 h–1) was close to that with 1.0 g/L Fe-ZSM-5 (0.8343 h–1).

Figure 2

Degradation kinetics of lindane. (A) Doses of Fe-ZSM-5 ranging from 0.1 to 1.0 g/L at 20 mA/cm2; (B) current density ranging from 6 to 30 mA/cm2 without Fe-ZSM-5; (C) current density ranging from 6 to 30 mA/cm2 with 0.5 g/L Fe-ZSM-5; (D) initial pH ranging from 3 to 11 with 0.5 g/L Fe-ZSM-5 at 20 mA/cm2; (E) initial pH ranging from 3 to 11 without Fe-ZSM-5 at 20 mA/cm2.

Lindane was also slowly degraded in the electrochemical oxidation system in the absence of Fe-ZSM-5 (Figure B). Lindane degradation in 2 h was 19.8, 24.1, 58.2, and 81.7% at the current densities of 6, 14, 20, and 30 mA/cm2, respectively. Also, the degradation rate constant values at a current density from 6 to 30 mA/cm2 were in the range of 0.0745 to 0.6367 h–1 (Table S4). Compared to the degradation in the absence of Fe-ZSM-5, lindane degradation was accelerated greatly in the presence of Fe-ZSM-5 zeolites (Figure C). The lindane degradation rate constant increased from 0.6230 to 0.8902 h–1 when the current density was increased from 6 to 30 mA/cm2 (Table S3). The degradation rate constant (0.8902 h–1) at a current density of 30 mA/cm2 was not much higher than that at a low current density of 6 mA/cm2 (0.6230 h–1) with Fe-ZSM-5. At a current density of 6 mA/cm2, lindane was almost completely degraded in 6 h in the presence of Fe-ZSM-5, while the degradation efficiency was less than 40% without Fe-ZSM-5. The degradation rate constant of lindane (0.6230 h–1) with Fe-ZSM-5 was more than eight-fold higher than that (0.0745 h–1) without Fe-ZSM-5 at 6 mA/cm2, close to that (0.6367 h–1) obtained at a current density of 30 mA/cm2 in the absence Fe-ZSM-5. Thus, adding Fe-ZSM-5 to the electrochemical system can accelerate lindane degradation, especially at low current densities.

The effects of initial pH on lindane degradation with or without Fe-ZSM-5 are shown in Figure D,E. Lindane was effectively degraded with Fe-ZSM-5 over a pH range of 3–11, with the degradation rate constant ranging from 0.6825 to 0.9458 h–1 (Figure D). The degradation became slower with increasing pH regardless of the existence of Fe-ZSM-5 or not. In the absence of Fe-ZSM-5, a high degradation rate of lindane was only obtained at pH ≤ 5 (Figure E), and the degradation rate constant at pH 11 was only 0.1812 h–1. Thus, the presence of Fe-ZSM-5 as an electrochemical catalyst can widen the working pH range. It is noted that the pH after 8 h was in the range of 2.89–4.22 when the initial pH was from 3 to 11 (Figure S2), suggesting the generation of acidic byproducts.

Identification of Active Radicals

To identify the radicals involved in the heterogeneous catalytic electrochemical process, EPR spectroscopic analysis was conducted with DMPO as the spin trap agent. The DMPO-H adduct characterized by nine peaks with a g-factor value of 2.00607 was evident at different reaction times (Figure A). The DMPO/•OH adduct characterized by four peaks (1:2:2:1) with a g-factor of 2.00625 was found at all the reaction times (Figure A), but the DMPO/•O2– adduct (1:1:1:1 quartet) was not observed. At 5 min (0.083 h), the signal intensity of DMPO-H was stronger than that of DMPO/•OH. However, it shrank over time and became lower than that of DMPO/•OH after 30 min. Atomic H* can be generated by the electroreduction of H2O or H+.[71] Moreover, both atomic H* and •OH were adsorbed on the surface of Fe-ZSM-5 because characteristic peaks of DMPO-H and DMPO/•OH were not observed in the solution with Fe-ZSM-5 excluded, while the signals of DMPO-H and DMPO/•OH in the solution containing catalyst Fe-ZSM-5 were high (Figure B). The result indicated that Fe-ZSM-5 could enrich atomic H* and •OH.

Figure 3

EPR spectra of radicals (A) at different reaction times ranging from 0.083 to 4 h; (B) in the filtrate and the solution containing Fe-ZSM-5 at 0.083 h.

Degradation Mechanisms and Pathways

The reactive •OH was usually regarded as the dominant oxidative species in the catalytic electrochemical systems, which could be scavenged by both IPA and TBA.[32] To evaluate the contribution of •OH to lindane degradation, IPA and TBA were added to the reaction solution as the radical scavengers. The degradation efficiency of lindane was apparently inhibited in the presence of IPA and TBA (Figure A), and the inhibition became more significant with the increase of the concentration of scavengers (Figure S3A,B). In the first 0.083 h, lindane degradation with IPA was higher than that with TBA at an equal concentration (Figure A). It is known that IPA reacts with •OH at a higher rate than TBA.[72,73] If •OH was the only species that was quenched, the result would be the opposite. Thus, there were other species responsible for lindane degradation in this study. Because H* was generated during the electrochemical process in this study (Figure B), and Mao et al. reported that TBA could react with H* rapidly, while H* could be formed because of water reduction in the cathode of the electrochemical reactor.[32] Therefore, it can be postulated that H* also contributed to lindane degradation, especially during the earlier phase of the reaction. HO2•/O2•– was another oxidative species that might exist in the catalytic electrochemical process. In this study, PBQ was applied to probe HO2•/O2•–. The results showed that the degradation rate of lindane was little affected by PBQ (Figure S3C), indicating that HO2•/O2•–, if formed, played a minor role in the degradation of lindane. Although with a high concentration of radical scavengers, lindane degradation was not completely suppressed (Figure A). This can be explained by the direct oxidation of the organic compounds through electron transfer on the anode. A similar result was observed: chlorinated hydrocarbons could be degraded through a direct oxidation pathway on anodes.[74,75]

Figure 4

Effect of scavengers on lindane degradation. (A) A comparison of lindane degradation with scavengers and 0.5 g/L Fe-ZSM-5 at 20 mA/cm2; (B) concentration of Cl– during the degradation process with scavengers at 20 mA/cm2 and 0.5 g/L Fe-ZSM-5.

To further explore the roles of H* and •OH, the concentrations of inorganic chlorine species during the degradation process were measured (Figure B). The time-course concentration of Cl– showed a similar trend in the presence and absence of scavengers; that is, the concentration of Cl– in the solution first increased and then decreased. This phenomenon is consistent with the results reported in the literature that dechlorination of lindane resulted in the formation of Cl–, and Cl– could be further oxidized to free chlorine and species with even higher oxidative states by •OH, such as ClO2–, ClO3–, and ClO4–, in electrochemical processes.[12,13,76] The concentration of Cl– in the solution without scavengers was higher than those with scavengers in the first hour, which means that it was easy to dechlorinate lindane without scavengers. However, the concentration of Cl– without scavengers was much lower than that with scavengers after 1 h. Both IPA and TBA can consume •OH, which would prevent the oxidation of Cl– to a higher state.

Although the time-course concentration of Cl– showed a similar trend in the presence of IPA or TBA, the concentration of Cl– with TBA was lower than that with IPA in the whole reaction process (Figure B). As a reduction agent, H* caused lindane to dechlorinate and release Cl–. TBA could consume H*, thus inhibiting reductive dechlorination of the targeted pollutant. As a result, Cl– concentration with TBA was lower than that with IPA. Therefore, H* played an important role during lindane degradation. Lindane degradation in the presence of TBA was expected to be slower than that in the presence of IPA if H* acted as the primary reactive species governing lindane degradation. However, degradation of lindane with TBA was significantly faster than that with IPA after the first hour. This indicates that there were other reactive species contributing to lindane degradation, especially in the later reaction phase. Both IPA and TBA can quench •OH, but IPA reacts with •OH more rapidly than TBA.[72] As a result, further transformation of Cl– to other states by •OH in the presence of IPA would be slower than that in the presence of an equal amount of TBA. More •OH were consumed in the presence of IPA, which made lindane degrade faster in the presence of TBA than IPA after the first hour although reductive dechlorination was inhibited by TBA. Thus, the lindane degradation process was initiated first by atomic H* dechlorination, followed by the oxidation of •OH.

In this study, H* and •OH in the filtrate solution were very low but high in the solution with Fe-ZSM-5 (Figure B), suggesting Fe-ZSM-5 could stabilize and enrich the H* and •OH. Fe-ZSM-5 could also adsorb lindane (Figure S4). Less than 20% lindane of 0.8 mg/L initial concentration was adsorbed onto Fe-ZSM-5, and the equilibrium was reached in 5 min (Figure S4). About 11% of adsorbed lindane could be desorbed from Fe-ZSM-5 with DI water wash (Figure S5). Thus, the Fe-ZSM-5 surface provided a local environment where both lindane and reactive species were enriched. This is similar to the previous findings that Pd/Fe3O4 or Pd–In nanoparticles could be used as electron and hydrogen atom carriers facilitating the degradation of organic compounds.[30,32] XPS spectra of Fe in spent Fe-ZSM-5 became broader, and its intensity decreased(Figure S6), and the Fe concentration on the surface of spent Fe-ZSM-5 through XPS surface composition analysis decreased, showing that Fe2O3 was consumed and released during the electrochemical process (Table S5) and that Fe2O3 may be used as an active site. It was reported that Fe2O3 used as the active site such as in Fe2O3–N-doped carbon, Fe2O3/FeP, Fe2O3/NiFe-LDHs, and so forth had a better electrocatalytic performance for the HER and OER,[77−79] which usually occur during the electrolytic process.[80,81] During the HER and OER processes, H* can be generated from the reduction of water through the Volmer process (eqs –5),[29,82] and •OH could be generated on the DSA anode by oxidation of water (eq ).[83]

A total of 11 reaction intermediates including pentachlorocyclohexene (PCCH), tetrachlorocyclohexene (TeCCH), trichlorobenzene (TCB), dichlorobenzene (DCB) (including 1,2-dichlorobenzene, 1,3-dichlorobenzene, and 1,4-dichlorobenzene), chlorobenzene, 4-chlorocyclohexene, and dichlorophenol (2,4-dichlorophenol, 2,5-dichlorophenol, and 2,6-dichlorophenol) were identified by GC–MS (Figure S7). Based on this, the degradation pathway of lindane is proposed (Figure ). With six chlorine atoms substituted, lindane is extremely electron-negative and thus resistant to oxidation. Its degradation is initiated with reductive dechlorination, generating PCCH, TeCCH, TCB, and DCB sequentially. With the decrease in the number of chlorine substitutions, the molecules became less electron-negative. When only two chlorine substitutions remained in the molecule, oxidation became possible. Thus, DCB could be oxidized to form hydroxylated products. The •OH probably played a central role in this step. Further oxidation of the molecule led to the breakdown of the ring, forming short-chain aliphatic acids such as hydracrylic acid, pyruvic acid, glyoxylic acid, and oxalate which were the last byproducts before mineralization, lowering the solution pH. This is confirmed by the final decreased pH (Figure S2).

Figure 5

Proposed degradation mechanism of lindane.

Energy Consumption and Mineralization of Lindane

The degree of mineralization of lindane was evaluated based on the reduction of TOC (Figure ). The mineralization efficiency of lindane increased with the current density increasing (Figure A). The TOC of the solution was reduced by 31.1, 50.0, 70.5, and 77.9% at the current densities of 6, 14, 20, and 30 mA/cm2, respectively, in the presence of 0.5 g/L Fe-ZSM-5 at 8 h. The mineralization efficiency of lindane also increased with the increase of the Fe-ZSM-5 dose(Figure B). The TOC degradation efficiencies were 52.5, 60.7, 70.5, and 76.2% at doses of 0.1, 0.3, 0.5, and 1.0 g/L Fe-ZSM-5, respectively. ECTOC increased with the increase of current densities (Figure C), but ECTOC decreased with increasing doses of Fe-ZSM-5 (Figure D), showing that Fe-ZSM-5 can lower energy consumption. At 30 mA/cm2 and 0.5 g/L Fe-ZSM-5, the EC value at hour 8 (77.9% TOC removal) was 1.2 kW h (g TOC)−1, which was higher than the value (0.74 kW h (g TOC)−1) at 20 mA/cm2 (70.5% TOC removal). The EC values (<1.2 kW h (g TOC)−1) were much lower than those reported.[12,13] Our results are similar to the recent studies that redox synergy between H* and •OH effectively lowered energy consumption and improved the degradation of persistent refractory chlorinated organic compounds.[26,84] In addition, the reusability of catalysts is crucial for practical application. To evaluate the reusability of Fe-ZSM-5, experiments were conducted by repeatedly spiking lindane to the same solution (Figure S8), and the content of Fe in ZSM-5 decreased from 0.98 to 0.56%. Fe-ZSM-5 remains highly efficient after seven cycles, indicating the excellent catalytic reusability of Fe-ZSM-5 catalysts. Thus, the Fe-ZSM-5 catalyst can be used as a promising catalyst for the degradation of lindane in an electrochemical system.

Figure 6

Mineralization of lindane and energy consumption in the presence of Fe-ZSM-5. (A) Effect of current density on lindane mineralization; (B) effect of dose of Fe-ZSM-5 on lindane mineralization; (C) effect of current density on the energy consumption; (D) effect of dose of Fe-ZSM-5 on the energy consumption.

Conclusions

In this study, we reported a highly efficient catalytic electrochemical process consisting of commercial DSA (Ti/RuO2–IrO2) and graphite electrodes and Fe-ZSM-5 particles as the catalyst. Fe-ZSM-5 zeolites could stabilize and enrich the H* and •OH, and they can also enrich lindane on its surface. The degradation of lindane was initiated with the reductive dechlorination of lindane by H*, followed by •OH oxidation that led to the cleavage of the ring ultimately. Fe-ZSM-5 significantly accelerated lindane degradation and reduced energy consumption. These results are valuable for the design and application of synthetic iron-loaded catalysts for the reduction–oxidation of organochlorine compounds in electrochemical processes.