Characterization and treatment of landfill leachate membrane concentrate by Fe2+/NaClO combined with advanced oxidation processes

Landfill leachate membrane concentrate (LLMC) is a type of non-biodegradable wastewater intercepted by the membrane filtration of the landfill leachate membrane bioreactor (MBR) effluent. The concentrations of chemical oxygen demand (COD) and ammonia nitrogen (NH4+-N) in the LLMC collected from a landfill in Beijing were determined to be 4700 mg/L and 487 mg/L, 2–5 times higher than those in the MBR effluent. The photoelectro oxidation (PEO) followed with the NaClO enhanced Fe2+ coagulation were more effective for the removal of COD than the Fenton oxidation followed with the enhanced coagulation. The final removal efficiencies of COD, UV254, NH4+-N and color degree were 86%, 95%, 93% and 95% with Fe2+ (90 mmol/L) and NaClO (60 mmol/L, Fe2+:NaClO = 1.5:1), and PEO for 3 hours with a current density of 400 A/m2. Due to the existence of Cl−, the chlorinated intermediates, which would be more toxic, were detected in the PEO treatment. However, the intermediates could be eliminated finally. As a result, the NaClO enhanced Fe2+ coagulation treatment combined with PEO treatment was efficient for the treatment of LLMC.

the in-situ formation of ferric ion was more efficient for the coagulation of organic compounds than the direct addition of ferric salts 14 . NaClO was usually used for the oxidation of NH 4 + -N 15,16 . Meanwhile, NaClO could enhance the in-situ generation of ferric ion from Fe 2+ . NaClO enhanced Fe 2+ coagulation has been firstly used for the treatment of the EMBR 12 . The removal efficiencies of COD in the EMBR increased from 24% with Fe 2+ coagulation treatment to 54% with NaClO enhanced Fe 2+ coagulation 12 . Therefore, it was supposed that NaClO enhanced Fe 2+ coagulation would also be efficient for the treatment of LLMC with higher concentration of both COD and NH 4 + -N. In order to further increase the removal efficiency of COD and NH 4 + -N, advanced oxidation processes (AOPs) were needed. AOPs were attractive methods for the elimination of non-biodegradable organics, as well as increase the biodegradability of the LLMC 5 . After 110 min continuously ozone generation-reaction treatment, COD and NH 4 + -N could be removed 56% and 0%, respectively 5 . During the Fenton oxidation, the highest removal efficiency of COD in the LLMC was 60% 8 . While with the treatment of a two-stage enhanced coagulation followed with photoelectro oxidation (PEO), 86% of COD in the LLMC could be removed 17 . So we proposed that the combination of the coagulation and PEO might significantly improve the removal efficiency of COD.
Herein, the objective of this study was to seek for a suitable technique for treatment of the organic compounds and NH 4 + -N in the LLMC. A typical LLMC collected from a landfill in Beijing was characterized. NaClO enhanced Fe 2+ coagulation combined with two kinds of AOPs, including Fenton oxidation and PEO, were used for the treatment of LLMC. Specifically, the chlorinated intermediates were identified in the PEO and electro oxidation (EO) treatment processes.

Results and Discussion
Characterization comparison of the LLMC with the EMBR. The regular chemical parameters of LLMC and the EMBR are listed in Table 1.
The average concentrations of COD and NH 4 + -N in the LLMC were 4700 mg/L and 487 mg/L respectively, about 2-5 times higher than those in the EMBR (1200 mg/L and 90 mg/L). Compared with the LLMC in Zhejiang province and Qingdao, the concentrations of COD and NH 4 + -N in Beijing were much higher 5,11 . The LLMC was weakly alkaline (pH = 8.0), while the EMBR was weakly acidic (pH = 6.3). The concentrations of Cl − (10458 mg/L) were also much higher than those in Zhejiang (1000-3999 mg/L).
The DOM in the LLMC and EMBR were fractioned by resin (Fig. 1). The DOM of both LLMC and EMBR, characterized with UV 254 , were mainly composed with hydrophilic matter (HiM) and hydrophobic acids (HOA). Compared with the EMBR, the LLMC contained more HOA and less hydrophobic neutrals (HON) and hydrophobic bases (HOB). The hydrophobic fractions were easier eliminated than the hydrophilic fractions using the chemical treatment 12 .  The ratios of the absorbance A 253 /A 203 reflected the types of the aromatic ring substituent groups, including the aliphatic chain with lower A 253 /A 203 values, the carbonyl, carboxyl, hydroxyl and esters with higher A 253 / A 203 values 18 . It was reported that the A 253 /A 203 ratios equaled to 0.02, 0.21 and 0.42 were estimated for aliphatic, hydroxyl and carboxylic substitutions of DOM, respectively 19 . Compared different fractions in LLMC and EMBR, HOA > HOB > HON > HiM, indicating that the substituent groups in the HiM fraction was mainly aliphatic chain, and the content of carbonyl, carboxyl, hydroxyl and esters were higher in the HOA fraction (Table 2). Similar result was obtained by a previous study 20 . The LLMC contained more carbonyl, carboxyl, hydroxyl and esters than the EMBR. Treatment of the LLMC using NaClO enhanced Fe 2+ coagulation. NaClO enhanced Fe 2+ coagulation effectively improved the removal efficiencies of COD and NH 4 + -N (53% and 80%) compared with individual Fe 2+ coagulation (28% and 25%) or NaClO oxidation (20% and 40%) (Fig. 2). Fe 2+ could eliminate organic compounds through coagulation. NaClO could oxidize organic compounds into small molecular weight matters, CO 2 and H 2 O. Additionally, the formation of ferric hydroxide from Fe 2+ could be enhanced by NaClO besides the dissolved oxygen in water Eq. (1) 12 . Thus, the removal of organic compounds and NH 4 + -N could be enhanced through coagulation. From the aspect of ammonia, NH 4 + -N could be oxidized by sufficient HClO into N 2 finally Eqs (2-6) 12 . The removal mechanism was similar to the "Breakpoint Reaction" 21 . Therefore, the removal efficiencies of COD and NH 4 + -N in the NaClO enhanced Fe 2+ coagulation treatment were 1-2 times higher than the individual Fe 2+ coagulation or NaClO oxidation treatment.  The removal efficiencies of color degree were higher than COD and NH 4 + -N (78-94%). The NaClO oxidation treatment (94%) was more effective than Fe 2+ coagulation treatment (78%). Additionally, the removal efficiency of color degree by the NaClO enhanced Fe 2+ coagulation treatment (92%) was similar the NaClO oxidation treatment, indicating that the organic matters with chromophore were mainly oxidized by NaClO.
To obtain the optimum condition for the removal of COD, UV 254 , NH 4 + -N and color degree by NaClO enhanced Fe 2+ coagulation treatment, different dosages of Fe 2+ and NaClO were conducted (Fig. 3). It could be noticed that with the same dosage of Fe 2+ , the removal efficiencies increased with the increase of NaClO dosage. Generally speaking, when the mole ratios of Fe 2+ to NaClO [M (FeSO 4 :NaClO)] were lower than 1.5:1, the removal efficiencies increased slowly. Thus, the optimum M (FeSO 4 :NaClO) was 1.5:1. With the same M (FeSO 4 :NaClO), the removal efficiencies increased with the increase of Fe 2+ dosage, and did not significantly increase from 90 mmol/L to 120 mmol/L, especially for NH 4 + -N. Additionally, the sediment yield increased much with the increase dosage of Fe 2+ dosage, probably due to the increase of the polymer yielded by Fe 2+ . Therefore, the condition of M (FeSO 4 :NaClO) = 1.5:1 and the dosage of Fe 2+ = 90 mmol/L was deemed as the optimum. The removal efficiency of COD was 53%, the concentration of which decreased from 4700 mg/L to 2213 mg/L. The removal efficiency of UV 254 was 79%, higher than that of COD, indicating that NaClO enhanced Fe 2+ coagulation treatment was more effective for the removal of unsaturated organic matters determined by UV 254 . The removal efficiencies of NH 4 + -N was 80% with a final concentration of 93 mg/L. The removal efficiency of color degree was 92%.
Further treatment with the Fenton oxidation and the photoelectro oxidation. To further remove the pollutants in the LLMC after the NaClO enhanced Fe 2+ coagulation treatment, the Fenton oxidation and the PEO treatment was used. The supernatant of the effluent from the NaClO enhanced Fe 2+ coagulation treatment was used for the further treatments. Fig. 4 that the removal efficiencies of COD and UV 254 increase with the increase dosage of Fe 2+ and H 2 O 2 . More ·OH was generated with a higher dosage of Fe 2+ and H 2 O 2 Eqs (7 and 8) 22,23 , resulting to the higher removal efficiencies of the dissolved organic matters. However, as the concentrations of Fe 2+ growing higher, the removal efficiencies increased slowly, because the ·OH was consumed by the surplus Fe 2+ Eqs (9 and 10) 24 .  It could be noticed that the dosage of Fe 2+ of 100 mmol/L did not increase much the removal efficiency of the COD and UV 254 (78% and 97%) compared with 80 mmol/L Fe 2+ (77% and 96%), although the removal efficiencies of COD and UV 254 in different conditions were significantly different (Paired sample t-test, p < 0.05) (Table S1). A higher dosage of Fe 2+ would yield a higher amount of sludge. Therefore, the dosage of Fe 2+ of 80 mmol/L was preferred. The highest removal efficiencies of COD and UV 254 were obtained at the mole ratio of Fe 2+ to H 2 O 2 of 1:10. As a result, the best condition was the dosage of Fe 2+ of 80 mmol/L and the mole ratio of Fe 2+ to H 2 O 2 of 1:10, from the aspect of COD (77%) and UV 254 (96%). Compared with the Fenton oxidation (60%), the combined treatment of coagulation and Fenton oxidation for the removal of COD was much better (77%) 8 . The removal efficiencies of NH 4 + -N and color degree did not vary much with different dosage of Fe 2+ and H 2 O 2 (Paired sample t-test, p > 0.05) ( Table S1). The final removal efficiencies of NH 4 + -N and color degree was 91% and 95%, respectively. Compared with the same treatment to the EMBR in our previous study, the final removal efficiency of COD was 86%, higher than the LLMC (77%) 12 . Thus, the PEO treatment was used in the following section.

Fenton oxidation treatment with different dosage of Fe 2+ and H 2 O 2 . It is shown in
Photoelectro oxidation treatment. In order to improve the removal efficiency of COD, the PEO was used for the treatment of LLMC after NaClO enhanced Fe 2+ coagulation treatment. The EO was compared with the PEO treatment (Fig. 5). It could be noticed that under the same current density, the removal efficiencies of COD were higher in the PEO treatment than in the EO treatment. The removal efficiencies of COD increased with the increase current density from 200 A/m 2 to 400 A/m 2 , both in the PEO and EO treatment. After 180 min, the final removal efficiency of COD was 86% with a final concentration of 660 mg/L. The generation of active chlorine in the PEO was higher than that in the EO treatment (Fig. S1), which might result to the higher removal efficiency of COD in the PEO treatment.
Compared with the Fenton oxidation treatment after the enhanced coagulation treatment, the removal efficiencies of COD were about 10% higher in the PEO treatment (Fig. 6). While the removal efficiencies of UV 254 ,

NH 4
+ -N and color degree were relatively similar between the two treatments. Compared with ozonation (56% for COD and 0% for NH 4 + -N) and Fenton oxidation (60% for COD), the combination of PEO and enhanced coagulation treatment were more effective for the removal of COD and NH 4 + -N in the LLMC (86% and 93% respectively) 5,8,17 .
The reusability performance of the electrode was tested. The removal efficiencies of COD, UV 254 , NH 4 + -N and color degree did not vary much (sd < 3%) for the PEO treatment at 400 A/m 2 (Figs 6 and S2). The RuO 2 /Ti mesh anode was a kind of dimensionally stable anode (DSA) usually used for electrolytic industry. The electrical conductivity, catalytical activity and stability were rather satisfied.
The generated halogenated intermediates in the photoelectro oxidation treatment. The halogenated intermediates could be formed from the reaction of DOM with halogen. The halogenated intermediates in drinking water were largely investigated, but in the landfill leachate were seldom concerned. Due to the large amount of active chlorine generated in the PEO and EO process and the high concentration of COD in the LLMC, the major kinds of intermediates, including the trihalomethanes (THMs: CHCl 3 , CHCl 2 Br, CHClBr 2 and CHBr 3 ), haloacetic acids (HAAs: C 2 H 3 O 2 Cl-MCAA, CH 2 O 2 Cl 2 -DCAA, C 2 HO 2 Cl 3 -TCAA, CH 2 O 2 BrCl-DBCAA, C 2 H 2 O 2 Br-MBAA, CH 2 O 2 Br 2 -DBAA and C 2 HO 2 Br 3 -TBAA) and haloacetonitriles (HANs: CHCl 2 CN) were measured in the PEO and EO process (Fig. 7).
Generally speaking, both in the PEO and EO treatment, the concentrations of the formed intermediates with more chlorine were higher than those with more bromine, due to the generation of the active chlorine. The result was similar to a previous study 25 . In the PEO treatment, regarding the THMs, the concentration of CHCl 3 increased to 10.5 mg/L in 60 min and decreased nearly to 0 mg/L in 120 min. As to the HAAs, the concentrations of MCAA, DCAA and TCAA increased to 8.0-13.0 mg/L in the first 60-120 min and decreased nearly to 0 mg/L in 180 min. Similar to the THMs, almost no HAAs with bromine was formed during the process. For HAN, only CHCl 2 CN was detectable. The highest concentration achieved in the first 30 min (0.2 mg/L), was lower than those of THMs and HAAs without bromine. In 180 min, the concentration of CHCl 2 CN decreased to 0 mg/L. The formation and degradation trends of the halogenated intermediates were different in the EO process. Regarding the THMs, the concentrations of CHCl 3 increased to 19.5 mg/L in 120 min, and decreased to 11.8 mg/L in 180 min. As to the HAAs, the concentrations of MCAA, DCAA and TCAA increased to 7.2-12.5 mg/L in the first 60 min, and almost keep steady to 180 min. The concentration of CHCl 2 CN (HAN) was 2.4 mg/L at 30 min, and decreased to 0 mg/L in 180 min. Therefore, intermediates with chlorine and without bromine could be formed during the PEO and EO treatment. The formed chlorinated intermediates could be eliminated in 180 min in the PEO treatment. However, most of the formed chlorinated intermediates still existed in 180 min in the EO treatment, indicating that the PEO treatment was more effective for the elimination of the chlorinated intermediates than the EO treatment.
It should also be noticed that before the treatment of the PEO and EO, some halogenated compounds existed with concentrations ranging from 0.1 to 5.0 mg/L, indicating that the some halogenated compounds was formed in the NaClO enhanced Fe 2+ coagulation treatment. After the PEO treatment, the halogenated intermediates formed both in the NaClO enhanced Fe 2+ coagulation and PEO treatment were eliminated. Therefore, from the aspect of the halogenated intermediates, the NaClO enhanced Fe 2+ coagulation treatment combined with the PEO treatment was suitable for the treatment of the LLMCs.
In conclusion, the most suitable technique for treatment of the LLMC was NaClO (60 mmol/L) enhanced Fe 2+ (90 mmol/L) coagulation followed with the PEO treatment (RuO 2 /Ti mesh anode and Ti mesh cathode) for 3 hours at a current density of 400 A/m 2 . The final removal efficiencies of COD, UV 254 , NH 4 + -N and color degree in the LLMC were 86%, 95%, 93% and 95%. All the formed halogenated compounds were eliminated.

Materials. The LLMC and EMBR samples were collected from a sanitary landfill in Chaoyang district in
Beijing. The landfill was built in 2002 and disposed 3400 tons of municipal solid waste per day. The landfill leachate was treated by two-state upflow anaerobic sludge blanket and the MBR followed with nanofiltration membrane (Fig. 8). The EMBR was collected for characteristic comparison with the LLMC. All the samples were stored  Analytical method. The basic water quality index was measured based on the water and wastewater detection and analysis methods 27 . COD was measured by heating digestion colorimetry method using a HACH DRB200 coupled with a DR2800 attachment. UV 254 was measured by an ultraviolet spectrophotometry. NH 4 + -N was tested using the sodium reagent method. Color degree was the value of absorbance at λ max = 436 nm measured with an ultraviolet spectrophotometry. Cl − was measured using ion chromatography. Active chlorine was measured using the N, N-two ethyl benzene two amines spectrophotometric method.
The dissolved organic matter (DOM) was separated based on a previous method 28 . After the samples filtered with 0.45 μm membranes, XAD-8 resin was used for the separation of the hydrophobic bases (HOB), hydrophobic acids (HOA), hydrophobic neutrals (HON) and hydrophilic matter (HiM).
The halogenated intermediates were extracted using hexane for the trihalomethanes (THMs), and using methyl tert-butyl ether for the haloacetic acids (HAAs) and haloacetonitriles (HANs). All these compounds were measured using a GC-ECD. A HP-5 fused silica capillary column (30 m × 0.32 mm × 0.25 μm) was used. Samples (1 μL) were injected in splitless mode. The carrier gas was high purity nitrogen. For the detection of THMs, the velocity of nitrogen was 6.3 mL/min. The injector and detector temperatures were 200 °C and 290 °C, respectively. The initial temperature was set at 35 °C (held for 4 min) and to 260 °C at a rate of 10 °C/min (held for 4 min). For the detection of the HAAs and HANs, the velocity of nitrogen was 2.0 mL/min. The injector and detector temperatures were both set as 250 °C. The temperature program procedure was: 35 °C (held for 4 min) to 65 °C at a rate of 2 °C/min.