Impact of electrochemically generated iron on the performance of an anaerobic wastewater treatment process

• E ‑ iron dosage of 40 – 200 mg Fe/L was tested in an anaerobic sewage treatment process. • An increase of 79.8 – 98.5 % in dissolved sul ﬁ de reduction was obtained. • Phosphate removal ef ﬁ ciency increased by 26.0 – 96.0 %. • CH 4 production rate increased by 12.0 – 21.3 % while CO 2 content in biogas decreased.


Introduction
In contrast to conventional aerobic treatment process, the anaerobic process has the advantages of low energy demand, low sludge production, Science of the Total Environment 875 (2023) 162628 and superior energy balance through methane (CH 4 ) recovery (Li and Yu, 2016;Wan et al., 2016;Garfí et al., 2017;Liu et al., 2021a;Liu et al., 2021b;Lim et al., 2023;Wang et al., 2022). Despite these benefits, this process still has several inherent issues, i.e., i) sulfate in influent can be reduced by sulfate-reducing bacteria to generate notorious hydrogen sulfide (H 2 S), leading to odour and safety concerns and risk of equipment corrosion (Vollertsen et al., 2008;Yongsiri et al., 2005); ii) as phosphorus cannot be removed, discharging the effluent rich in phosphorus can cause water eutrophication (Carpenter, 2005;Schindler et al., 2016); and iii) biogas normally contains high CO 2 (30-50 %), which significantly reduces the lower calorific value (LCV) of per unit volume biogas and makes the biogas difficult to be compressed (Angelidaki et al., 2018;Lai et al., 2021). Effluent of the anaerobic treatment process contains residual ammonia that can be primarily removed by an autotrophic nitrogen removal process Hausherr et al., 2022;Zheng et al., 2023).
Iron salts are widely used in urban water management system and also favor anaerobic wastewater treatment process (Rebosura et al., 2018a). Dosing of iron salts can form insoluble ferrous sulfide (FeS) and vivianite (Fe 2+ Fe 2 2+ (PO 4 ) 2 ·8H 2 O), results in dissolved sulfide and phosphate reduction, thereby mitigating corrosion, obnoxious odors emission and water eutrophication (Calderon et al., 2021;Hu et al., 2021;Rebosura et al., 2018b;Salehin et al., 2020;Wilfert et al., 2016;Yuan et al., 2020). Nevertheless, iron chemicals currently used in the wastewater industry are mostly manufactured as by-products in metallurgical processes, which are then trucked to wastewater treatment sites and facilities. This includes a long-distance transportation with high costs. Recently, a technology of using electrochemical cell to produce iron salts from iron plates was developed (Lin et al., 2017). The electrochemical cell was built by setting two iron plates in the reactor, where Fe 2+ was produced from sacrificial anodes (Fig. S1). Transportation of solid iron plates is much more cost-saving and enables safe handling, in comparison with the transportation of corrosive concentrated iron salts solution.
The electrochemical iron (e-iron) production and use were implemented and tested for sulfide control in sewers (Lin et al., 2017;Pikaar et al., 2019), and phosphorus control in effluents of wastewater treatment process (Hu et al., 2022). With a dosage of~80 mg Fe/L, the e-iron significantly reduced dissolved sulfide and phosphate in the effluent of anaerobic bioreactor and H 2 S in biogas (Hu et al., 2022). Moreover, e-iron production not only generates hydroxide ion (OH − ) but also H 2 that can be utilized by methanogens to consume CO 2 . Therefore, it is hypothesized that increasing an e-iron dosage can also enable in-situ biogas upgrading with a less CO 2 content. Together, it is possible to achieve simultaneous sulfide, phosphate, and CO 2 control by boosting the e-iron dosage, which requires more tests.
This study aims to perform a comprehensive assessment on the effect of different e-iron dosages on the performance of an anaerobic wastewater treatment process. Following our previous report at 80 mg Fe/L, three more different e-iron dosages (40, 60, and 200 mg Fe/L) were further studied in this work using experimental and control anaerobic bioreactors. During the 210 days of operation, effluent concentrations of total and volatile suspended solids (TSS and VSS), total chemical oxygen demand (TCOD), dissolved sulfur species, and phosphate, biogas production rates and its compositions in two reactors were monitored and compared.

Wastewater characteristics
Every two weeks, the fresh domestic wastewater was collected from a residential pumping station located at St Lucia, Queensland, Australia. After collection, the wastewater was immediately stored in a 1000-L storage tank located in a temperature-controlled (4°C) cold room to minimize biological transformation during storage (Duan et al., 2019). In the storage tank, the wastewater was mixed by a submerged pump (RSCD750, Ryobi) operated intermittently at an on/off ratio of 30/60 min. Prior to usage, the wastewater was warmed up to room temperature (22 ± 1°C) via a heater (IC-TH7100, RATEK). Table 1 summarizes characteristics of the used wastewater in this work.

Reactors set-up, operation, and monitoring
Two identical anaerobic baffled reactors (ABRs) were constructed. The ABRs were made of polymethyl methacrylate with a height of 450 mm and an inner diameter of 200 mm, giving a total volume of 12 L (10 L working volume and 2 L headspace). They were sealed to ensure gas-tightness and covered with aluminum foil to prevent light exposure. Each ABR had six replicate compartments, separated by six baffles (Fig. 1). 2 L of anaerobically digested sludge collected from a local full-scale anaerobic digester (Brisbane, Australia) was equally distributed and inoculated into six compartments. The information of digested sludge was detailed in our previous study (Calderon et al., 2021). Afterwards, 10 L of real domestic wastewater was fed to ABRs by using peristaltic pumps (BT100-2J, Longer, China), which gave a hydraulic retention time (HRT) of 1 d. The effluent was controlled by using peristaltic pumps with the same flow rate. The produced biogas was recorded by using a tipping bucket gas meter and collected in a gas bag. To ensure the reliability of gas counter, gas meter was recalibrated monthly.
The experimental reactor was installed with an electrochemical unit for in-situ e-iron production, namely R-ferrous. Another one was operated in parallel without the electrochemical unit, serving as R-control. In Rferrous, two iron plates were placed in the first compartment as anode and cathode of the electrochemical unit, respectively. Each iron plate had a dimension of 40 cm × 3 cm × 1 cm and a weight of about 960 g. Anode and cathode were symmetrically installed with interelectrode gap distance of 1.0 cm. Iron oxidation was achieved by maintaining a predesigned constant current in each phase via a bench power supply (72-2685, TENMA, China) (detailed below).
The two ABRs were operated for 210 days, during which the operating conditions of R-control were unchanged and that of R-ferrous was divided into three phases with varied iron dosages. From day 370 to day 436, a low dosage of 40 mg Fe/L was used by setting the constant current at 16 mA in the electrochemical unit. The dosage was increased to 60 mg and 200 mg from day 437 to day 518 and from day 519 to day 585, by controlling the current at 24 mA and 80 mA, respectively.
Throughout the operation, concentrations of TSS, VSS, TCOD, SCOD, sulfur species (sulfide, sulfite, thiosulfate, and sulfate), NH 4 + -N, and PO 4 3− -P in the influent and effluent were measured 2-3 times/week. The amount of produced biogas was monitored by a gas meter and manually recorded 2-3 times/week. The H 2 S concentration in biogas and the effluent dissolved methane were measured weekly. The contents of CH 4 , CO 2 , and H 2 in the biogas were analysed 1-2 times/week. The pH of influent and effluent was monitored weekly.

Chemical analysis
TSS and VSS concentrations were measured according to the standard methods (APHA, 2005). TCOD and SCOD concentrations were  , 2006). Analysis of total and dissolved iron concentrations were conducted by using an inductive Column Plasma Optical Emission Spectroscopy (Perkin Elmer Optima 7300DV, Waltham, MA, USA). The pH was measured using a portable pH monitor and probe (pH 5+, Oakton). The content of CH 4 , CO 2 , and H 2 in biogas were measured by using the gas chromatography method (7890A, Agilent), as described in literature (Zhang et al., 2019). The dissolved CH 4 concentration (Guisasola et al., 2008) and biogas H 2 S content (Rebosura et al., 2018b) were measured according to the methods detailed in literature.

COD mass balance assessment
To understand the fate of organic carbon in the raw wastewater, COD mass balance assessment was carried out in each phase. The COD in wastewater was consumed by three pathways, i.e., converted to gaseous CH 4 , used by sulfate reducing bacteria, and discharged with the effluent. Further, the dissolved CH 4 part in effluent was shown in the results separately. According to the Henry's Law and stoichiometric number, the consumption of 1 g COD should correspondingly produce 350 mL gaseous CH 4 or reduce 0.5 g dissolved sulfate as S.
Mass balance evaluation was conducted to analyse the conversion of COD to CH 4 (Table S2). The increased CH 4 production was associated with not only the high COD removal efficiency but also additional H 2 formed by the electrochemical unit (i.e., hydrogenotrophic methanogens, CO 2 þ 4H 2 ! CH 4 þ 2H 2 O). The calculation indicated that the increased COD removal of R-ferrous should contribute to more CH 4 production rates, as 67, 248, 193, and 227 mL CH 4 /d at 40, 60, 80, and 200 mg Fe/L, respectively. The more CH 4 production by the hydrogenotrophic process was estimated to be 35 mL CH 4 /d at 40 mg Fe/L, 50 mL CH 4 /d at 60 mg Fe/L, 68 mL CH 4 /d at 80 mg Fe/L, and 74 mL CH 4 /d at 200 mg Fe/L. Combined, the theoretically increased CH 4 production rate should be 102, 298, 261, and 301 mL CH 4 /d at 40, 60, 80, and 200 mg Fe/L. The calculated results were very close to the measured values of 124, 317, 266, and 373 mL CH 4 /d, thus indicating that the increased CH 4 production was due to both enhanced COD removal and H 2 utilization.
The efficiency of in-situ H 2 utilization was 86 %, 83 %, and 84 % at 40, 60, and 80 mg Fe/L, respectively (Table S2). This is generally higher than that (<80 %) of the continuous stirred tank reactor (Díaz et al., 2015), attributed to the ABR design, i.e., the H 2 produced in the first chamber was forced to pass through all chambers before emitting in the last chamber. The efficiency decreased to 37 % at 200 mg Fe/L stage, suggesting that additional gas recirculation is still needed to improve the efficiency.
The COD mass balance analysis also revealed similar flux in each phase of R-control (Fig. S4a), with~60 % of influent COD being converted to CH 4 (~52 % in gas phase and~8 % in the effluent), and~25 % COD (without CH 4 and dissolved sulfide) residuals in the effluent. In contrast, the Rferrous always had~5 % less of residual COD in each stage, resulted from the higher COD removal and more CH 4 production (Fig. S4b).

Effect of e-iron dosage on sulfide and phosphate control
The average dissolved sulfate, sulfide, and phosphate concentrations in the influent and effluent and H 2 S content in the biogas of R-control and Rferrous are shown in Fig. 4. The influent sulfate and sulfide concentrations were 10.5 ± 2.7 mg S/L and 6.6 ± 1.3 mg S/L, respectively. In R-control, the sulfate was almost fully converted to sulfide, showing low sulfate of 0.2 ± 0.1 mg S/L and high sulfide of 20.2 ± 2.7 mg S/L in the effluent (Figs. 4a,b,S5a,and b). This caused the H 2 S content as high as 1083.6 ± 133.1 ppmv in the biogas of R-control. The average effluent sulfide concentration in R-ferrous was 4.0 ± 2.5 mg S/L at 40 mg Fe/L, 1.8 ± 1.3 mg S/L at 60 mg Fe/L, 1.7 ± 0.2 mg S/L at 80 mg Fe/L, and 0.2 ± 0.1 mg S/L at 200 mg Fe/L, giving a relative reduction by 79.8 ± 12.6 %, 89.6 ± 6.7 %, 93.7 ± 7.3 %, and 98.5 ± 0.7 %, respectively. Similarly, the H 2 S content in the R-ferrous biogas declined by 75.5 ± 11.7 %, 84.4 ± 6.3 %, 90.1 ± 6.0 %, and 95.8 ± 1.7 %, respectively (Figs. 4c and S5c). The reduction of both dissolved sulfide in effluent and H 2 S in biogas was statistically significant (p < 0.05). Together, these results indicated that e-iron efficiently removed dissolved and gaseous sulfide, with removal efficiency positively correlated with dosage.
The effluent phosphate concentration in R-control was 7.5 ± 0.9 mg P/ L, slightly higher than the observation of 6.2 ± 1.4 mg P/L in the influent (Fig. 5), due to the hydrolyzation of biomass under anaerobic conditions. The phosphate removal efficiencies were 26.0 ± 20.6 %, 63.3 ± 17.8 %, 77.5 ± 5.2 %, and 96.0 ± 3.2 %, at 40, 60, 80, and 200 mg Fe/L, indicating the effective control of phosphate by e-iron. Fig. 6 shows the contents of CH 4 , H 2 , and CO 2 in the biogas of R-control and R-ferrous at different e-iron dosages. The gas compositions of the biogas produced in R-control were similar in the four phases (p > 0.05), show-ing~75 % CH 4 ,~25 % CO 2 and negligible H 2 . In comparison, lower biogas CO 2 contents were obtained in R-ferrous (Fig. 5b). The biogas CO 2 contents significantly (p < 0.05) decreased with the increase of e-iron dosage, from 20.5 % ± 0.7 % at 40 mg Fe/L to 18.9 % ± 0.5 % at 60 mg Fe/L, 14.2 % ± 0.6 % at 80 mg Fe/L, and 4.5 % ± 0.4 % at 200 mg Fe/L. The CO 2 was removed by two pathways: i) utilizing H 2 produced in the cathode via the methanogenetic reaction of CO 2 þ 4H 2 ! CH 4 þ 2H 2 O, and ii) utilizing OH − generated in the cathode via the reaction of CO 2 þ OH À ! HCO À 3 . At 200 mg Fe/L, the CO 2 production rate of Rferrous was about 480 mL CO 2 /d lower than that of R-control, among which~74 mL CO 2 was reduced to CH 4 by additional H 2 (approximately 15 % of the reduction of the total CO 2 production rate) (Table S2). It suggested that the remaining~85 % of CO 2 could be removed by OH − produced in the cathode. The H 2 content in the biogas increased to >20 % at 200 mg Fe/L (Fig. 6c). The biogas containing low CO 2 and high H 2 /CH 4 means a high energy density (30.5 MJ/m 3 of R-ferrous vs. 28.4 MJ/m 3 of R-control) and has a wide application, e.g., as a fuel in the transportation sector.

Implication for anaerobic wastewater treatment
In this study, e-iron was developed to be an alternative to the conventional iron salts. The e-iron can avoid occupational health and safety (OH&S) concerns caused by transportation and storage of the corrosive and acidic iron salt solution. It can also avoid a decrease of wastewater pH caused by hydrolysis of conventional iron ions . The performance of anaerobic process treating domestic wastewater was comprehensively assessed with four different dosages, and the results showed that the e-iron significantly improved the efficiency of anaerobic wastewater treatment process, including dissolved sulfide and phosphate and gaseous H 2 S and CO 2 reductions and CH 4 production enhancement. The reductions of effluent dissolved sulfide and phosphate over 75 % were obtained at 40 and 80 mg Fe/L, respectively. The CO 2 content in biogas was reduced to below 5 % with a higher dosage of 200 mg Fe/L. The e-iron dosing greatly enhanced the performance of the anaerobic wastewater treatment process, and the effectiveness relied on dosage. Nevertheless, the effluent iron concentrations at different dosages remained similar, as 4.4 ± 2.3 mg Fe/L, 5.8 ± 2.8 mg Fe/L, 6.4 ± 2.9 mg Fe/L, and 10.8 ± 5.2 mg Fe/L at 40, 60, 80, and 200 mg Fe/L, respectively (Table S1). The estimated operating cost of e-iron dosing was 34.6,60.5,92.2,and 576 AUD $/ML at 40,60,80,and 200 mg Fe/L, respectively (see calculation in Table S3).   The dissolved sulfide and phosphorus were mainly removed in the form of FeS and Fe 3 (PO 4 ) 2 ‧8H 2 O, respectively (Fig. S8). The phenomenon is consistent with previous reports using iron salts to remove dissolved sulfide and phosphate in wastewater management systems (Firer et al., 2008;Salehin et al., 2020;Wilfert et al., 2016). The removal efficiency of dissolved sulfide was higher than that of phosphate. More iron precipitated sulfide than phosphorus (Fig. S9) (Chen and Faust, 1974;Davison, 1991).
It is noteworthy that the extremely high e-iron dosage (200 mg Fe/L) was for the first time investigated in this study. The experimental result showed that e-iron dosing enabled superior effectiveness for sulfide and phosphate removal, i.e., dissolved sulfide and phosphate were reduced to 0.1 ± 0.1 mg S/L and 0.2 ± 0.1 mg P/L, respectively ( Fig. 4a and  Fig. 5). The effluent phosphate concentration meets the highest discharge standard in many countries (Li et al., 2012;Piekema and Valsami-Jones, 2004;Schellenberg et al., 2020). This dosing also enabled biogas upgradationthe contents of H 2 S and CO 2 in biogas were decreased to 25.2 ± 17.5 ppmv and 4.5 ± 0.4 %, respectively. The biogas is able to be compressed for value-adding applications (e.g. as a transport fuel), or further upgraded to biomethane for injection into the natural gas network (Angelidaki et al., 2018). This compares to the produced biogas used for thermal and electrical energy generation in current wastewater treatment plants (De Vrieze et al., 2018;Kougias et al., 2017;Szarka et al., 2013). Along with the continuous operation of anaerobic reactors, the sludge was not intentionally discharged. The majority of dosed iron was retained in the sludge, finally causing an iron-rich product (Table S4). The measurement showed that the amount of iron accounted for about 57 % of sludge mass, which is even in a range of iron ore (Holmes et al., 2022), indicating high recovery potential in future studies.
In this study, the performances of R-control and R-ferrous were compared to reveal the impacts of e-iron on the anaerobic domestic wastewater treatment process. Still, the elevated CH 4 production rate in each phase was potentially associated with the increase of activity of hydrogenotrophic methanogens. A previous study has shown that dosing iron can inhibit the activity of sulfate-reducing bacteria and methanogens in sewer systems (Zhang et al., 2009). In this study, the performance of increased biogas production rate was observed during anaerobic treatment of domestic wastewater by dosing e-iron. This result was due to two main reasons: 1) more COD (solid COD in particular) was retained in the reactor due to the coagulative effect of dosed iron, which was subsequently converted to CH 4 by methanogens; and 2) the addition of H 2 gas facilitated the reduction of CO 2 to CH 4 by hydrogenotrophic methanogens. Nevertheless, the shift of microbial community induced by e-iron may also play a role and needs to be investigated in further studies.

Conclusion
This study performed a comprehensive laboratory assessment of the impact of different e-iron dosages on the anaerobic domestic wastewater treatment process. The results of long-term demonstrated that the reduction of sulfide and phosphorus concentrations in liquid and the decrease of H 2 S and CO 2 contents in biogas were positively correlated with dosage. E-iron preferred to remove dissolved sulfide before removing phosphate. Also, the dissolved sulfide and phosphate were mainly removed as FeS and Fe 3 (PO 4 ) 2 ‧8H 2 O. In addition, the biogas quality was obviously improved by dosing e-iron at an extremely high dosage of 200 mg Fe/L. These results collectively demonstrated that dosing of e-iron greatly improved performance of anaerobic wastewater treatment process, while the specific design for enhanced sulfide, phosphorus and CO 2 control depends on dosage.