Effects of electrochemical intervention on the remediation of black-odorous water: insights into microbial community dynamics and functional shifts in sediments

Black-odorous water refers to water bodies with black or blackish color and terrible smells, which is an extreme manifestation of water pollution caused by urbanization (Chen et al. 2022). This phenomenon has been reported worldwide both in developing and developed countries, such as the St. Johns River in the United States, the Kanda River and the Shakujii River in Japan, the Siak River in Indonesia, and the Nanfei River in China (Cao et al. 2020). According to a previous report in September 2019, there are 2,100 black-odorous water bodies in China, of which 71% are distributed in relatively economically developed and densely populated cities in southeast China (Cao et al. 2020). The black-odorous phenomenon not only threatens the aquatic ecology but also seriously affects the daily life of the people. Therefore, it has aroused considerable interest among government agencies and academic researchers.

Black-odor is generally observed in extremely contaminated water bodies. The oxidation–reduction potential (ORP) is considered a key factor leading to the occurrence of the black-odorous phenomenon (Chen et al. 2020). A large number of pollutants enter water bodies, resulting in a rapid decrease in dissolved oxygen (DO) and a concomitant decrease in ORP (Liu et al. 2015). When the ORP in the sediments is lower than −50 mV, anaerobic microorganisms, such as sulfate-reducing bacteria (SRB), methanogens, fungi, and actinomycetes, dominate the microbial community structure (Chen et al. 2020). In particular, SRBs are capable of reducing sulfate to and H₂S under anaerobic conditions, both of which are critical compounds for black-odor formation (Wang et al. 2014; Liu et al. 2015; Song et al. 2017). Extensive research has shown that plays an important role in the water blackening process (Wang et al. 2014; Song et al. 2017). The prevalent odors such as NH₃, H₂S, and volatile organic sulfur compounds (VOSCs) are produced during the anaerobic decomposition of sulfur-containing organics (Song et al. 2017). Moreover, odorous substances such as geosmin and 2-methylisoborneo-l (2-MIB) are produced during the metabolism in fungi and actinomycetes, which may aggravate the odor in water (Watson 2003). In addition, dimethyl trisulfide (DMTS) and β-ionone released during the algal decomposition process are the sources of odor in algae-induced black-odorous water (Wang et al. 2014).

Over the past century, national and international researchers have conducted several research studies on the treatment of black-odorous water (Cao et al. 2020). Traditionally, black-odorous water treatment methods mainly include sediment dredging (Liu et al. 2015), artificial aeration (Pan et al. 2016), chemical methods (Liu et al. 2016), and biological methods (Yuan et al. 2018). Although sediment dredging and artificial aeration can rapidly reduce the pollution loads and increase DO, they are not permanent solutions and can easily lead to the transfer of pollutants from sediments to overlying water. The chemical methods can rapidly improve water transparency and eliminate odors, but the additional treatment techniques for residual sludge can increase economic costs. Even though the studies have demonstrated the ability of some microorganisms to degrade pollutants efficiently, this ability is governed by environmental conditions and have certain ecological risks. Recently, more attention has been paid to electrochemical water treatment technology for its extraordinary features and advantages. The long-term performance of any treatment for the removal of the black-odor depends on the growth and activities of several functional microorganisms. Therefore, more attention should be paid to the microbial community dynamics and functional variations in the treatment systems of black-odorous waterbodies. Electrochemical technology has been applied so far to different types of wastewater treatment, including actual textile wastewater, industrial wastewater, and toilet sewage (Zhang et al. 2019). However, few studies have investigated the performance of electrochemical technology in the treatment of black-odorous water. Moreover, little is known about the microbial community dynamics and functional variations during the remediation of black-odorous water with electrochemical intervention. In addition, there is a lack of clarity on the underlying mechanisms of electrochemical intervention affecting the biochemical remediation of black-odorous water.

In this study, DC voltages (2.5, 5, and 10 V) were used to increase the ORP of the surface sediments to accelerate the in situ remediation process of black-odorous water. The purpose of this study was: (i) to examine the effects of voltage intervention on the physicochemical properties and gas emissions in black-odorous water; (ii) to analyze the effects of voltage intervention on the microbial community structures and their predicted functions in surface sediments; (iii) to elucidate the underlying mechanisms of the effects of voltage intervention on water quality and odors during the remediation process. The focus of this study was to investigate the performance of electrochemical technology during the remediation of black-odorous water and provide fundamental support for bioelectrochemical water treatment technologies.

Methionine was selected to induce the black-odorous water formation in this study, as it contains C, N, and S and has the potential to cause a black-odor phenomenon (Lu et al. 2012). In total, 645.66 mg of methionine was added in each reactor, such that the concentration of methionine in each reactor was 117.44 mg/L. The water quality parameters and gas emission rates were continuously monitored after the addition of methionine (Supplementary material, Figure S1). Based on the classifications and standards for urban black-odorous water bodies in China (Supplementary material, Table S2), the experiment was started and recorded as day 0 when the water bodies exhibited an evident black-odor (DO < 2 mg/L, ORP < 50 mV, -N > 8 mg/L) (Cao et al. 2020).

The experiment reactors were divided into four groups, namely control, L, M, and H, as shown in Figure 1(b), wherein the improvement in the natural water quality of black-odorous water by self-purification was monitored in the control group. The DC voltages of 2.5, 5, and 10 V were applied to the L, M, and H groups, respectively, to monitor the effects of different voltage interventions on the remediation of black-odorous water. Three parallel devices were set up for each experimental group. The voltage-connected reactors were powered using an adjustable DC voltage regulator (MS-305D, Maisheng, China) and copper wires (1.5 mm²). Graphite electrodes (4 cm × 4 cm × 1 mm) were used for the reactors, with the cathode placed on the water surface and the anode placed on the sediment surface. The voltages were applied from 20:00 to 8:00 (the next day) for 12 h daily, which is determined based on the average light hours of natural water bodies.

The overlying water samples and headspace gas samples were collected every 2 days during the experiment. The water samples were collected from the middle zone of the water column using a 5-mL pipette gun, while the gas samples were obtained with a 10-mL plastic syringe after sealing the reactor for an hour. The sediment and interstitial water samples were collected only on days 0, 10, 20, and 30 to avoid disturbing the sediment-water stratification frequently. All the samples were collected in the power-off state. Furthermore, the overlying and interstitial water samples were filtered using a 0.45 μm membrane for subsequent analysis. The freeze-dried surface sediment samples were sent to the Biomarker Technologies Corporation, Beijing, China for high-throughput sequencing on Illumina Novaseq 6000 (Illumina, Santiago, CA, USA).

The physicochemical characteristics observed in this study included pH, electrical conductivity (EC), DO, ORP of overlying water (ORP-W), ORP of surface sediment (ORP-S), ammonia nitrogen concentration (⁠-N), turbidity, chroma, interstitial sulfate concentration (⁠⁠) and interstitial nitrate concentration (⁠⁠). EC, pH, DO, and turbidity of the water column were measured in situ by a multiparameter water quality sonde (YSI 6600, Yellow Spring, USA). ORP was measured using the Unisense Microsensor Monometer (Version 1.0, Unisense, Denmark). The -N concentration of the overlying water was determined by a Multiplate Reader (M2e, Molecular Devices, USA), while chroma was measured using a spectrophotometer (UV-5100, Yuanxi Instrument, China) (Gollnisch et al. 2021). The and concentrations in interstitial water samples were determined using ion chromatography (ICS-2000, Dionex, USA). The concentrations of H₂S and NH₃ gas were measured using portable gas analyzers (GT903-F-H₂S/GT903-F-NH₃, Kerno, China). The CH₄ and CO₂ gas concentrations were analyzed using gas chromatography (Trace GC, Themrmo Fisher, USA).

The experimental period was 30 days, during which most indicators, including pH, EC, DO, chroma, turbidity, ⁠, ⁠, H₂S, NH₃, CH₄ and CO₂, were observed. However, on the 25th day, the Unisense Microsensor Monometer measuring ORP encountered some uncontrollable accidents, rendering it unusable. Consequently, ORP observations were only carried out up to day 25. Moreover, since only a small amount of -N was degraded by the day 30, the observation period for this indicator was extended to 50 days to investigate the effects of voltage interventions on -N degradation. In addition, due to time constraints, the physicochemical analyses and sample collections could not be completed within a single day. Hence, the observation times of the physicochemical parameters and gas emission rates were not entirely consistent.

Three parallel samples of the same group at the same sampling time were mixed to account for sediment heterogeneity. DNA was extracted with the TGuide S96 Magnetic Soil/Stool DNA Kit (DP812, Tiangen, Beijing, China) according to the manufacturer's instructions. The concentration of the extracted DNA was tested using NanoDrop and the purity was determined using agarose electrophoresis at a concentration of 1.8%. The 515F (5′-GTGYCAGCMGCCGCGGTAA-3′) and 926R (5′-CCGYCAATTYMTTTRAGTTT-3′) universal primer set was used for the amplification of the V4–V5 hypervariable regions of 16S rRNA gene. The PCR amplicons were purified using VAHTSTM DNA Clean Beads (N411-03, Vazyme, Nanjing, China) and quantified by the ImageJ software. After the individual quantification step, amplicons were pooled in equal amounts. For the constructed library, Illumina Novaseq 6000 (Illumina, Santiago, CA, USA) was used for sequencing. The raw amplicon sequences were submitted to National Center for Biotechnology Information (NCBI) (BioProject ID: PRJNA901879; BioSamples ID: SAMN31736351–SAMN31736366).

The bioinformatics analysis in this study was performed using the BMK Cloud (www.biocloud.net). The raw data obtained were primarily filtered by Trimmomatic (version 0.33). The identification and removal of the primer sequences were processed by Cutadapt (version 1.9.1). The clean reads obtained from the previous steps were assembled by USEARCH (version 10). The dada2 method in QIIME2 software was used for denoising and removal of chimera sequences for obtaining amplicon sequence variants (ASVs) (Callahan et al. 2017). Taxonomy annotation of the ASVs was based on the Naive Bayes classifier in QIIME2 using the SILVA database. The microbial community structures at phylum and genus levels were represented by plotting stacked histograms using OriginPro. The distribution of dominant genera across four groups was analyzed using the Cytoscape network analysis. The alpha diversity indices (Chao1, Ace, Shannon index, and Simpson index) were calculated by QIIME2. The beta diversity was determined for evaluating the similarity of microbial communities in different samples. Principal component analysis (PCA) was performed to analyze the beta diversity by Chiplot (www.chiplot.online). The microbial functional profiles were predicted using FAPROTAX. The functional clustering heatmap was constructed in R using the package ‘pheatmap’. The redundancy analysis (RDA) was carried out in R using the package ‘vegan’ to investigate the correlation of predicted microbial functions with the physicochemical properties.

The DO of the water column in the control group started increasing on the 12th day and reached the maximum value of 3.25 mg·L⁻¹ on the 20th day. Subsequently, it exhibited a sharp decline and decreased to 0.81 mg·L⁻¹ on the 28th day (Figure 2(c)). A similar trend was observed for the DO in the L group, with a maximum value of 2.29 mg·L⁻¹ on the 24th day. In contrast, the DO in the M and H groups indicated a delayed but consistent increase on the 16th day, reaching the maximum values of 2.83 and 1.96 mg·L⁻¹ at the end of the experiment, respectively. The ORP-W increased from −129.38 ± 24.30 to +440.09 ± 29.23 mV, indicating that the oxidation conditions in the water column were improved. The maximum value of ORP-W was observed in the L group, followed by the control, H, and M groups (Figure 2(d)). The ORP-S in all the groups exhibited a fluctuating increase from −154.18 ± 16.94 to +157.11 ± 21.62 mV (Figure 2(e)). Moreover, the ORP-S values exhibited significant inter-group differences (ANOVA: p < 0.05), with higher values observed in voltage treatments compared to the control group (mean value: L > M > H > control), indicating that voltage interventions are effective in improving the oxidation conditions of the surface sediment.

An increasing trend was observed for -N concentrations in all the groups during the early stages of the experiment, reaching the maximum values of 30.03, 31.29, 36.09, and 28.69 mg·L⁻¹ in the control, L, M, and H groups, respectively. However, the concentration in the control group decreased sharply from 30.03 to 1.02 mg·L⁻¹ in the later stages of the experiments, whereas it showed a slowly declining trend in the M and H groups. -N concentrations in the M and H groups decreased to 16.44 and 17.5 mg·L⁻¹, respectively on the 50th day, which was still much higher than that in the control (Figure 2(f)).

The chroma in the control group increased to the maximum value (31.15) on the 7th day and then decreased to 5.22 on the 11th day. The chroma in the L group exhibited a similar trend, with a maximum value (42.85) slightly higher than that in the control. However, the chroma in the M group varied within a small range (0.75–11.58). The chroma in the H group rapidly increased to 62.93 and then gradually declined to 2.88 by the end of the experiment (Figure 2(g)). The turbidity of the water column in the control and L groups showed extreme values (57 and 31.18, respectively) around the 12th day. Except for these values, the turbidity in the M and H groups was significantly higher than that in the control (Figure 2(h)). The turbidity and chroma are used to describe the appearance of water, both of which were significantly higher in the H group.

The interstitial concentration in the control group exhibited a slight decrease from 0.15 to 0.13 mg·L⁻¹ during the early stage and consistently increased to 0.49 mg·L⁻¹ with time during the later stage (Figure 4(b)). The interstitial concentration in the L group demonstrated a trend similar to that of the control. However, the interstitial concentrations in the M and H groups at the end of the experiment (0.17 and 0.20 mg·L⁻¹, respectively) were significantly lower than that in the control (0.49 mg·L⁻¹). Nitrates primarily originate from nitrification, an aerobic microbial process that involves the sequential oxidation of -N to -N by ammonia-oxidizing bacteria (AOB) and -N to -N by nitrite-oxidizing bacteria (NOB). On the 20th day of the experiment, the higher DO levels in the control (3.25 mg/L) and L (1.87 mg/L) groups facilitated nitrification, the process by which -N is oxidized to -N or -N while oxygen is consumed. In contrast, the lower DO levels in the M (0.48 mg/L) and H (0.65 mg/L) groups hindered nitrification, resulting in a negligible increase in concentrations. Moreover, nitrates can serve as electron acceptors for microbial oxidation of organic pollutants under anaerobic conditions, a process known as nitrate reduction. The lower DO levels in the M and H groups favored nitrate reduction, which also contributed to the lower concentrations observed in these groups.

A total of 864,701 effective sequences were obtained from 16 sediment samples using Illumina sequencing of 16S rRNA gene amplicon, with at least 47,401 effective sequences obtained in a single sample. A total of 5,603 ASVs were obtained by rarefication of the minimum number of effective sequences (47,401), which contained 30 bacterial and five archaeal phyla.

The species richness and microbial community diversity of each group increased with time during the remediation process of black-odorous water (Supplementary material, Figure S2). The highest ACE, Chao1, and Shannon indices were observed in the L group, followed by the control, M, and H groups. This indicated that low-voltage intervention might lead to an increase in the species richness and microbial community diversity, while high-voltage might lead to a decrease in these parameters. Furthermore, the change in the Shannon index was more significant than that in the Simpson index because the Shannon index gives greater weightage to the low-abundance species and is more susceptible to them (Dušek & Popelková 2017). Thus, it can be stated that the voltage interventions affected species richness and microbial community diversity primarily by affecting the low-abundance species and has little effect on the survival of the most abundant indigenous microorganisms.

PCA was performed using ASVs from 16 samples, which were divided into four groups according to different voltage treatments. The distribution of points in the control and L groups was more concentrated, while the points in the M and H groups were more scattered and different from the control (Supplementary material, Figure S3), indicating significant effects of voltage intervention on the microbial community at higher voltages.

The major phyla in the sediments were Proteobacteria (24–52%), Firmicutes (18–42%), Desulfobacterota (10–15%), Bacteroidota (2–6%), Verrucomicrobiota (1–9%), and Halobacterota (1–6%) (Figure 4(a)). Proteobacteria were the most dominant phylum in almost all the samples. The average relative abundance of Proteobacteria in the control group was 36.8%, while those in the voltage treatments were 33.6, 41.4, and 44.1% for L, M, and H groups, respectively. In contrast, the average relative abundance of Desulfobacterota in the control group was 12.1%, slightly higher than that in the H group (10.6%). Furthermore, the relative abundance of Halobacterota was slightly lower in the M (3.2%) and H (1.7%) voltage treatments than that in the control (4.3%). Overall, the microbial community structures at the phylum level indicated that there was an increase in the relative abundance of Proteobacteria and a decrease in those of Desulfobacterota and Halobacterota at a voltage of 10 V.

At the genus level, 605 genera were identified in 16 samples. Dechloromonas, Desulfovirga, and Symbiobacterium were the top three abundant genera, followed by Citrifermentans, Azospira, Azospirillum, Sedimentibacter, Zoogloea, Methobanolus, Lysobacter, and Thiobacillus (Supplementary material, Figure S4). The clustering network analysis was performed by Cytoscape to elucidate the differences among the four groups. Dechloromonas was the most abundant genus, accounting for a larger proportion in the M and H groups. Similarly, the genera Azospira, Azospirillum, and Pannonibacter were relatively more abundant in the M and H treatments (Figure 4(b)). On the contrary, the relative abundances of Desulfovirga, Symbiobacterium, Methanolobus, and Thiobacillus in the M and H groups were lower than those in the control. In addition, some species with lower abundance also demonstrated significant differences. Prosthecobacter was unique in the control group. The relative abundance of Zoogloea in the L group was significantly higher than that in the other three groups. The highest abundance of Methylotenera was observed in the M group, while Comamonas, Clostridium-sensu-stricto-7, and Cloacibacterium were predominant in the H group.

FAPROTAX analysis was performed to investigate the effects of voltage intervention on the potential microbial functions in the sediments. The functional processes that dominated the sediments included chemoheterotrophy, nitrate reduction, sulfate reduction, nitrogen fixation, fermentation, ureolysis, dark iron-oxidation, dark oxidation of sulfur compounds, nitrite reduction, chitinolysis, methanogenesis, and phototrophy (Figure 6(a)). It was predicted that sulfate reduction, dark oxidation of sulfur compounds, and phototrophy functions were relatively more abundant in the control group, while chemoheterotrophy, nitrate reduction, nitrogen fixation, ureolysis, and dark iron-oxidation were predominant in the M and H groups (Figure 6(b)). Moreover, the relative abundance of methanogenesis in the H group was the lowest among the four groups. Furthermore, the PCA was also applied to analyze the similarity of the microbial functions among different samples, using the number of reads for each predicted function (Supplementary material, Figure S5). The results indicated that the distribution of points in the M and H groups were more concentrated, while the points in the L group were more dispersed, which demonstrated that the voltage interventions of 5 and 10 V might enhance the selectivity of the environment, thereby leading to concentrated microbial functions.

Correlation analysis was performed to analyze the relationship among all the physicochemical properties (Figure 7(a)). pH showed a significant negative correlation with emission rate of CO₂ (r = −0.74, p ≤ 0.01). ORP-W indicated significant positive correlations with DO (r = +0.86, p ≤ 0.001) and ORP-S (r = +0.74, p ≤ 0.01), while negative correlations with the emission rates of CH₄ (r = −0.92, p ≤ 0.001), H₂S (r = −0.74, p ≤ 0.01), and NH₃ (r = −0.85, p ≤ 0.001). This can be explained by the fact that all these gases were mainly produced under anaerobic conditions. However, sulfate concentrations were not significantly correlated with H₂S emission rates (r = −0.37, p = 0.16), indicating that H₂S was emitted not only from the sulfate reduction process but also from the anaerobic degradation of sulfur-containing proteins, such as methionine.

The relationships between the physicochemical properties and the relative abundances of predicted microbial functions were analyzed by RDA (Figure 7(b)). The emission rate of H₂S was positively correlated with the relative abundances of aromatic compound degradation, sulfate reduction, and sulfite reduction functions, while negatively correlated with sulfate concentration and the relative abundance of dark oxidation of sulfur compounds. These relationships corroborated that the H₂S emission in this study were mainly from the anaerobic degradation of sulfur-containing organics and sulfate reduction. Moreover, the emission rate of CH₄ was positively correlated with methanogenesis, while negatively correlated with methanotrophy and phototrophy. This was because enhanced phototrophy function was generally accompanied by increased oxygen production and ORP, which would lead to the inhibition of methanogenic activities under anaerobic conditions. Furthermore, the emission rate of CO₂ was positively correlated with chemoheterotrophy, indicating that CO₂ emission might be partly enhanced due to the activities and respiration of chemoheterotrophic bacteria in the sediments. In addition, the nitrate concentration demonstrated a negative correlation with the relative abundance of nitrate reduction, which explained that the lower nitrate concentrations in the M and H groups were partially due to the enhanced nitrate reduction functions. Generally, it can be stated that the voltage interventions significantly affected the potential microbial functions involved in S, C, and N element cycles, which further affected the physicochemical environment and gas emissions in black-odorous water.

The results indicated that the voltage interventions significantly affected the physicochemical properties of black-odorous water. Firstly, the oxidative conditions in the surface sediments were effectively improved upon application of the voltages. The anode electrode placed on the sediment surface can aid in the oxidation of reductive gases and pollutants, thereby improving ORP-S (Moreira et al. 2017). Meanwhile, the electric field induces the discharge of water to produce hydroxyl radicals with strong oxidation capacity, which can oxidize reductive substances such as H₂S, NH₃, CH₄, and organic pollutants (Panizza & Cerisola 2009). The ORP-S values are influenced by both the anode electrode and the redox environment of the overlying water. However, the experimental results suggest that the voltage groups did not exhibit a significant increase in DO or ORP-W levels, and in fact, they were lower than those in the control group. This finding suggests that the electrode could not effectively generate O₂ through water electrolysis. Moreover, the cathode electrode placed on the water surface can lead to the cathodic reduction of oxygen at voltages of 5 V and above, thus affecting the reoxygenation efficiency and leading to delayed improvements in DO and ORP-W (Kaneko et al. 2006). Considering the mutual influence of the sediment and the overlying water, this can further lead to a decrease in ORP-S. Therefore, the improvement effect of voltage interventions on the ORP-S did not increase with higher voltages, as the highest ORP-S value was observed in the L group (Figure 2(e)).

In the voltage treatments, pH continued to decrease with an increase in voltage, partly due to the oxidation of H₂S to more acidic H₂SO₄ or H₂SO₃ near the anode. Meanwhile, H⁺ was also produced during the oxidation of organics by the hydroxyl radicals (Panizza & Cerisola 2009), which was corroborated by the significant negative correlation between pH and CO₂ emission rate (Figure 7(a)). Moreover, the intermittent existence of the electric field was not conductive to the settling of suspended matter, while augmented gas emissions also added hydrodynamic disturbances, ultimately resulting in increased chroma and turbidity of the water column.

Briefly, these results had illustrated that voltage intervention could be effective in increasing ORP-S and inhibiting the emissions of reductive odors such as H₂S, NH₃, and CH₄. However, the water disturbance caused by the electrodes and gas emissions could lead to an increase in the chroma and turbidity of the water column. Meanwhile, the cathodes could lead to delayed improvements in DO and ORP in the overlying water.

The results indicated that the voltage interventions of 5 and 10 V inhibited the emissions of H₂S, NH₃ and CH₄ while significantly enhancing the emission of CO₂ (Figure 3). As mentioned in Section 4.2, NH₃ and H₂S emitted from the reactors mainly came from sulfate reduction and degradation of sulfur-containing organics. It has not yet been confirmed that electron acceptors have an inhibitory effect on the mineralization of organic sulfur compounds. No significant difference was observed in the variations of -N concentrations between the control and voltage treatments during the early experimental period (Figure 2(f)), indicating that voltage intervention had a minimal effect on the decomposition of methionine. However, SRBs are generally considered strict anaerobic microorganisms. Prior studies have shown that the metabolism of sulfate-reducing microorganisms is hindered when the ORP exceeds −100 mV (Zhang et al. 2022). Consequently, the increased ORP-S would impede the activities of SRB, leading to a reduction in H₂S emission rates in the voltage treatments. Desulfovirga is a typical SRB that can reduce ⁠, ⁠, or to H₂S using a variety of organics as electron donors (Davidova et al. 2022). The lowest abundance of Desulfovirga was observed in the H group, which was in agreement with the low abundance of sulfate reduction and sulfite reduction functions in this group (Figure 6(b)). On the contrary, the genera Dechlormonas and Azospira, which were generally considered perchlorate-reducing or nitrate-reducing microorganisms, were significantly predominant in voltage groups (Chakraborty & Picardal 2013). Previous studies have demonstrated that all the known dissimilatory perchlorate-reducing microorganisms are naturally able to oxidize H₂S (Mehta-Kolte et al. 2017). In addition, the electrogenerated hydroxyl radicals can oxidize H₂S and NH₃ in the water column, which also leads to a decline in their emission rates.

The effects of voltage interventions on CH₄ emission were similar to those on H₂S emission. On the one hand, the increase in ORP-S inhibited the methanogenesis process in sediments. Methanolobus and Methanosarcina were the predominant methanogens in the sediments (Supplementary material, Figure S4), both of which can utilize a range of substrates for methanogenic activities even in extreme environment (Gao et al. 2017). The lowest abundances of both Methanolobus and Methanosarcina were observed in the H group (Figure 5(b)), indicating that voltage intervention of 10 V was detrimental to the survival and activities of methanogens. Accordingly, the lowest relative abundance of predicted methanogenesis function was observed in the H group (Figure 6(b)), which was consistent with the lowest CH₄ emission rate observed in this group (Figure 3(c)). Furthermore, CH₄ can be oxidized into CO₂ by hydroxyl radicals, which also leads to a decrease in the CH₄ emission rate.

In this study, the effects of voltage interventions of 2.5, 5, and 10 V on the remediation of black-odorous water were investigated by conducting simulation experiments. The voltage interventions of 5 and 10 V can effectively increase the ORP of the surface sediments and inhibit the emissions of reductive odor gases such as H₂S, NH₃, and CH₄. On the contrary, the emission of CO₂ was significantly enhanced after voltage treatments. Moreover, voltage interventions have a significant effect on the microbial activities in the sediments. The relative abundances of typical methanogens (Methanosarcina and Methanolobus) and SRB (Desulfovirga) decreased while the chemoheterotrophic microorganisms (e.g., Dechloromonas, Azospira, Azospirillum, and Pannonibacter), which play important roles in the anaerobic oxidative decomposition of organics, became more abundant. Meanwhile, voltage application also led to a significant enhancement in the predicted microbial chemoheterotrophy and nitrate reduction functions in the sediments, implying the improved biochemical degradability of sediments.

In summary, the results demonstrated that electrochemical remediation of black-odorous water is feasible and the implementation effect is significant. The oxidation conditions in the sediments were improved and the water odor was effectively eliminated after voltage treatments. The findings of this research provide insights into the electrochemical treatment of black-odorous water. However, further research is still needed to avoid the adverse effects of voltage devices, such as water disturbance and impeded reoxygenation processes.

This work has been jointly supported by the National Natural Science Foundation of China (No. 42077221), the Fundamental Research Funds for the Central Universities (No. 2019B72714), the Postgraduate Research and Practice Innovation Program of Jiangsu Province (No. SJKY19_0472).

Y.S. conceptualized the study, did data curation, did formal analysis, investigated the study, performed methodology, visualized the study, and wrote the original draft. Z.W. did formal analysis, performed methodology, supervised the study, wrote, reviewed, and edited, and did project administration. Y.X. performed the methodology, supervised the study, wrote, reviewed and edited the article, and did project administration. X.L. did formal analysis, supervised the study, wrote, reviewed, and edited the article, and did project administration. A.R. supervised the study, wrote, reviewed, and edited the article, and did project administration.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.