Application of Hydrogen-Based Denitrification: Simultaneous Removal of Nitrate and Reactive Black 5 Dye from Textile Wastewater Containing Organic Matter

Abstract

In this study, a hydrogen-based denitrification (HD) reactor was used to investigate the simultaneous treatment of nitrogen and decolorization in textile wastewater contaminated with organic matter. The reactor operated in two phases: without and with organic matter. Despite the short hydraulic retention time, the HD system successfully removed all pollutants, including nitrate, nitrite, reactive black-5 dye and chemical oxygen demand. The unhindered treatment efficiency for nitrogen and decolorization in the presence of organic pollutants was observed. With the addition of organic matter, the nitrogen removal efficiency increased slightly from 85% to 90–100%, and the decolorization rate doubled from 25% to 50–60%. Organic matter played a crucial role in stimulating heterotrophic bacteria during biological denitrification and acted as a carbon source facilitating biological denitrification and azo bond cleavage during dye degradation. Despite the generation of toxic byproducts and changes in the dominant microbial community, the treatment efficiency remained stable and improved. This approach offers a promising solution for enhancing treatment efficiency in textile wastewater, providing a cost-effective and environmentally friendly option for developing countries to treat wastewater before discharge.

1. Introduction

Textile wastewater (TW) constitutes a significant source of wastewater, posing severe environmental and health risks. This type of wastewater is typically generated in large quantities and contains highly toxic pollutants. Previous studies have estimated that approximately 1000–3000 m³ of TW, weighing between 12 and 20 tons daily, is discharged into natural sources after textile processing [1]. However, the characteristics of TW vary depending on factors such as the type of fabric (e.g., cotton, linen, wool, synthetic) and the color used to enhance fabric qualities [2]. Furthermore, various pollutants are released at different stages of textile processing, including sizing, scouring, bleaching, mercerizing, dyeing, printing and finishing [3]. Traditional wastewater treatment processes fail to eliminate nearly 30 toxic compounds found in dyes and chemicals, leading to the accumulation of harmful substances in TW and their subsequent discharge into the environment [4].

The primary pollutants commonly detected in effluent from the textile industry include suspended solids (SS), nitrate (NO₃-N) and nitrite (NO₂-N) as forms of nitrogen, organic pollutants represented by chemical oxygen demand (COD) and biochemical oxygen demand (BOD), dye color and other soluble substances, as indicated in Table 1 [2,5]. Generally, biological, chemical and physical treatment methods have been employed as standard practices to enhance TW quality [6]. As demonstrated in Table 1, it is evident that the influent of textile wastewater exhibits significantly high concentrations, and despite undergoing treatment processes, its quality remains poor. This observation highlights the need for alternative treatment processes to effectively address textile wastewater, ensuring proper treatment and compliance with discharge limits. However, the combination of anaerobic and aerobic conditions in biological treatment has been frequently utilized due to its cost-effectiveness, simplicity and efficacy in dye removal. Nevertheless, conventional biological processes are unable to fully degrade certain nitrogen forms, such as NO₃-N, NO₂-N and toxic dye molecules [7]. For instance, reports indicate that effluent TW can contain approximately 40–100 g-NO₃-N/L discharged from sodium nitrate (NaNO₃) used to enhance textile fibers in dyeing baths, with around 50% of the influent dye often contaminating the discharged wastewater [8]. Consequently, tertiary treatment processes such as adsorption, membrane filtration, ozonation, photocatalytic degradation and electrochemical processes are required after biological treatment, albeit at high costs [7,9,10,11,12,13].

Conversely, Singhopon et al. [13,14] discovered that a simple and cost-effective system known as the hydrogen-based denitrification (HD) system can simultaneously remove NO₃-N, NO₂-N and certain dyes. HD, or hydrogenotrophic denitrification, is an autotrophic denitrification method that utilizes H₂ gas as an electron donor and NO₃-N and NO₂-N as electron acceptors, and HCO₃⁻ or CO₂ serves as an inorganic carbon source [7]. The process offers several advantages, such as a low hydrogen cost, minimal biomass accumulation, the generation of clean and non-toxic waste, the elimination of the requirement for a carbon source, a high nitrogen removal rate compared to that of heterotrophic denitrification as well as easy installation and maintenance [15]. A reliable reason for the claim can be found in the study conducted by Singhopon et al. [13,14]. Their research demonstrated that a hydrogen-based denitrification (HD) system, which is a simple and cost-effective approach, has the ability to simultaneously remove NO₃-N, NO₂-N and certain dyes.

The HD system, also known as hydrogenotrophic denitrification, operates by utilizing H₂ gas as an electron donor and NO₃-N and NO₂-N as electron acceptors. In this process, HCO₃⁻ or CO₂ acts as an inorganic carbon source [7]. This denitrification method offers several advantages, one of which is the low cost associated with hydrogen usage. By utilizing hydrogen gas, the HD system provides an economically feasible option for denitrification and simultaneous dye removal. The HD system has been widely employed for NO₃-N and NO₂-N reduction in drinking water treatment, particularly in developing countries [16,17,18]. Various HD systems, such as suspended growth, fixed-bed and sequencing batch reactor (SBR)-membrane bioreactors, have been developed, exhibiting high nitrogen removal rates [7,19,20]. However, some studies have also applied this system to reduce nitrogen in domestic wastewater. For example, Li et al. [21] established an HD system using a membrane diffusion packed-bed bioreactor for tertiary nitrogen removal from municipal wastewater, demonstrating relatively high nitrogen removal efficiency (NRE). Similarly, Visvanathan et al. [22] observed a high nitrogen removal rate when combining the HD system with a membrane bioreactor for NO₃-N removal in synthetic aquaculture wastewater intended for recirculation purposes. Additionally, their results confirmed that certain organic pollutants, such as low COD concentrations, experience a decrease of 70–90% without affecting the nitrogen removal rate in a simple HD reactor [23]. Nevertheless, the HD system is not commonly employed in industrial wastewater treatment due to the presence of various pollutants and toxic chemicals.

Hence, the present study aimed to evaluate the treatment performance of nitrogen and decolorization in synthetic TW by adapting the HD system based on previous research conducted by Singhopon et al. [13,14]. The study objectives involved investigating the nitrogen removal efficiency (NRE) and decolorization efficiency (DRE) of synthetic textile wastewater by introducing organic matter (OM) at different COD concentrations, a factor that has not been extensively explored. An HD reactor was operated over a period of approximately 80 days, with and without organic matter contamination, to assess the treatment process and identify potential methods for enhancing the system’s efficiency.

Table 1. Summary of textile wastewater characteristics, for which the standards of industrial wastewater prior to discharge of different countries are compared.

Parameters	Industrial Wastewater Discharge Standards			Influent		Effluent
Reference	[13]	[24]	[24]	[13]	[1]	[13]	[2]
pH	5.5–9	6–9	5.5–9	7.0	6–10	7.77	7.7–8.0
BOD (mg/L)	<60	60	30	800–1200	100–4000	1050	272–310
COD (mg/L)	≤400	200	250	6000–8000	150–10,000	3056	678–932
TN (mg/L)	≤100	-	-	100–700	-	154	-
TSS (mg/L)	≤150	150	100	1500–3000	100–5000	975	120–180
TDS (mg/L)	≤3000	-	2100	8725	1800–6000	4000–9000	1632–1902
Cr3+ (mg/L)	≤0.5	-	100	-	-	13	-
Color (ADMI)	-	800–1200	-	-	-	-
Color (mg/L)	-	-	-	40–500	-	-	-
Color (Pt-Co)	-	80	-	-	50–2500	-	240–290

Table 1. Summary of textile wastewater characteristics, for which the standards of industrial wastewater prior to discharge of different countries are compared.

Parameters	Industrial Wastewater Discharge Standards			Influent		Effluent
Reference	[13]	[24]	[24]	[13]	[1]	[13]	[2]
pH	5.5–9	6–9	5.5–9	7.0	6–10	7.77	7.7–8.0
BOD (mg/L)	<60	60	30	800–1200	100–4000	1050	272–310
COD (mg/L)	≤400	200	250	6000–8000	150–10,000	3056	678–932
TN (mg/L)	≤100	-	-	100–700	-	154	-
TSS (mg/L)	≤150	150	100	1500–3000	100–5000	975	120–180
TDS (mg/L)	≤3000	-	2100	8725	1800–6000	4000–9000	1632–1902
Cr3+ (mg/L)	≤0.5	-	100	-	-	13	-
Color (ADMI)	-	800–1200	-	-	-	-
Color (mg/L)	-	-	-	40–500	-	-	-
Color (Pt-Co)	-	80	-	-	50–2500	-	240–290

2. Materials and Methods

2.1. Experimental Setup

The HD reactor, with a working volume of 2 L, was constructed using a cylindrical plastic beaker measuring 14 cm in diameter and 15 cm in height. Initially, the reactor was inoculated with enhanced HD sludge that had been operating for over 400 days. The enriched sludge was added at a rate of 0.3 gVSS/L. This particular sludge exhibited a high nitrogen removal rate of approximately 85–95% under specific operating conditions, including a hydraulic retention time (HRT) of 8 h, influent feeding rate of 8.7 mL/min, H₂ supply of 30 mL/min and a nitrogen loading rate (NLR) of 313 gN/m³/day. The reactor was operated continuously in accordance with the procedures outlined in a previous study [13,14].

To assess the treatment efficiency of nitrogen and decolorization, the reactor was prepared on day 0 of the experiment by introducing organic pollutants. It was placed inside a water bath to maintain a temperature range of 30–35 °C, and a magnetic stirrer was utilized to facilitate thorough mixing of the liquid and sludge. Furthermore, plastic beads were employed to cover the top of the reactor, ensuring the maintenance of anaerobic–anoxic conditions. The reactor was connected to an influent synthetic feeding tank and a gas feeding line. A schematic diagram illustrating the experimental setup is presented in Figure 1.

2.2. Synthetic Textile Wastewater Preparation

The synthetic TW used in this study was prepared based on the quality of wastewater in Thailand, as described in a previous study [13,25]. The TW was synthesized by combining tap water with specific chemical reagents, and the details of this composition can be found in

Table 2. Composition of synthetic TW.

Substrate	Concentration
NaNO3	0.486 g/L as 80 mg NO3-N/L
KH2PO4	0.055 g/L as 12.5 mg P
NaHCO3	1.0 g/L as an inorganic carbon source
CaCl2	0.021 g/L as 7.5 mg Ca
MgSO4·7H2O	0.038 g/L as 3.75 mg Mg
Trace elements I a	1 mL/L
Trace elements II b	1 mL/L
Reactive Black 5 dye	0.030 g/L as color of 30 mg/L
CH3COONa	2.085 g/L as 1500 mg COD/L

^a Containing 5 g/L EDTA and 5 g/L FeSO₄. ^b Containing 15 g/L EDTA, 0.43 g/L ZnSO₄·7H₂O, 0.24 g/L CoCl₂·6H₂O, 1 g/L MnCl₂·4H₂O, 0.25 g/L CuSO₄·6H₂O, 0.22 g/L MnCl₂·5H₂O, 0.22 g/L NaMoO₄·5H₂O, 0.19 g/L NiCl₂·6H₂O, 0.21 g/L NaSeO₄·10H₂O and 0.014 g/L H₃BO₄.

. The initial concentration of NO₃-N remained consistent within the range of 80–85 mg-N/L. Regarding the color model utilized in this research, reactive black 5 (RB-5) dye, sourced from Fujifilm Wako Pure Chemical Corporation, Co., Ltd. (Sigma-Aldrich, Burlington, MA, USA), was employed. This dye exhibits a disazo-vinylsulfnyl structure (C₂₆H₂₁N₅Na₄O₁₉S₆), with a molecular weight of 991.82 g/mol and a maximum identifiable wavelength (λ_max) of 597 nm [26]. The initial concentration of RB-5 in this study was fixed at 30–40 mg/L. Additionally, sodium acetate (CH₃COONa) was introduced as an organic carbon source, contributing to the concentration of COD.

2.3. Investigation of Nitrogen Removal and Decolorization Efficiencies

The experiment was conducted in two phases: one with OM contamination and the other without, as outlined in

Table 3. Summary of the textile wastewater characteristics.

Operating Day	No OM Addition	OM Addition
	Day 0–20 RUN 1	Day 21–40 RUN 2	Day 41–60 RUN 3	Day 61–80 RUN 4
Initial concentrations
NO3-N (mgN/L)	80	80	80	80
RB-5 dye (mg/L)	30	30	30	30
COD (mg/L)	-	5000	3000	1500
H2 supply conditions	Continuous supply at 50 mL/min

. During RUN 1 (days 0–20), the HD reactor was operated with initial influent concentrations of 80 mg NO₃-N/L and 30 mg/L of RB-5 dye. Subsequently, in RUN 2–4, various concentrations of COD as OM were introduced to the reactor from days 21 to 80. This was carried out to investigate the impact of OM on the treatment performance of the HD system in terms of nitrogen removal efficiency (NRE) and dye removal efficiency (DRE). Specifically, the initial COD concentrations were set at 5000 mg/L for RUN 2, 3000 mg/L for RUN 3 and 1500 mg/L for RUN 4. However, the influent concentrations of NO₃-N and RB-5 dye remained the same across all runs. Gas supply feeding was conducted following the methods described in a prior study [13]. The gas pipeline was fixed and consistently supplied with H₂ gas at a rate of 50 mL/min. The H₂ gas was generated using an H₂ generator (HG-260, GL Sciences) and then delivered through an air stone diffuser. Throughout the experiment, the HD reactor was tested with an HRT of 12 h under continuous feeding conditions.

2.4. Water Sample Analyses

Prior to sampling, the measurements of dissolved oxygen (DO), pH, dissolved hydrogen (DH) and temperature were carried out within the reactors using specific instruments. The probe (YSI 58 Dissolved Oxygen Meter, Xylem Inc., Washington, DC, USA), pH meter (Horiba-B712, HORIBA Advanced Techno, Co., Ltd., Kyoto, Japan), DH meter (ENH-2000, TRUSTLEX Inc., Osaka, Japan) and digital thermometer (WT-6, Jinan Retekool Inc., Jinan, China) were utilized for this purpose. Subsequently, influent and effluent samples were subjected to centrifugation at 10,000 rpm for 5 min, and the resulting supernatant was analyzed for concentrations of NO₃-N, NO₂-N, RB-5 and COD. These samples were then stored at a temperature of −18 °C to facilitate further analysis. All parameter determinations were carried out in accordance with standard methods for water and wastewater analysis [27]. Specifically, the concentrations of NO₃-N, NO₂-N and RB-5 were analyzed using a spectrophotometer (UV-1800 Shimadzu-Spectrophotometer, Shimadzu Corp., Kyoto, Japan) employing the colorimetric method. The RB-5 concentration was determined using a standard curve generated by plotting known concentrations of RB-5 against their corresponding absorbance values at a maximum wavelength (λ_max) of 597 nm [28]. Additionally, the COD concentration was determined using the closed reflux method.

2.5. Bacterial Community Analysis

The sludge samples collected at various stages of the experiment, including the initial sludge at the beginning and the end of each phase (on days 20, 40, 60 and 80 from RUN 1 to 4), were utilized for bacterial community analysis. Total DNA extraction was performed using a FastDNA^® SPIN Kit for soil analysis (MP Biomedicals LLC, Santa Ana, CA, USA), following previously described methods [14] by utilizing the collected sludge samples of approximate wet weight of 0.12–0.13 g. The concentration of DNA was determined using a QuantusTM Fluorometer (Promega Corp., Madison, WI, USA) through nanodrop analysis. Subsequently, the DNA samples were sent to a commercial service for Next-Generation Sequencing (FASMAC Co., Ltd., Kanagawa, Japan). Sequencing targeted the V4 region of the bacterial 16S rRNA gene, and amplification was carried out using Univ-515F (5-GTG YCA GCM GCC GCG GTA A-3) and Univ-806R (5-GGA CTA CNV GGG TWT CTA AT-3) primers. The MiSeq platform was employed for obtaining amplified metagenomic sequences. To classify the raw sequence data taxonomically, QIIME version 1.9.0 was utilized, resulting in the identification of operational taxonomic units that were subsequently clustered at 97% similarity. The relative abundances of these taxonomic units during different operational stages were visualized using a heatmap generated using R software version 4.0.2, along with the Heatplus (version 2.30.0) and Vegan (version 2.5.6) packages.

2.6. Calculations and Statistical Analyses

The values of NLR [gN/m³/day] and NRE [%] were calculated using Equations (1) and (2), and Equations (3) and (4) were employed to determine the DRE of RB-5 [%] and the COD removal efficiency (CRE) [%]. C₀ and C₁ represent the initial and final concentrations, respectively, of the specific parameters indicated in each variable (NO3 for NO₃ and NO2 for NO₂). Q represents the flow rate of the influent feeding, and V denotes the volume of the reactor. Furthermore, ABS₀ and ABS₁ correspond to the measured absorbances of the initial and treated water, respectively. To assess potential variations in decolorization under different operational conditions, a 95% confidence interval was obtained using an analysis of variance (ANOVA) and least significant difference (LSD) test conducted with SPSS statistical software version 25. These analyses aimed to identify significant differences in decolorization among the experimental conditions.

$${\mathrm{NLR}=(\mathrm{C}}_{0,{\mathrm{NO}}_3}[\mathrm{g}\mathrm{N}/\mathrm{L}]\times \mathrm{Q}[\mathrm{L}/\mathrm{day}]/\mathrm{V}[{\mathrm{m}}^3]$$

(1)

$${\mathrm{NLR}=(\mathrm{C}}_{0,{\mathrm{NO}}_3}[\mathrm{mg}\mathrm{N}/\mathrm{L}]-{\mathrm{C}}_{1,{\mathrm{NO}}_3}[\mathrm{mg}\mathrm{N}/\mathrm{L}]-{\mathrm{C}}_{1,{\mathrm{NO}}_2}[\mathrm{mg}\mathrm{N}/\mathrm{L}]\times 100{/\mathrm{C}}_{0,{\mathrm{NO}}_3}[\mathrm{mg}\mathrm{N}/\mathrm{L}]$$

(2)

DRE = (ABS₀ − ABS₁) × 100/ABS₀

(3)

CRE = (C_0,COD [mg/L] − C_1,COD [mg/L]) × 100/C_0,COD [mg/L]

(4)

3. Results and Discussion

3.1. HD Performance on Nitrogen Removal

In this study, the experiment was conducted and maintained over a long-term period with an HRT of 12 h. Figure 2 displays the effluent concentrations of NO₃-N and NO₂-N, along with the corresponding NRE, obtained from an HD reactor under two conditions: without the presence of OM (represented as COD concentrations) and with various dosages of OM. The findings revealed that, due to the short HRT, there was frequent accumulation of NO₃ and NO₂ in the effluent.

At the beginning of the experiment (RUN 1), in the absence of OM contamination, both parameters exhibited higher levels in the treated water, leading to an unstable NRE during this phase. In accordance with the experimental details, varying concentrations of OM ranging from 1500 to 5000 mg COD/L were subsequently introduced into the reactor for RUN 2–4. The results showed that, in RUN 2, with an initial COD concentration of 5000 mg/L, some NO₃-N concentrations remained incomplete, ranging from 10 to 15 mg/L in the effluent. However, in RUN 3–4, where a lower initial COD concentration in the range of 1500–3000 mg/L was utilized, high ranges of NRE approached close to 100%, and the concentrations of effluent NO₃ and NO₂ were nearly zero after day 50. This improvement may be attributed to the gradual adaptation of the bacteria in the reactor to these activities when exposed to organic pollutants. The results indicate that bacteria were more effective in reducing NO₃ and NO₂ at low COD dosages [29]. Thus, a small amount of organic carbon sources might be suitable for denitrification in this reactor.

On the other hand, it is possible that some heterotrophic bacteria occurred and utilized the OM as carbon sources, resulting in improved NRE after the addition of organic matter in RUN 2–4. Previous studies have confirmed that heterotrophic denitrification can occur within HD systems when supplied with various organic carbon sources [30]. The average value of NRE under different operating conditions indicated that the NRE was close to 100% when organic matter was introduced to the reactor. In contrast, the average value of NRE without and with organic matter contamination only increased from 85% to 90–100%. Therefore, the results of this study indicate that no significant effect of OM on NRE was observed. This might be due to the dominance of autotrophic denitrifying bacteria in the HD system, which required no organic carbon for their respiration. Previous studies have suggested that bacteria involved in denitrification processes, such as heterotrophic denitrifiers, rely on organic carbon as an electron donor for respiration [22]. Hence, organic carbon plays a crucial role in supporting the activities of these bacteria in biological denitrification, thereby enhancing the treatment performance of the denitrification system. Moreover, several reports have reported different stoichiometric ratios for heterotrophic denitrification. Some suggest that 4 mg of COD as an organic carbon source is required to reduce 1 mg/L of NO₃-N to N₂, whereas others propose that 1 mg of NO₃-N reduction necessitates 2.86 mg of COD, and 6 mg of COD oxidizes 1 mg of NO₃-N [31]. Previous studies recommend designing conventional wastewater treatment systems with a COD/NO₃-N ratio in the range of 5.3–11 or higher to optimize the denitrification process [32,33].

Therefore, this study confirmed that the presence of OM in the reactor had no inhibition on the denitrification process through the HD system; instead, it showed a slight improvement in NRE. The organic matter employed in this study served as an organic carbon source for the bacteria in the reactor, enhancing the respiration process and resulting in increased rates of NO₃-N and NO₂-N removal, ultimately leading to an elevated NRE. Similarly, previous findings indicate that the presence of OM has no effect on the removal of NO₃-N and NO₂-N due to the co-existence of denitrifiers [29]. However, varying COD concentrations had no significant impact on the treatment performance of the HD system in this study.

3.2. HD Performance on Decolorization of RB-5 Dye

Figure 3 illustrates the RB-5 dye concentrations and DRE from the HD reactor under different operating conditions (RUN 1–4). The results suggest that, due to the short operation time and lack of electron donors, the degradation of the dye in this study exhibited fluctuations and remained incomplete. Furthermore, NO₃-N and NO₂-N were more favorable electron acceptors than the dye, and they tended to compete faster for electron donors, such as H₂ and organic carbons [34]. H₂ gas and carbon sources had the advantage of acting as electron donors, facilitating the transformation of dye molecules into aromatic amines through anaerobic processes [35]. Therefore, biodegradation typically occurred after denitrification was completed.

The findings confirm that RB-5 can undergo reduction through the HD process alongside NO₃-N and NO₂-N in the presence of organic pollutants. However, significant amounts of RB-5 dye accumulated, primarily ranging from 10 to 30 mg/L in the effluent wastewater, compared to the initial dye concentration of 30–40 mg/L. In the absence of OM, the effluent dye concentrations remained high, and the average DRE value was relatively low, with a range of 25 ± 14%. Subsequently, the addition of OM in RUN 2–4 led to a doubling of the DRE ratio, reaching the range of 50–60 ± 18%. OM has been reported to be a major factor in the cleavage of azo bonds during the biodegradation process, serving as a carbon source for bacterial activities involved in dye degradation [25]. Carbon sources significantly influence energy availability and act as electron donors for the reduction required to break down azo bonds, resulting in dye removal. Some microorganisms in treatment systems cannot directly break down dye molecules; hence, carbon and nitrogen sources were introduced to facilitate co-metabolism and to enhance the rate of dye removal. The mechanism of dye degradation typically consists of two steps: the cleavage of the azo bond (-N=N-) by azo reductase in the presence of suitable electron donors, leading to the formation of amino groups (NH₂) and colorless intermediates via anaerobic processes [36]. Subsequently, the amino groups and other intermediate molecules were degraded under aerobic conditions [37,38]. Therefore, organic matter played a crucial role in the breakdown of azo molecules, thereby enhancing DRE. Additionally, the results indicate that varying COD dosages had no significant impact on the decolorization rate in this study. However, previous studies suggest that higher COD dosages can improve the decolorization rate by providing sufficient electron donors, whereas low COD concentrations resulted in an insufficient supply of electron donors for the degradation process [39]. However, some studies have also indicated that higher concentrations of carbon or nitrogen sources have no pronounced effect on the decolorization rate because bacteria primarily act as co-metabolites alongside dye molecules [40]. Generally, the reductive cleavage of a single azo bond requires four electrons, equivalent to 32 mg/L of COD as the carbon source, to reduce 1 mmol of monozo dye [36,41].

In conclusion, this study demonstrates that RB-5 dye can be partially reduced alongside NO₃-N and NO₂-N in the presence of organic pollutants, but the process remains incomplete due to the dominance of the denitrification process. However, the addition of organic matter can improve the removal efficiency of RB-5. Higher amounts of organic carbon sources lead to increased DRE, whereas there is no significant effect of initial COD concentrations on the decolorization rate.

3.3. HD Performance on COD Removal

Effluent in the textile industry contains a variety of toxic pollutants resulting from different processes. One crucial parameter is the organic pollutant, which is typically quantified as the COD concentration and is often found in high levels. Under anaerobic conditions, this parameter is generally reduced, whereas under aerobic conditions, its presence is minimal. In this experiment, our focus was on assessing the treatment performance of the HD process in terms of COD removal efficiency (CRE) when applied to textile wastewater (TW) contaminated with NO₃ and RB-5 dye. Figure 4 illustrates the results obtained from averaging the COD concentrations and correlating them with CRE values under various initial COD dosage conditions.

The findings reveal that a high concentration of COD still accumulated in the discharged wastewater. However, CRE remained relatively stable under different COD dosage conditions, ranging from 40% to 55%, compared to the nitrogen removal efficiency (NRE) and dye removal efficiency (DRE). The results confirm that varying COD dosages had no significant impact on the COD removal efficiency in this study. Previous studies focusing on the treatment performance of HD and the combination of HD with aerobic processes for treating organic-matter-contaminated domestic wastewater with influent COD concentrations between 1500 and 2000 mg/L have reported high CRE values ranging from approximately 70% to 90% [23]. In contrast, the observed relatively low CRE in this study may be attributed to the persistence of certain toxic chemicals following the processes of dye degradation, as dye colors often have the potential to convert into other toxic forms. Normally, certain toxic forms, including -NH₂ compounds, accumulate under anaerobic conditions and are subsequently degraded under aerobic conditions [37,38]. Thus, the dye present in the effluent under anaerobic conditions was observed as both residual dye and some aromatic amines, leading to decreased COD removal rates [42]. Toxic by-products of dye degradation, such as sulfanilic acid and 1-amino-2-napthol, can significantly inhibit bacterial activity [43]. In this study, using RB-5 dye as a representative colorant, the dye was typically converted to sulfanilic acid, which easily inhibited bacterial activity in the HD reactor, which is consistent with previous findings [39,44,45,46]. Additionally, previous studies have shown that CRE increases as the dye concentration decreases due to a lack of dye metabolism after the degradation process [42]. However, some results have indicated that a high dye concentration has no significant impact on the COD removal rate, possibly because some dyes are non-toxic to bacterial activities [39,45]. A slight decrease in the COD removal rate was observed as the dye concentration increased, indicating that the dye concentration had no influence on the activity of the bacteria involved in the process.

Therefore, although the HD reactor demonstrated some capability in reducing the COD concentration, the process remained incomplete. The presence of toxic chemicals from RB-5 degradation and residual dye concentration could contaminate the discharged wastewater, resulting in low CRE. However, there was no effect of the presence of NO₃-N and NO₂-N on the COD removal rate. Similarly, various initial COD concentrations had no significant impact on the CRE.

3.4. Bacterial Community under Different Operating Conditions

The initial sludge of the HD reactor used in this study consisted of HD sludge that had been operated for over 400 days with an initial NO₃-N concentration of 40 mg/L. The analysis of the initial sludge revealed the presence of Proteobacteria, Bacteroidetes, Firmicutes and Planctomycetes as the dominant phyla, accounting for 78.8%, 10.3%, 5.6% and 2.6%, respectively. At the class level, the microbial community showed variations, with Betaproteobacteria (59.3%), Alphaproteobacteria (12.8%), Cytophagia (4.6%) and Gammaproteobacteria (3.8%) being the most prominent classes. Specifically, Thauera spp., belonging to the class Betaproteobacteria, were identified as the dominant bacterial community at the family and genus levels during the initial stage of the reactor.

Subsequently, the reactor was operated under different conditions for more than 250 days following the methodology described by Singhopon et al. [13,14]. RUN 1 was initiated to assess the treatment performance of the HD reactor and to study the microbial community structure when contaminated with organic pollutants. Sludge samples were collected from the HD reactor at the end of each phase, on days 20, 40, 60 and 80 for RUN 1 to 4, corresponding to periods without and with OM contamination.

At the phylum level, the results of RUN 1 indicate the presence of Proteobacteria (85.6%), Chloroflexi (7.3%) and Firmicutes (4.7%). However, the composition slightly changed to Proteobacteria (60–75%) and Firmicutes (18–34%) after the addition of organic pollutants in RUN 2 to 4. The class-level analysis showed significant differences as well. In RUN 1, the microbial community consisted of Betaproteobacteria (62.4%), Gammaproteobacteria (17.1%), Anaerolineae (7.3%) and Alphaproteobacteria (6.0%). However, after introducing OM, Gammaproteobacteria, Erysipelotrichia, Clostridia and Alphaproteobacteria accounted for 55–67%, 10–19%, 7–15% and 4–7% of the community, respectively. These results indicate that the distribution of the microbial community structure was significantly affected by organic pollutants. The abundance of Gammaproteobacteria and Erysipelotrichia increased with the addition of OM, and the abundance of Betaproteobacteria and Anaerolineae decreased. Conversely, there was no impact of OM on the distribution of Alphaproteobacteria, which remained stable.

Figure 5 provides a detailed overview of the microbial communities at the family and genus levels. The main bacterial communities identified in this reactor were closely related to denitrifying bacteria, as well as some autotrophic and heterotrophic denitrifying bacteria. Previous studies have highlighted the importance of factors such as NO₃-N concentration, initial dye concentration and H₂ gas volume in influencing the distribution of bacterial communities in the HD process [14]. Additionally,

Table 4. Comparison of the relative abundance of bacterial communities at the genus level with the average performance in nitrogen treatment, decolorization and COD removal.

Sludge Samples	Bacterial Relative Abundance (%)
	Initial Sludge	RUN 1	RUN 2	RUN 3	RUN 4
Bacterial communities
Thauera spp.	51.6	15.5	4.2	28.3	28.9
Xanthomonadaceae	0.6	16.3	-	-	-
Rhodocyclaceae	4.6	46.0	0.1	1.1	0.2
Azoarcus	-	-	48.6	9.7	11.8
Erysipelothrix	-	-	18.6	10.5	10.5
Alishewanella	-	-	0.7	15.4	9.2
Alkaliphilus	-	-	0.4	7.5	6.8
Parameters		Treatment efficiency (%)
Nitrogen		86 ± 9	90 ± 3	95 ± 4	99 ± 2
Decolorization		25 ± 14	50 ± 18	58 ± 17	51 ± 9
COD		-	47 ± 446	43 ± 434	50 ± 359

lists the relative abundances at the genus level and compares them with the average removal efficiency of each parameter.

In RUN 1, without OM contamination, the dominant bacterial community consisted of Rhodocyclaceae (class Betaproteobacteria), Xanthomonadaceae (class Alphaproteobacteria) and Thauera spp. (class Betaproteobacteria). The relative abundance of Rhodocyclaceae and Xanthomonadaceae increased, whereas the relative abundance of Thauera spp. decreased. Previous research has indicated that the relative abundance of Rhodocyclaceae increases with RB-5 concentration, whereas unclassified Xanthomonadaceae decreases with RB-5 concentration. Despite this, the nitrogen removal efficiency remained close to 90%, and decolorization efficiency reached approximately 25% due to the activities of these bacteria.

Upon the addition of organic pollutants in RUN 2 to 4, Thauera spp., Azoarcus spp., Alishewanella spp., Alkaliphilus spp. and Erysipelothrix spp. (all belonging to different bacterial classes) became the dominant bacterial communities. The results confirm that the relative abundances of Xanthomonadaceae and Rhodocyclaceae were inhibited by the presence of OM. In RUN 2, characterized by a high dosage of COD concentration, Azoarcus spp. and Erysipelothrix spp. showed high relative abundances. Nevertheless, nitrogen removal efficiency remained close to 90%, and the COD removal efficiency was approximately 50%, indicating a doubled decolorization rate. These bacteria played a commensal role and significantly enhanced the treatment efficiency under these conditions.

In RUN 3 and RUN 4, a low COD concentration was supplied to the reactor. The relative abundances of Thauera spp., Alishewanella spp. and Alkaliphilus spp. increased, whereas Azoarcus spp. and Erysipelothrix spp. decreased. However, organic pollutants had no inhibition on the relative abundance of Thauera spp., which exhibited a high relative abundance at low COD concentrations. This could be attributed to the adaptability of these bacteria to highly toxic pollutants, although they typically require some adaptation time. The results confirm that NRE, DRE and CRE remained stable, indicating that these bacteria effectively removed pollutants in the reactor. Therefore, this study concludes that organic pollutants significantly influenced the bacterial communities in the HD system, with the different bacterial species functioning together to enhance treatment efficiency.

The previous literature suggests that Xanthomonadaceae, often considered a suitable candidate for autotrophic denitrification, is commonly found in HD systems that use substantial amounts of salt and bicarbonate [47]. Similarly, Rhodocyclaceae is an important denitrifying bacterium involved in the reduction of NO₃ to NO₂ in the HD process. Previous findings have shown that its relative abundance increases with higher H₂ flow rates [16]. Thauera spp., frequently detected in both autotrophic and heterotrophic denitrification conditions, plays a crucial role in HD systems [48]. The low relative abundance of Thauera spp. observed in this study could be linked to the low volume of H₂ gas. Additionally, many Thauera spp. strains have been recognized as efficient degraders of aromatic compounds and exhibit versatile aromatic compound-degrading capabilities under denitrification conditions compared to aerobic conditions. These bacteria can utilize intermediate metabolites generated during dye degradation as the sole carbon source in the decolorization process [49].

The genus Azoarcus, a facultatively anaerobic, mesophilic, non-motile, Gram-negative bacterium, isolated from wastewater treatment plants, is known for its ability to degrade o-phthalate and a wide range of aromatic compounds using nitrate as an electron acceptor [50]. Azoarcus has been reported to be commonly found in various environments, including anoxic wastewater sludge, grass root soil, and nitrogen-fixing bacteria. Some Azoarcus strains also possess the capability to degrade aromatic compounds through anaerobic degradation linked to nitrate reduction [51,52,53]. They have been identified as predominant potential denitrifying genera contributing significantly to nitrogen removal in wastewater treatment plants and have been observed in activated sludge reactors treating ammonium-rich, high-organic tannery and coking wastewater [54,55]. Azoarcus has also been involved in the degradation of polycyclic aromatic hydrocarbons in municipal and industrial wastewater treatment processes. Additionally, it has been found to remove toluene and phenol under denitrification conditions [56].

Similarly, Alishewanella has been reported as an efficient degrader of textile dyes and exhibits efficient denitrification capacity under high alkaline conditions and high initial nitrogen concentrations [57]. The Erysipelothrix and Alkaliphilus genera are commonly found in textile industry effluents [58].

In conclusion, Xanthomonadaceae, Rhodocyclaceae and Thauera spp. acted as commensal bacteria in the absence of organic pollutants, and Thauera spp., Azoarcus spp., Alishewanella spp., Erysipelothrix spp. and Alkaliphilus spp. became dominant microbial communities following the addition of organic pollutants. This shift in bacterial composition contributed to enhanced nitrogen removal performance and reduced RB-5 dye and COD concentrations in the reactor. The results highlight the significant influence of organic pollutants on bacterial communities in the HD system. Despite their differences, these bacterial species worked together to enhance treatment efficiency in the reactor.

4. Conclusions

An HD reactor was established with an HRT of 12 h in the continuous feeding mode to investigate the simultaneous treatment performance of nitrogen and decolorization when TW was contaminated with OM. The reactor was operated under two distinct phases: without OM and with OM. RB-5 dye was chosen as the representative color source, and sodium acetate served as the organic source in this experiment. Although the HD system was capable of removing all pollutants, including NO₃-N, NO₂-N, RB-5 dye and COD, the process remained incomplete due to the relatively short HRT. It is worth noting that the presence of organic pollutants did not hinder the treatment efficiency of nitrogen and decolorization. The overall NRE was higher than the DRE, as NO₃-N and NO₂-N acted as more favorable electron acceptors, thus competing for electron donors at a faster rate compared to the dye. Following the addition of OM, the NRE slightly increased from 85% to 90–100%, and the decolorization rate doubled from 25% to 50–60%. Consequently, the results confirm that organic matter had a slight enhancing effect on NRE and a greater impact on DRE, although the different doses of organic matter had no significant influence on both parameters. OM played a crucial role in stimulating the activities of heterotrophic bacteria during the biological denitrification process. It also acted as a carbon source for bacterial activities, facilitating the cleavage of azo bonds during the dye degradation process. Consequently, some heterotrophic denitrification bacteria likely played a role alongside autotrophic denitrification bacteria in reducing nitrogen and dye within the reactor. Furthermore, the results indicate that a certain amount of organic matter, represented by the concentration of COD, was reduced within the HD system, typically ranging from 40% to 55% across different initial COD concentrations. As a result of the dye degradation process, various toxic chemicals, such as aromatic amines and other hazardous forms, were often generated and presented as organic pollutants once the azo bond was cleaved. Moreover, the degradation products of RB-5, acting as toxic molecules, inhibited bacterial activities within the HD reactor, leading to a decrease in the COD removal rate in this experiment. The findings of this study indicate that the HD system exhibits superior performance in the simultaneous treatment of COD, dye color and NO₃-N compared to other existing treatment methods, as outlined in Table 1. Among the bacterial community, Xanthomonadaceae, Rhodocyclaceae and Thauera spp. functioned as commensal bacteria in the absence of organic pollutants. However, after the addition of organic pollutants, Thauera spp., along with some Azoarcus spp., Erysipelothrix spp., Alishewanella spp. and Alkaliphilus spp., were found to dominate the microbial community. Despite these changes, the treatment efficiency of all parameters remained stable and improved through the activities of these bacteria. In practical terms, this approach presents an opportunity to enhance treatment efficiency and enable the treatment of textile wastewater through cost-effective and environmentally friendly processes prior to discharge, making it particularly suitable for developing countries.