Achieving enhanced biological nitrogen removal via 2450 MHz electromagnetic wave loading on returned sludge in anaerobic-anoxic-oxic process

Anaerobic-anoxic-oxic (A/A/O) process, which allows simultaneous biological removal of organic matter, nitrogen and phosphorus compounds, has been widely applied in wastewater treatment (Lim et al. 2009; van Loosdrecht & Brdjanovic 2014). However, along with the increasing severity of eutrophication, wastewater with low carbon/nitrogen (C/N) ratios would not provide sufficient carbon sources for denitrifying bacteria, which causes low nitrogen removal efficiency (Wang et al. 2015a). This serious phenomenon has hindered the development of the A/A/O process and made it difficult to meet stricter regulations. A conventional method to improve the denitrification effect is to add extra carbon sources (i.e. acetate, glucose, methanol, and ethanol) as carbon complements (Cho et al. 2004), but these processes would generate additional expense and produce more excess sludge (Gao et al. 2011).

In order to tackle the pressing problem, much interest has been observed among the scientific community in adopting strategies to simultaneously enhance nitrogen removal and in-situ sludge reduction (Nolasco et al. 2002). Sludge contains a significant amount of organic substances, and the organic substances released upon sludge disintegration can be used as a carbon source to improve the biological denitrification efficiency (Zubrowska-Sudol & Walczak 2015). Such a method is also considered to be the mechanism of the lysis-cryptic growth for sludge reduction, which is of great significance for reducing the excess sludge production (Wei et al. 2003; Xu et al. 2018). Thus, some in-situ improvements to conventional A/A/O techniques were established to facilitate the efficient release of carbon source from excess wasted activated sludge or returned activated sludge (Gao et al. 2011; An et al. 2017). Currently, sludge pretreatment technologies based on carbon source release have been developed widely, such as physical, chemical, mechanical, biological hydrolysis or a combination of these methods (Nolasco et al. 2002; Zubrowska-Sudol & Walczak 2015; Liu et al. 2017). In view of full-scale applications, the in-situ pretreatment method must be cost-effective, with simple design and management, without negative impacts on effluent quality, and no secondary pollutants.

As an effective sludge pretreatment method, electromagnetic wave loading has gained widespread popularity (Yu et al. 2010). In particular, sludge treatment applications generally use a frequency of 2,450 MHz with a wavelength of 12.24 cm (Tang et al. 2010). It is generally believed that the electromagnetic wave loading causes the dipole molecules in the sludge to rotate at high speed, causing the sludge flocs and microbial cells to rupture; at the same time, the polarized parts of the macromolecules are aligned with the poles of the electromagnetic field, resulting in possible breakage of hydrogen bonds and the death of the microorganisms (Hong et al. 2004; Eskicioglu et al. 2007). As reported, electromagnetic wave loading changes the physico-chemical properties of sludge and destroys its micro-structure (Yu et al. 2010; Yeneneh et al. 2015), which may positively affect the subsequent treatment and disposal of this very complex material (Tyagi & Lo 2013). Possessing the ability to disintegrate extracellular polymeric substances and microbial cells, it had been established that electromagnetic wave and its hybrid treatment processes could bring about the solubilization of sludge, release of soluble organics and enhancement of sludge anaerobic digestion (Eskicioglu et al. 2007; Yeneneh et al. 2015; Sang et al. 2020). Considering its convenience, little cost and great cell solubilization, electromagnetic wave loading could also be employed to the enhancement of biological nitrogen removal based on carbon source release and sludge reduction based on lysis-cryptic growth (Wang et al. 2015b).

In essence, the biological treatment technology of activated sludge is the growth and metabolism process of various microorganisms. Therefore, treatment efficiency and stable operation of the process are determined by the microbial community and physico-chemical composition of the activated sludge (Bell et al. 2005). However, to our knowledge, the effects of electromagnetic wave loading alone on process performance and corresponding microbial community structures of A/A/O have been rarely reported. Hence, comprehensive analysis on process performance of combining electromagnetic wave loading with A/A/O at the level of the microbial community structure is necessary and promising.

The objective of this research was to evaluate the possibility of increasing the effectiveness of biological nitrogen removal by using electromagnetic wave loading on the returned sludge. In our work, the A/A/O reactor was operated, in which the returned sludge was loaded by 2,450 MHz electromagnetic wave. Continuous operation was implemented, and excess sludge production and pollutant removal were evaluated. In parallel with the changes in microbial activity, the dissolution effect of organic compounds and nutrients in the activated sludge flocs was monitored. Furthermore, the mechanism of electromagnetic wave loading on returned sludge to enhance biological denitrification was explained by Illumina MiSeq sequencing technology. This study could propose a strategy for improving the biological nitrogen removal performance of wastewater with low C/N ratio.

Artificial simulated wastewater was used for the influent of this experiment. The specific characteristics were as follows: chemical oxygen demand (COD) of 240–250 mg/L, total nitrogen (TN) of 35–40 mg/L, total phosphorus (TP) of 3.5–4.0 mg/L and pH of 6–8. The concentrations of MgSO₄·7H₂O, CaCl₂ and trace solution were set as 27, 30 and 1 mg/L, respectively.

A laboratory-scale A/A/O reactor was used in this study (Figure 1), which had been continuously operated for over half a year. The inoculated activated sludge was collected from the secondary sedimentation tank of Shahu wastewater treatment plant in Wuhan, China. Effective volumes of the anaerobic, anoxic, oxic and secondary sedimentation tanks were 10, 15, 40 and 29 L, respectively. During the experiment, the flow rate of influent was 5 L/h, temperature in the reactor was maintained at 25 °C approximately, and dissolved oxygen (DO) concentration of the oxic tank was about 4 mg/L. The hydraulic retention time (HRT) of each reaction tank was 2, 3 and 8 h, respectively; and excess sludge was periodically discharged from the secondary sedimentation tank to maintain a sludge retention time (SRT) of 15 days. The returned sludge rate of A/A/O was controlled at 50% of the influent flow rate, and the mixed liquor recirculation ratio was maintained at 200% for denitrification. In order to add the element of the electromagnetic wave loading unit, a branch pipe was added to the returned sludge pipe of the A/A/O reactor, and a 2,450 MHz electromagnetic wave was used to load on returned sludge.

In the experiment, the electromagnetic wave loading power was set to 265 W by adjusting the output voltage, and loading time was maintained at 45 s by changing the length of the pipe inside the loading unit. Under each operating condition, 0%, 30, 60 and 100% of the returned sludge was respectively loaded by electromagnetic wave and then returned to the system. Meanwhile, the remaining (100, 70, 40 and 0%) returned sludge was recycled directly to the system via the sludge return. The reactor was operated for 50 days under each loading condition. During operation, the influent and effluent of the reactor were collected from the reactor every two days. The mixed liquor and returned sludge were collected daily for the last three days of operation, and 30 mL collected samples were centrifuged at 4,000 × g for 15 min, then the supernatant was filtrated through the 0.45 μm membrane, and the filtrate was used for further analysis.

The conventional tests for COD, ammonium nitrogen (NH₄⁺-N), nitrate nitrogen (NO₃⁻-N), TN, TP, mixed liquor suspended solids (MLSS), and mixed liquor volatile suspended solids (MLVSS) concentrations were measured according to the standard methods (APHA 2005). The measuring methods for specific oxygen uptake rate (SOUR) and endogenous respiration rate are summarized in Supplementary Text S1. The denitrification rate (ρ_NO3-N) was measured as described by Xu et al. (2018). Each test was performed at least in duplicate, and the average values were obtained.

The cumulative sludge production (W_n, g SS) and the sludge growth rate (Y_sgr, g SS/day) were selected to evaluate the sludge reduction ability of the EW-A/A/O system, and the specific calculation method referred to Xu et al. (2018).

Samples from anaerobic, anoxic, oxic and returned sludge in the reactor were taken on the last day of each operating condition. Propidium monoazide (PMA) dye pre-treatment was performed before DNA extraction to achieve high-throughput sequencing based on viable bacteria (Rizzotti et al. 2015). The PMA-pretreated sludge samples were subjected to DNA extraction by employing the E.Z.N.A^™ Mag-Bind Soil DNA Kit. The purity and concentration of nucleic acid were accurately measured with Qubit 3.0 fluorescent agent. After DNA extraction, PCR amplification was carried out. Bacterial V3-V4 region of the 16S rRNA gene was amplified using forward primer 341F: 5′-CCCTACACG ACGCTCT-T-CCGATCTG(barcode)CCTACGGGNGGCWGCAG-3′ and reverse primer 805R: 5′-GACTGGAGTTCCTTGGCACCCGAGAATTCCA(barcode)GACTACHVGGGTATCAATCC-3′. PCR products were produced using QIAquick PCR Purification kit, and then utilized for pyrosequencing on the Illumina MiSeq platform (Sangon Biotech, Shanghai, China). The obtained results (raw reads) were deposited into the NCBI short reads archive database under the accession number SRP256521.

Cutadapt software was used to cut the primer sequence, and PEAR was used for sequence splicing, for which the maximum mismatch ratio allowed in the overlap region of the spliced sequence was 10%. Prinseq was used for quality control to remove low-quality sequences shorter than 200 bp, with one or more ambiguous bases and quality score inferior to 25. In addition, chimeras were identified with the ‘uchime’ command. Sequences were clustered into operational taxonomic units (OTUs) by setting a distance limit of 0.03 (equivalent to 97% similarity) by using Usearch. The ribosomal database project (RDP) classifier was used to compare the OTU representative sequence with the Silva database (http://www.arb-silva.de/) for species taxonomy, where the confidence threshold was set to 80%. Additionally, alpha diversity statistics (i.e. Chao1, ACE, Shannon, Simpson and Good's Coverage indexes) were generated for each sample in Mothur software (http://www.mothur.org).

MLSS measurement was performed to evaluate the excess sludge reduction effect of the system with electromagnetic wave loading on returned sludge. W_n and Y_sgr are identified as the vital factors to evaluate sludge minimization. Normally, W_n of the four systems (original, LR = 30%, LR = 60% and LR = 100%) increased linearly with the longer operation time. In Figure 2, the mean sludge production after 50 days’ monitoring in the original system was 13.1 g SS/day, while the Y_sgr of the other three systems was 9.6 (LR = 30%), 8.6 (LR = 60%) and 12.2 g SS/day (LR = 100%), with the reduction effects of 27, 34 and 6.8%, respectively. At the end of the operation, the four systems cumulatively produced excess sludge for 630.5 (original), 454.1 (LR = 30%), 422.9 (LR = 60%) and 582.9 g SS (LR = 100%).

The decrease of the sludge production indicated that the electromagnetic wave loading on returned sludge in the A/A/O process had great potential in excess sludge reduction, which was mainly attributed to the release and subsequent degradation of organic matter (Xu et al. 2018). Simultaneously, it was found that there was no linear correlation between the change of sludge reduction effect and the change of LR. Within a certain range, sludge reduction effect was improved with the increase of LR. However, when the LR of the returned sludge continued to increase, the organic load could be significantly increased, while the reduction of biomass in the system was even more obvious (Text S2; Table S1). Therefore, the organic components in the decomposition liquid cannot be fully utilized by microorganisms, so these redundant organic components were continuously accumulated in the activated sludge, which led to a downward trend in the sludge reduction effect.

The pollutant removal performances of the systems are denoted in Figure 3. During the stable stage, all of the four processes achieved good removal of COD (Figure 3(a)). Mean effluent COD concentration in four systems was enumerated 35.1 ± 2.3 (original), 38.7 ± 3.1 (LR = 30%), 41.9 ± 2.1 (LR = 60%) and 49.2 ± 1.8 mg/L (LR = 100%), respectively. The reason for higher effluent COD in three treated systems was the attribution of the higher COD load in the reactor due to substrate release with electromagnetic wave loading on returned sludge (Yasui & Shibata 1994; Wang et al. 2015b).

Figure 3(b) depicted variations of effluent NH₄⁺-N concentration. The effluent NH₄⁺-N concentration in four systems was 0.5 ± 0.11, 1.9 ± 0.4, 2.0 ± 0.5 and 0.8 ± 0.17 mg/L, which showed that electromagnetic wave loading on the returned sludge had unobvious adverse impact on NH₄⁺-N removal. By comparison, effluent NH₄⁺-N concentration in four systems could all meet the Grade 1A (<5 mg/L) in the discharge standard of pollutants for municipal wastewater treatment plant of China (GB18918–2002).

The performance of TN removal by the four processes is shown in Figure 3(c). With TN concentration in the influent of 40.3 ± 1.3 mg/L, mean effluent TN concentrations were lower than 15 mg/L in four systems, showing the concentrations of 14.7 ± 0.7, 12.8 ± 1.2, 13.6 ± 1.1 and 14.6 ± 0.2 mg/L, respectively. This meant that the denitrification effect was improved with electromagnetic wave loading on returned sludge. In treated systems, COD in returned sludge could be released along with the cell lysis and organic substances hydrolysis, and therefore more carbon source was provided for denitrification (Liu et al. 2017). Although effluent NH₄⁺-N concentration increased slightly, the removal effect of TN did not deteriorate, which was associated with the abundance of denitrifying bacteria. Moreover, it could be speculated that denitrification of microorganisms in the system enriched after a proper percentage of returned sludge was loaded by electromagnetic wave, and the reaction of NO_x⁻-N with reduction to N₂ could be carried out more completely.

It has been recognized that biological phosphorus removal mainly relies on the discharge of polyphosphate accumulating organisms in waste activated sludge (Saktaywin et al. 2005). What can be seen from Figure 3(d) is that loading of electromagnetic wave on returned sludge had an adverse effect on the TP removal. The effluent TP concentration of the original and LR = 30% systems was 0.5 ± 0.10 and 0.5 ± 0.17 mg/L. However, as the LR continued to increase, the phosphorus removal performance decreased more significantly. It was found that the LR = 60% system acquired the best sludge reduction (Y_sgr = 8.6 g SS/day) but the maximum effluent TP concentration (0.8 ± 0.17 mg/L). The reduction of phosphorus removal efficiency in the sludge reduction system based on lysis-cryptic growth is inevitable. Uan et al. (2013) also found that the TP removal of the A/A/O process decreased from 83–87% to 81–83% when the sludge was thermochemically pretreated.

Based on the comprehensive analysis of excess sludge production and pollutant removal performance under different electromagnetic wave loading conditions, the system with LR of 30% exhibited a relatively high total nitrogen removal capacity while achieving sludge reduction. Therefore, the original system (A/A/O) and LR = 30% system (named as EW-A/A/O) were selected for further study and analysis.

Table 1 reflects the release of COD, NH₄⁺-N, TN and TP in returned sludge and the variation of pollutant loading rate, which found that the concentrations of COD, NH₄⁺-N, TN and TP in the supernatant of returned sludge increased effectively. Previous studies revealed that sludge flocs and cells were destroyed under the action of electromagnetic fields (Yeneneh et al. 2015). Thus, organic matter in the sludge could be released into the liquid phase, then converted into soluble form that could be utilized by microorganisms (Appels et al. 2013).

The pollutant load of the system increased with the loaded sludge returning back to the biological system, which affected the degradation of organics by microorganisms. After calculation, the actual removal efficiency of COD, NH₄⁺-N, TN and TP in the A/A/O system was 85 ± 0.40%, 85 ± 0.63%, 63 ± 0.65% and 87 ± 0.34%, respectively. However, the actual pollutant removal efficiency of the EW-A/A/O system was 86 ± 0.36%, 92 ± 0.39%, 70 ± 0.16% and 89 ± 0.45%. Although the daily treatment amount of organic matter and nutrients has increased, the actual pollutant removal efficiencies (except for NH₄⁺-N) of the EW-A/A/O system were still higher than A/A/O, which was attributed to the strong utilization ability of heterotrophic bacteria in activated sludge for easily degradable organics. Previous studies (Sun et al. 2017) have also shown that the disintegration of these microorganisms in sludge released endogenous carbon resulted in sludge reduction. The quantity of TN removal increased with the amount of the organic carbon source and thus the abundance of sludge reduction increased accordingly. Therefore, the increased 4.6 g/d SCOD can be used as a carbon source to supply denitrifying electron donors in the denitrification process, which is beneficial to the denitrification effect of EW-A/A/O. As a result, the actual removal efficiency of TN increased by 7%.

With the purpose of evaluating the effects of electromagnetic wave loading on the microbial activity of sludge, the SOUR, endogenous respiration rate, and denitrification rate (ρ_NO3-N) were assessed in two systems. The average SOUR was 60.1 ± 5.3 and 68.3 ± 1.9 mg O₂/(g VSS·h) in A/A/O and EW-A/A/O processes, respectively. The increase in SOUR might be ascribed to a decrease in the transfer resistance of oxygen from the liquid phase to sludge solid phase, which was due to the loose structure of sludge flocs under the action of the electromagnetic wave (Yeneneh et al. 2015). Li et al. (2009) also obtained a similar result that SOUR of activated sludge increased owing to the stimulation of microbial activity within an appropriate ultrasonic density. At the same time, the phenomenon of elevated SOUR explained the fact that the EW-A/A/O system holds a higher pollutant removal efficiency. The endogenous respiration rates of aerobic activated sludge in the A/A/O and EW-A/A/O systems were 5.14 ± 0.25 and 5.11 ± 0.09 mg O₂/(g MLSS·h), respectively. This slightly reduced change also reflected that better nutrient conditions were provided to the microorganisms in the EW-A/A/O system, which was mainly due to the replenishment of the carbon source by the returned sludge decomposition solution.

The ρ_NO3-N in the anoxic activated sludge of the two systems was detected and analyzed. It was found that the ρ_NO3-N increased from 5.2 ± 0.4 to 6.7 ± 0.6 mg N/(g VSS·h), which indicated that the biological activity of the activated sludge in the anoxic tank was improved. This was because the organic matter dissolved in the returned sludge, which supplied the denitrifying carbon source in the anoxic tank, thereby increasing the activity of heterotrophic denitrifying bacteria. These results are very helpful for understanding the improvement of biological nitrogen removal.

The statistical results of diversity analysis of sludge samples gathered from the two systems are displayed in Table 2. The coverage indexes of each sample were both above 0.97 at a 3% distance, indicating that the coverage of gene sequence detection for each sludge sample was relatively high, and the sequences produced at this sequencing depth could well symbolize the actual bacterial communities. Diversity analysis results indicated that a profound influence in microbial community structure of the A/A/O system was induced with returned sludge loaded by electromagnetic wave. ACE and Chao1 indexes in Table 2 both denoted that the EW-A/A/O system exhibited greater microbial richness than A/A/O, which indicated that electromagnetic wave loading offered more substrates for the system by disintegrating the floc structure of returned sludge (Akgul et al. 2017). From the perspective of evenness, the slightly higher Simpson index suggested that the enriched OTUs in EW-A/A/O communities were distributed more unevenly than in the A/A/O system, which also meant that the distribution of dominant microorganisms has changed. Compared to R-1 sample in the EW-A/A/O system, OTUs, ACE and Chao indexes contained in L-1 sample were all smaller. This finding indicated that an electromagnetic wave could inactivate some species in returned sludge and reduce species abundance of the microbial community surviving under electromagnetic wave loading, which verified the existence of biological effects in the process of electromagnetic wave loading on sludge.

To recognize the influence on the bacterial community structure and composition with electromagnetic wave loading on returned sludge in the A/A/O system, the phylogenetic spectrums were assigned from phylum to genus level (Figure 4). In total, more than 30 bacterial phyla were recovered from the samples, of which Proteobacteria, Bacteroidetes, Planctomycetes, Firmicutes, and Acidobacteria dominated in all the community compositions (Figure 4(a)). Proteobacteria (53.6–68.3%) and Bacteroidetes (10.5–19.1%) were the most dominant phylum, which were reported as the commonly predominant phylum in activated sludge systems (Ma et al. 2013). Electromagnetic wave irradiation provided a sharp increase in the SCOD, and thus the abundance of these heterotrophic bacteria moved up because the dissolvable organic materials were available for anaerobic digestion. Planctomycetes was responsible for autotrophic ammonium oxidizing, the abundance of which was lower in EW-A/A/O (1.1–3.2%) than in A/A/O (6.2–8.9%). This was due to the fact that the released organic matter could replenish the organic carbon source of the heterotrophic denitrifying bacteria in the system, causing the limitation of the survival of the anammox bacteria (Tao et al. 2013). Figure 4(a) also showed an increase in the relative abundance of Firmicutes, which could resist some extreme environments (Filippidou et al. 2016). Phylum Actinobacteria was capable of a slower growth rate and carbohydrate removal, which was more abundant in A/A/O (2.4–4.3%) than EW-A/A/O (1.7–1.8%).

In order to further explore the influence on the microbial community structure of the system with electromagnetic wave loading on returned sludge, the microbial community distribution characteristics at the genus level was analyzed in the current study (Figure 4(b)). As shown in Figure 4(b), the dominant species of the bacterial community in EW-A/A/O system were quite different from those in the A/A/O system. Zoogloea (3.0–13.0%), Dechloromonas (6.0–11.5%) and Azospira (2.0–3.8%) were the most dominant genera in each module. As a typical bacterium that benefit cell aggregates during flocculation of active sludge, Zoogloea was also considered as denitrifying bacteria (Unz et al. 1967; Shinoda et al. 2004). As the returned sludge was loaded by electromagnetic wave, the sludge flocs became dispersed. In this process, Zoogloea became the first dominant genus (10.7–13.0%) in EW-A/A/O, and the repolymerization process of sludge flocs was also completed. Dechloromonas was widely found in wastewater treatment plants and played an important role in denitrification and phosphorus removal (Shinoda et al. 2004). It was the most abundant genera in A/A/O system (6.5–8.0%) and the second dominant genera in EW-A/A/O (6.0–11.5%). However, Dyella, with efficient desulfurization ability, was eliminated and disappeared in EW-A/A/O system on account of the weak resistance to electromagnetic wave.

Figure 5 reflects the transformations of SCOD, NH₄⁺-N, and NO₃⁻-N in every different unit. The concentration of SCOD in the anaerobic tank of A/A/O and EW-A/A/O was 45.2 ± 5.0 and 93.9 ± 5.3 mg/L, respectively. The rise of SCOD concentration represented the increase of organic load in EW-A/A/O, which was the main reason for the higher effluent COD concentration. However, the smaller variation of COD removal effect was mainly attributed to the enrichment of heterotrophic microorganisms (i.e. Dechloromonas, Azospira, Tolumonas, etc.).

In order to further illustrate the mechanism of improving the denitrification performance with returned sludge loaded by electromagnetic wave, the relative abundances of ammonia oxidizing bacteria (AOB), nitrite oxidizing bacteria (NOB) in the oxic tank and denitrifying bacteria (DNB) in the anoxic tank were counted. The results are shown in Figure 6.

The total relative abundance of AOB (i.e. Nitrosomonas and Prosthecobacter) was 0.77% in A/A/O and 0.75% in EW-A/A/O. As for NOB, the relative abundance of Nitrospira changed noticeably, declining from 3.1 to 0.4%. The decrease in relative abundance of AOB and NOB (from 3.9 to 1.1%) supposed that the selection pressure was provided in the condition of electromagnetic wave loading. Therefore, although it was detected that the aerobic activated sludge had a higher SOUR in the EW-A/A/O system, the reduction of AOB relative abundance weakened the degradation capacity for NH₄⁺-N. Accompanied with a higher NH₄⁺-N load in the EW-A/A/O system (Table 1), a lower actual NH₄⁺-N removal of 92% was caused.

Results presented in Figures 5 and 6 revealed that nitrogen removal efficacy corresponded to the relative abundance of DNB. As shown in Figure 5(c), lower NO₃⁻-N concentration was 4.8 ± 0.2 mg/L in the anoxic tank of the EW-A/A/O system, and 6.7 ± 0.3 mg/L in A/A/O, which had a great relationship with the enrichment of DNB. Zoogloea, Dechloromonas, Azospira and Thauera were reported to be important denitrifying bacteria (Unz et al. 1967; Shinoda et al. 2004; An et al. 2017; Sun et al. 2017), which have been well enriched in the EW-A/A/O system. It is worth noting that the Sulfuricurvum, taking nitrate as electron acceptor (Kodama & Watanabe 2004), existed as a new genus (1.1%) in EW-A/A/O. In Figure 5(a), the higher SCOD concentration (93.9 ± 5.3 mg/L) in the anaerobic tank of the EW-A/A/O system meant that more abundant substrates would enter the anoxic tank, which favored the growth of DNB (Figure 6) and further enhanced nitrogen removal (Figure 3(c)). Overall, the total relative abundance of DNB in the anoxic tank of A/A/O and EW-A/A/O was 18.5 and 33.3% respectively, indicating that electromagnetic wave loading on returned sludge can promote the evolution of the microbial community structure in the A/A/O system.

In general, a large amount of organic matter in the returned sludge would be released under the loading of electromagnetic wave, thereby supplementing the carbon source to the A/A/O system, which inhibited the survival of autotrophic AOB. In addition, the removal of NH₄⁺-N in wastewater treatment process could also be attributed to the synthesis of microorganisms, thus the reduction in sludge yield also limited the adequate removal of NH₄⁺-N in the system.

The organic matter released from the returned sludge promoted the enrichment of DNB, thus contributing to the improvement of the effectiveness of denitrification. At the same time, since the relative abundance of AOB was not significantly reduced, while the relative abundance of NOB decreased obviously, it could be considered that NO₂⁻-N would be enriched in the oxic tank, thus NO₂⁻-N accounted for a large proportion of the NO_x⁻-N that returned to the anoxic tank. It is well known that the organic content required for denitrification with NO₂⁻-N as the electron acceptor is lower than that of NO₃⁻-N. Therefore, from another perspective, it can be considered that the deterioration of the nitrification effect would reduce the carbon source content required in the denitrification process. As for the worsening nitrification effect, increasing the amount of aeration can be considered to solve this.

For the practical application of electromagnetic wave loading on the improvement of biological nitrogen removal in A/A/O, convenient operation and economic efficiency should be considered crucial. However, it should be noted that this is only a proof-of-concept study and is the first step in exploring the proposed method for improving biological nitrogen removal performance by electromagnetic wave loading on returned sludge. Therefore, in order to expand the application of electromagnetic wave in the sludge treatment field, comprehensive testing is needed to better evaluate this method in the future. For example, economic analysis of electromagnetic wave technology is necessary to discuss in detail for the improvement of practical application value.

This work presented the novel application of the 2,450 MHz electromagnetic wave for returned sludge in the anaerobic-anoxic-oxic process to improve the biological nitrogen removal effect. When 30% of the returned sludge was loaded by electromagnetic wave, the system had the potential to achieve better biological nitrogen removal with less sludge production. The release of organic matter in the returned sludge supplemented the carbon source for denitrification and increased the nitrogen removal efficiency by 7%. The enhancement of microbial activity and the enrichment of denitrifying bacteria illustrated the mechanism of the enhancement of the denitrification effect at a biological level.

This work was supported by National Natural Science Foundation of China (51108360, 51208397); Science and Technology Infrastructure Program in Hubei (2015BCA304); and Fundamental Research Funds for the Central Universities of China (2018-zy-059, 2019-zy-121).

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