1 Introduction

Aerobic granular sludge (AGS) is a novel biotechnology used to treat municipal and industrial wastewater in recent years [1, 2]. With its compact microbial structure, excellent settling capacity, and high biomass retention, it has been considered a promising technology for wastewater treatment in the coming future [3, 4]. Compared with the traditional activated sludge, AGS can simultaneously remove COD, phosphorus, nitrogen, and other pollutants in a single reactor [5]. Thus, it was estimated that AGS can reduce the operating costs of wastewater treatment by 20–25% [6, 7]. The feasibility of this technology has been widely used to treat urban domestic wastewater, and the aquatic, food, and textile industrial wastewater [8,9,10]. However, this technology still has problems such as long start-up time and easy destabilization of the AGS structure in a long-term operation [11, 12]. These problems are the main bottlenecks for the engineering application of AGS technology.

To solve these problems, extensive research has been conducted to accelerate granulation and enhance the stability of AGS. Among them, the addition of carrier media (nucleus) to aerobic sludge systems has been considered an effective method [13]. Based on the nucleus hypothesis, the formation process of AGS is similar to the crystallization process [14]. The nucleus acts as support media for microorganisms to attach onto their surface, eventually forming a mature AGS. It was found that the zero-valent iron can significantly reduce the time required for sludge granulation, and also improve the physical and chemical properties of the cultivated AGS [15]. The magnetic nanoparticles can also shorten the AGS formation time and enhance the denseness of the obtained AGS [16]. Besides, the zeolite powder was also demonstrated as an effective material to enhance the sludge granulation process and reduced the granulation time to 18 days [17].

Recently, an alternative material with a stable chemical property is required to further improve the granulation process. The conductive polymers have received increasing attention because of their special structure and excellent stable chemical properties [18,19,20,21]. Polyaniline (PANI) with the advantages of simple preparation, low cost, good bio-affinity, low toxicity, and good environmental stability has been widely used in industrial and biomedical fields [22,23,24,25,26]. The conductive PANI had also been added into the anaerobic wastewater treatment systems, and has been demonstrated to have good affinity with microorganisms [27]. In addition, PANI has a positively charged surface in the wastewater treatment system, which facilitates the attachment of microorganisms with a negatively charged surface. This in turn promotes the granulation process of sludge [28]. Therefore, the PANI is supposed as a promising nucleus material to accelerate the AGS granulation process.

In this study, the PANI was synthesized and was used to improve the granulation of aerobic sludge. In addition, the effect of the PANI on the morphology and EPS production of AGS was determined. Finally, the performance of the formed AGS on water treatment efficiency was investigated, and the effect of PANI on the bacterial community of the AGS was detected.

2 Material and methods

2.1 PANI synthesis

The PANI was synthesized by a chemical oxidation polymerization method [29]. Briefly, 3.2 mL aniline was dispersed in 200 mL HCl solution (54 mM) by magnetic stirring to get a homogeneous reaction system. Then, 8.22-g ammonium persulfate was dissolved in another 100 mL deionized water and added dropwise into the aniline solution. The reaction system was stirred continuously in an ice bath at 600 rpm for 4 h. After the polymerization, the bottle green slurry was obtained and washed with ethanol and deionized water by a filtration method until the filtrate was colorless. Finally, PANI samples were obtained after vacuum drying at 70 °C for 12 h.

2.2 SBR set-up and operation

Three lab-scaled sequencing batch reactors (SBR) with a working volume of 1.3 L and a height/diameter (H/D) ratio of 10 (60 cm of height and 6 cm of diameter), were inoculated with activated sludge. The activated sludge was collected from Xiaohongmen Wastewater Treatment Plant, Beijing, China, with a MLSS of approximately 5.3 g/L. Then, 0, 0.4, and 0.8 g/L PANI were added into the three reactors, named R1, R2, and R3, respectively. Each SBRs was operated with a 4-h operating cycle which contained five parts: the water inlet, anaerobic, aeration, settling, and the water outlet. The reactors were operated at room temperature for 33 days, and four phases were included in the entire experimental process (Table S1). The volumetric exchange rate of the three SBRs was set to 50% and the airflow rate was 2.5 L/min.

The SBRs were fed with the artificial wastewater with the composition (g/L) as follows: 0.76 ~ 1.09 NaAc (600 ~ 850 mg COD/L), 0.23 ~ 0.32 NH4Cl, 0.04 KH2PO4, 0.08 CaCl2·2H2O, 0.02 MgSO4·7H2O, and 0.01 peptone. One liter of the synthetic domestic wastewater contained 1 mL of trace element solution, which contained (mg/L): 1500 FeCl3·6H2O, 150 H3BO3, 30 KI, 120 MnCl2·4H2O, 30 CuSO4·5H2O, 58 ZnCl2, 150 CoCl2·6H2O, and 50 Na2MoO4·2H2O. The COD/N ratio was kept at 10 ~ 9:1.

2.3 Analytical methods

The water quality index (COD, NH4+–N, total nitrogen (TN), total phosphorus (TP)), and sludge index (MLSS and sludge volume index (SVI30)) was measured by the Standard methods [30]. The concentrations of protein (PN) and polysaccharide (PS) in the EPS were detected by Folin method and phenol–sulfuric acid method [31]. The particle size distribution of AGS in each SBR was determined by a Mastersizer3000 Malvern laser light scattering instrument. The morphology of AGS and PANI was observed by a JSM-6700F scanning electron microscope (SEM).

2.4 EPS extraction and measurement method

EPS of sludge was extracted by using a thermal treatment method [32]. In brief, 20 mL of sludge mixture was collected after 30 min of aeration, and centrifuged at 3000 rpm for 10 min and then at 4000 rpm for 3 min. Twenty milliliters of phosphate buffer (0.01 M) was added after decanting the supernatant. After mixing, the mixture was heated at 80 °C for 1 h and then was centrifuged at 10,000 rpm for 30 min to collect EPS solution. The EPS was measured by EEM fluorescence spectrometry using a Hitachi F-7000 spectrophotometer. The five regions of the EEM peaks (Table S2) were analyzed using the fluorescence region integral (FRI) technique. The percent fluorescence response (Pi, n, %) was calculated using the normalized Ex/Em area volumes (Фi,n, ФT,n) by following Eqs [33].

$$\Phi_{i,n} = MF_{i} \sum\limits_{ex} {\sum\limits_{em} {I(\lambda_{ex} \lambda_{em} )\Delta } } \lambda_{ex} \Delta \lambda_{em}$$
(1)
$$\Phi_{T,n} = \sum {\Phi_{i,n} }$$
(2)
$$P_{i,n} = {{\Phi_{i,n} } \mathord{\left/ {\vphantom {{\Phi_{i,n} } {\Phi_{T,n} \times 100}}} \right. \kern-\nulldelimiterspace} {\Phi_{T,n} \times 100}}\%$$
(3)

where \(\Delta \lambda\) ex and \(\Delta \lambda\) em were the \(\lambda\) ex wavelength interval and the \(\lambda\) em wavelength interval, respectively, which were both taken as 5 nm in this study. MFi was the multiplication factor and \({\text{I}}({\uplambda }_{{{\text{ex}}}} {\uplambda }_{{{\text{em}}}} )\) was the peak intensity (au) of each region.

2.5 Bacterial community analysis

Three AGS samples were collected on the 27th day from the three SBRs. The bacterial community was analyzed using the high-throughput sequencing method by the Illumina HiSeq platform of Allwegene (Beijing, China).

3 Results and discussion

3.1 Cultivation of AGS

It was found that AGS was successfully cultivated in all the three SBRs. Visible granules were firstly observed in the SBRs added PANI (R2 and R3) on the 4th day, while it was found in R1 on the 10th day during the operation. This indicated that PANI could reduce the granulation time of sludge and facilitate the formation of AGS. The changes of MLSS in the three reactors are shown in Fig. S1A. After inoculation, the MLSS in the three SBRs were 5.11 g/L (R1), 5.41 g/L (R2), and 5.35 g/L (R3), respectively. At the initial stage of the start-up period, the MLSS concentration decreased significantly due to the poor settling performance of inoculated activated sludge. Compared to R1, R2 and R3 had less sludge loss, probably because the positively charged surface of PANI provides abundant sites for the attachment of the microorganisms with negatively charged surface, avoiding the loss from the reactors (Fig. 1). From the 4th day, the MLSS in the three reactors increased, and R2 and R3 have lower MLSS than R1. The maximum MLSS in R2 and R3 reached 5.57 and 6.14 g/L, which was also lower than 6.45 g/L in R1. With a gradual reduction in settling time to 5 min (at day 17), the MLSS in R1 gradually decreased. A large amount of flocculated sludge was suspended in R1 at the end of the settling process on the 31st day, which indicated that sludge disintegration occurred in R1 (Fig. S3A). As shown in Fig. S3B, lots of rotifer-dominated post-zooplankton were found in the sludge of R1, which further indicated the imminent disintegration of the granular [34]. In addition, the foams were observed on the water surface in R1 on the 32nd day of incubation, while this phenomenon was not observed in R2 and R3, probably due to the proteins released from AGS into the water after disintegration. It was found that PS was mainly distributed in the outer structure of the granules, while PN was mainly present in the interior of the AGS [35]. The protein content in R1 was higher than that in R2 and R3 by 27.6% and 54.2%, further indicating the disintegration of AGS in R1.

Fig. 1
figure 1

Zeta potential of the PANI

Full size image

The SVI reflects the settling performance of sludge in the reactors. As shown in Fig. S1B&C, the SVI30/SVI5 of each reactor gradually increased to 1, indicating that the granulars in the reactors had good settling performance. The SVI30/SVI5 of R2 and R3 reached above 0.9 on the 8th day and then remained stable, while it reached above 0.9 in R1 on day 12 and then decreased on day 20, which indicated that the PANI could accelerate the granules formation to improve the settling performance and the stability of the AGS. At the early stage of cultivation, the SVI30 of each reactors increased and then decreased due to the poor settling performance of activated sludge. From day 4, the decrease of SVI30 in R1 was more obvious than that in R2 and R3. Finally, it reached a minimum of 45.79 mL/g in R1 on day 16, which was lower than 53.87 mL/g and 53.66 mL/g in R2 and R3, respectively. The granules formed in R1 have large size and low MLSS, which contributed to the good settling performance [36].

3.2 Morphology and size distribution of AGS

The SEM was used to observe the morphology of the PANI and the formed AGS. It was observed that the synthesized PANI has a rough surface (Fig. 2A), which facilitates the attachment of microorganisms on its surface [37]. As shown in Fig. S2, the morphology of AGS formed in the three reactors was almost similar, with a regular elliptical shape. It revealed that the mature AGS had a compact structure with a surface dominated by rod-shaped and spheric bacteria embedded in EPS (Fig. 2), which was consistent with the characteristics of AGS cultured with sodium acetate as substrate [38]. It was also observed that the AGS formed in R2 and R3 have more pores than that in R1. As documented in the literature, the small granule size leads to high porosity and large channels in the AGS [39]. These channels could diffuse the substrates and oxygen to the inner layer, thus reducing the space of the anaerobic layer. In contrast, the AGS in R1 had a dense structure, which inhibited the diffusion of the nutrients and the metabolites in the granules. This also inhibited the bacterial activity inner the granulars, thus reducing the long-term stability of the AGS [40].

Fig. 2
figure 2

SEM images of A the PANI, and formed AGS in B R1, C R2, and D in R3

Full size image

Figure 3 shows the size distributions of AGS at 24th and 31st day of incubation in all the SBRs. When the reactors were operated up to day 24, a significant increase in AGS size in all the three reactors compared to inoculation sludge. In addition, the AGS with the size larger than 0.9 mm accounted for 31.83% in R1, while 3.35% and 8.63% in R2 and R3, respectively (Fig. S4). On day 31, the average AGS size decreased in R1 compared to that on the 24th day due to the sludge disintegration. The AGS size increased in R2 and R3 compared to that on the 24th day. The granule size ranging from 0.5 to 0.9 mm accounted for 46.57% and 40.79% in R2 and R3. This indicated that the PANI was beneficial to the maintenance of the AGS. The percentage of AGS with a size bigger than 0.9 mm in R3 was obviously higher than that in R2, which also showed that the PANI could improve the granulation process of sludge. The positively charged surface of PANI could attract bacteria with the negatively charged surface and enhanced the microbial aggregation, resulting in a large AGS in R3 [41].

Fig. 3
figure 3

Size distribution of AGS of each reactor at A day 24, B day 31

Full size image

3.3 EPS content and composition of AGS

EPS are the substances secreted by microbes, mainly consisting of PN and PS, and play an essential role in the formation of AGS [42,43,44]. The variation in PN and PS content during AGS formation was evaluated in this study (Fig. 4). The content of PN in the three reactors firstly increased in the granulation period and gradually decreased in the mature period. It was found that the PN content of the AGS in R1 was lower than that in R2 and R3. For example, 94.67–129.95 mg/(g MLSS) of PN was detected in R1 during the AGS formation stage (5 to 10 days), which was lower than that in R2 (100.17–133.57 mg/(g MLSS)) and R3 (98.52–149.26 mg/(g MLSS)). It was documented that the increase in PN content at the granulation stage contributes to the aggregation of microorganisms [45, 46]. The added PANI increased PN production in the sludge, thus facilitating the granulation process. In contrast, the PS content in sludge gradually increased with the operation of the reactors. It was also found that R2 and R3 had higher PS/PN ratios than R1, and high PS proportion was beneficial in maintaining the stability of AGS in R2 and R3 [47].

Fig. 4
figure 4

PN and PS variation of AGS in A R1, B R2, and C R3

Full size image

Figure 5 shows the 3D–EEM spectra of EPS in the three reactors during the granulation process. Information of five different substances in EPS was evaluated by the FRI technique (Table S2). As shown in Fig. 6, tryptophan-like protein (regions II) and soluble microbial byproduct (regions IV) were detected as the domain fluorescence response (Pi, n) among the five \(\lambda\) ex/\(\lambda\) em regions, which accounted for 35 ± 5% and 30 ± 5%, respectively. In contrast, the intensity of the fulvic acid-like material (region III) had no significant changes throughout the operation. In particular, tyrosine-like proteins (region I) and tryptophan-like proteins are two hydrophobic substances that facilitate the maintenance of structural stability of AGS [48]. The Pi, n of the two types of substances in the three SBRs were 42.08 ± 6%, 45.43 ± 2%, and 44.33 ± 4% during the whole granulation process, indicating that the added PANI could promote the sludge for secretion of hydrophobic proteins to enhance the formation of AGS. It was also found that the Pi, n of region V in R1 was higher than that in R2 and R3. As documented, the humic-acid-like organics (region V) is a kind of hydrophilic substance, which is not conducive to microbial aggregation in wastewater biological treatment system [49].

Fig. 5
figure 5figure 5

3D-EEM spectra of EPS in the AGS in the SBRs at A 5th, B 10th, C 15th, D 20th, and E 25th day of incubation

Full size image
Fig. 6
figure 6

FRI analysis of EPS composition during the granulation process in A R1, B R2, and C R3

Full size image

3.4 Pollutants removal performance

The pollutants performance of the SBRs was also evaluated (Fig. 7). From the start-up stage, all the three SBRs showed good performance on COD removal (Fig. 7A). During phase I, the COD removal percentage decreased because the sludge was adapting to the substrates in feeding water. When the reactors were operated to phase II and III, the reduction of aeration time affected the activity of heterotrophic bacteria, resulting in a gradual decrease of COD removal. In order to maintain the F/M ratios at 0.4 ~ 0.5 g COD/g VSS·day, COD concentration in the influent increased from phase IV (Fig. S5). It was reported that the microbial diversity in AGS at this sludge load was high and the pollutant removal was effective [50]. At the initial time of phase IV, the COD removal performance of the three reactors was stable and remained above 93%. R2 and R3 had better performance than R1, which was probably due to the biomass loss caused by sludge disintegration in R1. There was no obvious difference between the three reactors in terms of the NH4+–N removal (Fig. 7B). After 7 days of incubation, the NH4+–N concentration in the effluent of all reactor was below 1 mg/L, which means that more than 99% of NH4+–N could be transferred by AGS. As shown in Fig. 7C, the SBRs had poor removal effect on TN during the granulation period because most of the microorganisms were exposed to an aerobic environment before granules formation, which inhibited the activity of denitrifying bacteria. With the granulation of sludge and the increase of AGS size, the TN removal in all the reactors obviously increased. In the late IV stage of operation, the TN removal percentage of R1, R2, and R3 was 58.73%, 56.35%, and 58.83%, respectively. It was inferred that the large granule size of AGS led to a high TN removal due to good anoxic condition inner the AGS. As shown in Fig. 7D, the not significant difference in TP removal was observed in the three reactors throughout the operation, and the TP removals were up to 65.56% (R1), 69.54% (R2), and 70.86% (R3).

Fig. 7
figure 7

A COD, B NH4+–N, C TN, and D TP removal by R1, R2, and R3

Full size image

3.5 Bacterial community in AGS

Table S3 summarizes the abundance and diversity of the bacterial community in the SBRs. Goods coverage reached 1 in all three AGS samples, revealing that the bacterial diversity of the samples could be covered well [51]. The Chao1 index of the AGS samples was 202.55, 199, and 197.25, indicating that the added PANI had a little effect on the abundance of the bacterial community. The similarities and differences of microbial species in the three AGS could be assessed by the Venn diagram (Fig. S6). 221 OTUs were detected for the AGS samples, in which 172 OTUs were the same, which indicated a close relationship between the microorganisms in the AGS.

At the phylum level (Fig. 8A), Proteobacteria (73.89% in R1, 68.44% in R2, and 73.71% in R3) and Bacteroidetes (23.76% in R1, 28.89% in R2, and 23.11% in R3) were the two dominant phyla in the three AGS samples. Proteobacteria and Bacteroidetes are reported to be the commonly shared phylum and are also considered to be the widespread nitrification–denitrification phylum in biological wastewater treatment processes [52,53,54]. Other prominent phylum, such as Verrucomicrobia (0.80%), Actinobacteria (0.54%), Patescibacteria (0.33%), and Chloroflexi (0.10%), were also detected in AGS in R1. In contrast, Verrucomicrobia (0.57% and 0.47%), Actinobacteria (0.56% and 0.65%), Patescibacteria (0.95% and 0.68%), and Chloroflexi (0.12% and 0.13%) were detected in the AGS of R2 and R3, respectively. The increase in the abundance of these phyla may contribute to the granulation in R2 and R3 [55, 56]. In particular, Actinobacteriota could maintain the structure of sludge, and Chloroflexi could enhance the granule structure. At the class level (Fig. 8B), γ-proteobacteria and Bacteroidia were the predominant classes, accounting for 93.08%, 93.46%, and 93.21%, respectively. γ-proteobacteria, a member of the Proteobacteria phylum, contributes to the stability of aerobic granules and has the ability to oxidize ammonia, which contributed to the excellent performance on ammonia nitrogen removal by the SBRs [57]. Bacteria belonging to the classes α-proteobacteria, Verrucomicrobiae, Polyangium, Blastocatellia were also detected in the three samples, but no significant change was observed for the relative abundance.

Fig. 8
figure 8

Bacterial communities of AGS at A phylum level, B class level, and C genus level in the SBRs

Full size image

Figure 8C shows the 20 top abundant bacterial genera in the three AGS samples to evaluate the changes in functional bacterial composition. The dominant genus detected in three samples were Thauera, Diaphorobacter, Acinetobacter, Ferruginibacter, Luteimonas, Comamonas, Alicycliphilus, and Dechloromonas. In detail, Thauera is a key bacterial genus in AGS with the ability on degrading various organic substances and denitrification [58, 59]. Diaphorobacter and Ferruginibacter are also genera with denitrification ability, among which Diaphorobacter is related to the accumulation of nitrite in the aqueous environment [60]. It could be found that the relative abundance of these genera in AGS was affected by the added PANI, and the bacteria related to the denitrification process were enriched in R2 and R3. In addition, the genus Dechloromonas is a denitrifying polyphosphate accumulating bacteria and a slow-growing microorganism, which facilitates the complete granulation [61, 62]. Previous studies have shown that slow-growing organisms have a positive effect on the stability of AGS [63, 64], which contributed to the high stability of the AGS formed in R2 and R3.

4 Conclusions

In this study, the PANI material was added to promote the sludge granulation and improve the stability of the formed AGS. The results showed that PANI with a positively charged surface could act as the support medium for attachment of the microorganisms in activated sludge, facilitating the granulation process. It was found that the added PANI can shorten the granulation time. PANI can control the AGS size and promote the secretion of PS, which facilitates the long-term stability of the AGS. The formed AGS has a good performance on removal of COD, NH4+-N, TN, and TP from wastewater. In addition, bacteria with denitrification ability were enriched in the AGS by the added PANI, which improved the denitrification performance of AGS.