Effects of ferric nitrate additions under different pH conditions on autothermal thermophilic aerobic digestion for sewage sludge

Abstract

The effects of ferric nitrate additions at different pH values on stabilization of sewage sludge and microbial communities were investigated in autothermal thermophilic aerobic digestion (ATAD). The lowest pH value but highest VS removal was achieved at optimal pH of 6.5 when Fe(NO₃)₃ was added, compared to other dosing groups, and the stabilization time was shortened by 7 days. The increased dosing pH reduced the effectiveness of Fe(NO₃)₃ on disinhibition of excessive volatile fatty acids, and even caused failure of stabilization. However, highly basic dosing conditions (such as pH 9.5) could restore the ATAD ability by promoting the processes of hydrolysis and acidification, and then take back control of the pH. Illumina high-throughput sequencing revealed the group at optimal dosing pH significantly enhanced the abundance of phylum Firmicutes (from 61.0% to 96.6%), while the rise of dosing pH decreased the richness of phylum Firmicutes. The structure of the microbial community was totally changed under strong basic conditions of Fe(NO₃)₃ addition.

---

1 Introduction

Biological processing of wastewater will inevitably produce large amounts of sludge, and the price of treatment and disposal of the sludge could account for up to 50% of a wastewater treatment plant’s (WWTP) operating cost.¹ Due to the lower investment and simpler control requirements compared to the technology of anaerobic digestion, aerobic digestion process is much more suitable for small- and medium-sized WWTPs, especially the autothermal thermophilic aerobic digestion (ATAD).² ATAD process can meet class A biosolids demand for a wide range of organic sludge, such as animal sludge, sewage sludge and food processing wastes, and significant advances of optimization and adaptation of ATAD technology have been achieved.³

Despite the superiority presented by ATAD,³ however, recently inhibition of excessive volatile fatty acids (VFA) had been found to be a severe issue in the initial stages of the ATAD process for sludge with high volatile solid (VS).⁴ A chemical method had been devoted to relieving the inhibition caused by over-high VFA concentration in the ATAD system, through a complex-precipitation process involved in reactions between sludge components and ferric nitrate.⁵ Although there have been some effort to reveal the influences of the factors related to the agent itself on the ATAD for sludge, for instance, the dosage and timing of ferric nitrate additions,^6,7 there are still lots of difficulties unsolved, such as effects of pH, temperature and so on. As a chemical reaction, the crucial impact of pH on the product and the reaction rate is obvious. In addition, the involvement of hydroxyl in the reaction of the complex will make the pH conditions more significant as well.⁵

On the other hand, the importance of pH to the biological processes in ATAD is evident. The most immediate response to the change of pH was the fluctuations of VFA levels in the procedures of hydrolysis and acidification,⁸ as well as the variations of dominant VFA species.⁹ Especially under thermophilic conditions, the VFA concentrations changed drastically at different pH values.¹⁰ Furthermore, the optimal pH was divergent in different systems. It was found that the optimal pH was 10 for VFA production from excess sludge,¹¹ through the break of sludge matrix which increased the effective contact between extracellular organic matters and enzymes, and created a favourable environment for microbes to accumulate VFA.¹² In the presence of alkyl glycoside, short chain fatty acid productions from membrane bioreactor sludge at initial alkaline pH values were also more efficient than those at acidic and near-neutral pH conditions, but the optimum initial pH was 11.¹³ A pH of 12 was adopted to obtained continuous volatile fatty acid production from waste activated sludge as well.¹⁴ In addition, hydrolytic enzymatic assays demonstrated that abiotic effect was responsible for increased solubility of organic matter in sludge under high alkaline condition.^1,12 Nevertheless, VFA accumulation was optimized at pH 8 despite higher solubility at higher pH, and the pH at 9 and above would lead to disruption of biological activities and VFA production eventually.¹ Meanwhile, some research also indicated that the alkaline pH improved the solubility and biodegradation of proteins in the sludge, and significantly influenced the biodiversity and bacterial community in the system.¹⁵ Lastly, the pH level had a close relationship with the complexing state of metalorganics, particularly the iron organic complex, as well as their migration and transformation in aerobes.¹⁶

As we know, the initial stage of the ATAD system was much like a facultative anaerobic environment, in view of the limited aeration as well as the low oxidation–reduction potential (ORP).¹⁷ The survival of anaerobic microbes in ATAD also supported this argument.¹⁸ Hence, the stress of pH changes to anaerobic microbial metabolism, especially the hydrogen production, should also be taken into consideration.^19–21 Besides, the pH condition affected the ammonia stripping²² and phosphorous release^23,24 as well. Therefore, the effects of ferric nitrate additions at different pH values on ATAD performance for sewage sludge were investigated in this study. At the same time, the corresponding microbial communities at different pH values were compared and analysed.

2 Experimental

2.1 Start-up of the experiments

Sewage sludge used in this study was secondary sludge collected from a WWTP in Shanghai, China. This WWTP owns an anaerobic–anoxic–oxic process with a capacity of 45 000 m³ wastewater treatment daily. The sludge sample was screened to remove particles coarser than 0.5 mm before going through centrifugation at 2200 g for 3 min to obtain total solid (TS) between 5% and 6%. The main properties of raw sludge are shown in Table 1.

Table 1 Properties of initial sludge employed in simulated one-stage ATAD process^a

Parameter	pH	TS (g L^−1)	VS (g L^−1)	SCOD (mg L^−1)	TN (mg L^−1)	NH4^+–N (mg L^−1)	TP (mg L^−1)
a SCOD, soluble chemical oxidation demand; TN, total nitrogen in supernatant; TP, total phosphate in supernatant.
Value	6.5 ± 0.2	54.8 ± 0.4	36.0 ± 0.3	1205 ± 20	221 ± 10	35 ± 1	182 ± 8

---

The batch digestions were conducted in five simulated autothermal thermophilic aerobic digesters of 200 mm (D) × 400 mm (H), and the available volume of a cylinder reactor was 4 L. The self-heat process was imitated through a water bath connected to the heating water jacket outside the body of the digester, and the temperature of digestion was rising from 35 °C to 55 °C at a rate of 5 °C per day. After the temperature reached 55 °C, the digestion process would stay at this temperature until the end. A continuous aeration rate of 0.13 L min⁻¹ and a constant stirring rate of 120 resolutions was supported.¹⁷

The entire process of digestion last 21 days. Fe(NO₃)₃ was added to the reactors on the 6^th day⁷ with dosage of equal to reducing 1000 mg L⁻¹ of acetic acid,⁶ but the control one was not fed. The decreased amount of sludge by sampling before the 6^th day was taken into consideration when the chemical reagent was added into the reactors on the 6^th day. Fe(NO₃)₃ was put into the digester 6 hours before sampling on the 6^th day, for the sake of adequate reaction between chemical reagent and sludge. The pH of the four systems were adjusted to 6.5, 7.5, 8.5 and 9.5 just after the additions of Fe(NO₃)₃ by using of NaOH and HCl, considering that the pH of a well operating ATAD system is always between 6 and 9, and sometimes can reach up to 9.5,²⁵ while the control was still not given any treatment. Samples were taken on the 4^th, 6^th, 8^th, 11^th, 14^th, 17^th, 21^st day and the start of the digestion, and microbial communities were analysed in raw sludge and digestion sludge on the 21^st day.

2.2 Chemical analysis

The pH of sludge was measured using a pH meter (pHs-3C, Leici Co. Ltd., Shanghai). VS and TS were determined according to Standard Methods²⁶ with values of influence caused by chemical reagents subtracted. The sampled sludge went through centrifugation at 12 000 g for 5 min followed by filtration through a 0.45 μm mixed cellulose ester membrane to obtain the supernatant. Then the values of NH₄⁺–N, SCOD, TN and TP in the supernatant were analysed on the basis of Standard Methods.²⁶ As for the determination of VFA level in the supernatant, the filtrate was blended well with 3% H₃PO₄ (to keep the pH of the filtrate staying at approximately 4.0) and then injected into a Shimadzu GC-2010 gas chromatograph with a flame ionization detector and DB-FFAP column (30 m × 0.25 mm × 0.25 mm) according to a method by Chen et al.⁸ The value of VFA was expressed in terms of COD. All of indicators were measured in triplicate and the standard deviations were acquired. The software SPSS version 19.0 for Windows (SPSS, IBM) was applied for statistical analysis and statistically significant correlations were decided at a 95% confidence interval (P < 0.05; Tukey’s test).

2.3 Molecular biological analysis

The five DNA samples of sludge after 21 days digestion were sent for Illumina Miseq sequencing of V3–V5 region of bacterial 16s rRNA gene (Genewiz, South Plainfield, USA). The software of Trimmomatic (v 0.30) was applied to the optimizing process of Illumina raw reads. The barcodes and primers were trimmed off. The quality values less than twenty on both ends were removed as well as the bases of average quality less than twenty. Every two sequences were compared and joined using the tool of pandaseq to filter out the reads containing nitrogenous bases and the sequences shorter than 400 bp or longer than 480 bp in the overlap region of the end. Afterwards these sequences were optimized, removing the sequences outside the target region by comparing with the 16s rDNA sequences in Greengene database and wiping off chimera sequences. Taxonomic classification was carried out to classify the optimized sequences into operational taxonomic units (OTUs), with a confidence threshold of 97% using QIIME standard pipeline.²⁷ To compare the microbial diversity and richness, Shannon rarefaction curves, clustering tree and species distribution diagrams of multi-samples were covered.

3 Results and discussion

3.1 Change of pH

The changes of pH values in five digestion systems are illustrated in Fig. 1. Few differences of the pH values were observed among all digesters in the first 4 days (P < 0.05). Then the variations of pH values in ATAD of Fe(NO₃)₃ dosed at pH 6.5, 7.5 and the control had more or less the same tendency after the ferric nitrate additions, while the other two were much similar. The former three curves had a sharp increase from the 6^th day to 14^th day before fluctuating moderately until the end, because of the consumption of VFA by precipitation with Fe(NO₃)₃ and increase of alkaline substances releasing.⁵ However, the other two systems had a fall from the 8^th day to 12^th day after an increase on the 6^th day, and went back to normal as the others afterwards.

Fig. 1 Changes of pH values in simulated one-stage ATAD systems.

The reduction in pH should be due to the increase of VFA, NH₄⁺ and cations, especially the VFA, which results from the acceleration of cell rupture and promotion of solubility of organic matters under alkaline conditions.^12,28 As for the faster growth of pH 8.5 compared to pH 9.5 afterwards, it should be attributed to the optimum pH 10 (closer to 9.5) in hydrolysis and acidification, which would make contribution to the control of pH and even the digestion later.¹² On the other hand, an out of control pH would lead to the failure of the ATAD system when Fe(NO₃)₃ is added at pH 8.5, and would result in a higher pH as well. As seen in Fig. 1, the range of pH changes were all between 6 and 9 except the one dosed at pH 8.5, agreeing with the domain of well operation ATAD systems’ pH,²⁵ and also supported the design of pH set in this study. The pH of the control got the lowest level throughout the digestion, while the pH of ATAD process with Fe(NO₃)₃ addition at pH 8.5 reached as high as nearly 10 in the end.

3.2 Removal efficiency of VS

VS removal always plays an important role in stabilization of sludge for aerobic digestion. The VS removals of ATAD processes for sewage sludge are shown in Fig. 2. There were few differences in the VS removal values in all reactors in the first 6 days (P < 0.05), and there was some delay compared to the changes in pH values. Afterwards, there was a plateau in either curve at pH 8.5 or pH 9.5, while the others increased continuously until the end. As seen in the Fig. 2, the ATAD systems had all stabilized for sewage sludge (>38%) after 21 days digestion except the group at pH 8.5.²⁹ As for the failure in stabilization of the ATAD process with Fe(NO₃)₃ addition at pH 8.5, the loss of control of pH should be the main reason. The most efficient VS removal was obtained in the group at pH 6.5, and its final VS removal was 45.16%, which was 6.42% more compared to that of the control. After a lag phase, the VS removal in the group at pH 9.5 had sharply increased and exceeded that of the control finally, indicating the advantageous pH condition had played a part in the digestion as mentioned above. Nevertheless, the stabilization time obtained by the group at pH 9.5 was almost the same as that of the control, but still lagged behind those achieved by the groups at pH 6.5 and pH 7.5 (reached removal of VS above 38%), 7 days and 4 days behind, respectively. Given the pH in the control was only a little higher than 6.5 on the 6^th day, the addition of NaOH or HCl for adjustment of pH was very small. Thus, the pH of 6.5 at which Fe(NO₃)₃ was dosed was certainly the right choice. These results demonstrated that the effect of ferric nitrate played a leading role on enhancement of ATAD performance rather than the pH at which the ferric nitrate was added, and the optimal pH was 6.5.

Fig. 2 Variations of VS removals in simulated one-stage ATAD systems.

3.3 SCOD and VFA in the supernatant

The variations of soluble chemical oxygen demand (SCOD) in the supernatant of five digesters are presented in Fig. 3A. The change of SCOD concentration in the supernatant of the control was in line with the trend revealed by Liu et al.² The lowest SCOD level was obtained by the group at pH 6.5 throughout the digestion. As illustrated in Fig. 3A, the SCOD of all reactors was almost the same in the first 4 days (P < 0.05). Then the SCOD in the group at pH 6.5 decreased until the end after a plateau on the 6^th day. However, the SCOD value in the group at pH 8.5 had a reduction when Fe(NO₃)₃ was added, and then increased to the maximum on the 11^th day. Afterwards, it fluctuated moderately to the end, agreeing with the response of VS removal, indicating the microbial activity had been inhibited under the conditions of losing control of the pH.¹ Unlike the SCOD value in the group at pH 8.5, the SCOD concentration in the group at pH 9.5 still increased as that in the control on the 6^th day, though the microbial activity was suppressed (reflected by the plateau of VS removal on the 8^th day) when the dosing pH was adjusted at 9.5,¹ implying the acceleration of hydrolysis and acidification had made up the lost SCOD value by inhibition of microbes. After a fluctuation, the SCOD level in the group at pH 9.5 rose again, just like the reflection of VS removal.

Fig. 3 Changes of (A) SCOD; (B) TVFA (total VFAs); (C) individual VFA of the control; (D) individual VFA of pH 6.5; (E) individual VFA of pH 8.5; (F) individual VFA of pH 9.5 in the supernatant of simulated one-stage ATAD systems.

The total volatile fatty acids (TVFA) included acetic acid, propionic acid, n-butyric acid, isobutyric acid, n-valeric acid and isovaleric acid. As shown in Fig. 3B, the changing trend of TVFA concentration was roughly similar to that of the SCOD value, respectively, in accordance with the internal relationship between TVFA and SCOD reported by Liu et al.¹⁷ The TVFA level in the group at pH 6.5 was very low (<1000 mg COD L⁻¹) after digestion of 14 days, denoting the stabilization had been achieved in another respect. As for the TVFA value in the group at pH 8.5, although the level of TVFA was almost undetected after the 14^th day, the fact that the increase of VS removal nearly stopped and the fluctuation of SCOD demonstrated that the metabolism of microbes were restrained.

The variations of individual VFA in the supernatant are illustrated in Fig. 3C to F. As seen in Fig. 3C, either the concentration of propionic acid or content of n-butyric acid was larger than that of isovaleric acid during the whole digestion process. However, the level of isovaleric acid turned out to be higher after Fe(NO₃)₃ addition (as shown in Fig. 3D and E), and the concentration of propionic acid was much less than 1000 mg COD L⁻¹, indicating that disinhibition of propionic acid had been achieved by relieving the stress of acetic acid through precipitation,⁵ and then promoting n-butyric acid degradation and transformation to shorter VFA.³⁰ On the other hand, the pH increased by Fe(NO₃)₃ addition and adjustment of itself also had a strong impact on the distribution of individual VFA, such as increase of isovaleric acid and decrease of propionic acid.¹⁵ Lower pH conditions revealed to be more favourable to accumulation of n-butyric acid as well.³¹ Nevertheless, the variations of individual VFA in the group at pH 9.5 (as seen in Fig. 3F) demonstrated that the key role played in promotion of microbial metabolism was the pH condition in this system, rather than the Fe(NO₃)₃ addition, in view of the abovementioned relatively higher concentration of isovaleric acid and lower content of n-butyric acid as well as the second highest level of propionic acid (>1000 mg COD L⁻¹)⁵ throughout the entire digestion process.

3.4 NH₄⁺ and TN in the supernatant

The change of pH value is a macroscopic phenomenon of variations in acidic and alkaline substances. The NH₄⁺ concentration in the supernatant of the control was the highest during the whole digestion process (as seen in Fig. 4A), and the TVFA level in the control was also the maximum, both of which contributed to the lowest pH value.²⁸ On the other hand, the concentration of NH₄⁺ was closely related to the content of TVFA (as shown in Fig. 3B and 4A). Because the main constituents of buffer system in the ATAD were NH₃–NH₄⁺ and NH₃–VFA, the level of NH₃ in the supernatant would decrease when TVFA was reduced, which would in turn lead to the reduction of NH₄⁺. Therefore, the lowest TVFA concentration in the group at pH 8.5 caused the lowest NH₄⁺ content after the 8^th day, and the NH₄⁺ level in the group at pH 9.5 rose on the 11^th day when the TVFA concentration increased. The other digesters’ situations were very similar. Additionally, the introduction of OH⁻ into the systems had also made the NH₄⁺ level decrease (as presented in Fig. 4A) because of the following reaction: NH₄⁺ + OH⁻ ↔ NH₃·H₂O.

Fig. 4 Variations of (A) NH₄⁺ and (B) TN in the supernatant of simulated one-stage ATAD systems.

The changes of TN in the supernatant of five systems are illustrated in Fig. 4B. As we know, NH₄⁺–N was the dominant component of TN in the supernatant, due to nitrification and denitrification were inhibited under thermophilic conditions.¹⁷ Hence, the tendency of variation in TN was very similar to that in NH₄⁺–N. Nevertheless, the lowest level of TN in the supernatant was obtained by the ATAD process of Fe(NO₃)₃ dosed at pH 6.5, but not the one of Fe(NO₃)₃ added at pH 8.5, which was different to the situation of NH₄⁺–N. As seen in Fig. 4B, the concentration of TN in group at pH 8.5 rose continuously to the maximum on the 11^th day, while the content of NH₄⁺–N started to reduce from 8^th day (as shown in Fig. 4A), indicating a large amount of organic amines were not degraded, considering that the microbial activities had been inhibited as mentioned above. The higher TN level in the group at pH 9.5 compared to that in the group at pH 7.5, which was contrary to the conditions of NH₄⁺–N, supported the presence of mass of organic amines in the supernatant as well.

3.5 TP in the supernatant

The variations of TP in the supernatant are shown in Fig. 5. The ATAD system was a closed system for resource of phosphorus,² so the TP in the supernatant was derived from the degradation of microbial cells and substances, and reduced by the metabolism of microbes,¹⁷ as well as precipitation and adsorption by ferric nitrate.³² As seen in Fig. 5, the TP level in the control was highest during the digestion except for the TP value in the group at pH 8.5 after the 14^th day, due to death and lysis of massive microbial cells under over-high pH conditions. The concentration of TP in the group at pH 9.5 was fluctuating moderately throughout the entire digestion process, which should be attributed to the acceleration of hydrolysis and acidification as mentioned above. The TP values in the groups at pH 6.5 and pH 7.5 kept pace with each other, and increased constantly until the end after a decline when Fe(NO₃)₃ was dosed on the 6^th day, which differed from the results in a previous study.⁵ This difference should be due to the adjustment of pH conditions when Fe(NO₃)₃ was added as hydroxyl is a strong competitor to phosphate while reacting with ferric.

Fig. 5 Changes of TP in the supernatant of simulated one-stage ATAD systems.

3.6 Microbial community

After removal of low-quality readings using the QIIME standard pipeline, a range of 118 420 to 264 538 high-quality readings (average length of 264 bp) was obtained from six sludge samples, with a high coverage of bacteria sequences (ranging from 0.95 to 0.96). The diversity curve became flat after 10 010 readings, and each of the six samples had many more sequences than 10 010, which perfectly shows the community diversity (as illustrated in Fig. 6A). The largest total number of operational taxonomic units (OTUs) estimated by Chao1 estimator with infinite sampling was observed in raw sludge (1346), and that of the control on the 21^st day second (1046), indicated that the addition of ferric nitrate had a negative impact on the richness of bacteria phylotypes, though the stress environment in ATAD had the ability as well.³³ The total number of OTUs in the group at pH 8.5 was the lowest (414), implying that the lowest richness of microbes was maintained. The other three were 702 for the group at pH 6.5, 637 for the group at pH 7.5 and 567 for the group at pH 9.5, respectively, denoting the increase of dosing pH also had a reducing effect on abundance of bacteria phylotypes. As seen in Fig. 6A, the raw sludge had the highest diversity (Shannon = 8.34), and the control ranked second (6.47) while the group at pH 8.5 ranked third (5.28). The highest diversity but lowest richness was achieved in the group at pH 8.5 compared to the other three dosing groups, contributing to the lowest microbial activities as mentioned above. Furthermore, the microbial community in the group at pH 8.5 was the closest in composition compared to raw sludge (as shown in Fig. 6B), demonstrating that the level of stabilization in the group at pH 8.5 was lowest of all. As for the other three dosing groups, the group at pH 6.5 had the highest diversity among them (5.07), and the one at pH 9.5 had larger a Shannon index (4.74) than the group at pH 7.5 (4.29), indicating that a strong alkali environment would increase diversity of aerobic microbes to a certain extent, considering that the abundance of anaerobic fermentative microorganisms under alkaline or acidic pH conditions was less than that under neutral pH condition.¹⁵ The microbial community in the group at pH 7.5 was closer in comostition with that in the group at pH 9.5 than that in the group at pH 6.5 (as presented in Fig. 6B), denoting the alkaline pH conditions had an influence on structure of microbial community.

Fig. 6 Results of Illumina high throughput sequencing: (A) Shannon rarefaction curves; (B) clustering tree; (C) taxonomic compositions of bacterial communities at phylum level in the simulated one-stage ATAD systems. Relative abundance was defined as the number of sequences affiliated with that OTU divided by the total number of sequences per sample. The relative abundance of phyla less than 1.0% of total composition in the three libraries was defined as “other”.

The species distribution diagrams of six sludge samples at phylum level are illustrated in Fig. 6C. After a period of 21 days digestion, the abundance of phylum Firmicutes increased in all digesters except for that in the group at pH 9.5. Phylum Proteobacteria was the dominant bacterium (77.6%) in the group at pH 9.5, indicating that the structure of microbial community had been totally changed under strong alkali conditions when Fe(NO₃)₃ was added. That is why the diversity in the group at pH 9.5 was higher compared to that in the group at pH 7.5, while the diversity in the group at pH 7.5 was observed reduced compared to that in the control. As seen in Fig. 6C, the composition of the microbial community in the group at pH 8.5 was similar to that in the raw sludge, supporting the closet relationship of microbial communities between them among the five groups as mentioned above. The abundance of phylum Firmicutes in the digestion sludge (except for the group at pH 9.5) had increased a lot compared to that in raw sludge, indicating a restricted phylum distribution into the Firmicutes, who played an important role for thermophilic aerobic degradation of waste sludge.^18,33 Especially in the group at pH 6.5, the abundance of phylum Firmicutes had reached as high as 96.6%, demonstrating the great impact on selection and reinforcement of bacteria phylotypes by ferric nitrate. Additionally, the increase of pH had an effect of inhibition of the phylum Firmicutes as the abundance of these bacteria phylotypes decreased in the group at pH 7.5 as well. This finding was opposite to the results revealed by Piterina et al.,¹⁸ which should be due to the combined action of ferric nitrate and the alkaline environment.

4 Conclusion

The addition of ferric nitrate at different pH values had distinct effects on stabilization of sewage sludge in a one-stage ATAD system. The increase of pH when ferric nitrate was added had a negative influence on promotion of stabilization by ferric nitrate, while a strong alkaline environment accelerated the hydrolysis and acidification, and even totally altered the structure of the microbial community, which restored the digestion process. The optimal pH of ferric nitrate addition was 6.5, at which the abundance of phylum Firmicutes was significantly enhanced. However, a moderate increase of pH would inhibit phylum Firmicutes and even cause failure in ATAD performance.

Acknowledgements

This study was financially supported by the National Hi-Tech Research and Development Program of China (863) (No. 2011AA060906), the Key project of Science and Technology Commission of Shanghai Municipality (No. 14DZ1207306), SMC “Chenxing” project by shanghai jiaotong University (2014) and Open Funding Project of National Key Laboratory of Human Factors Engineering, Grant No. HF2012-K-0.

References

1. Y. Maspolim, Y. Zhou, C. H. Guo, K. K. Xiao and W. J. Ng, Bioresour. Technol., 2015, 190, 289–298.
2. S. G. Liu, N. W. Zhu and L. Y. Li, Chem. Eng. J., 2011, 174, 564–570.
3. N. M. Layden, H. G. Kelly, D. S. Mavinic, J. Bartlett and R. Moles, J. Environ. Eng. Sci., 2007, 6, 665–678.
4. D. Hayes, L. Izzard and R. Seviour, Syst. Appl. Microbiol., 2011, 34, 127–138.
5. N. B. Jin, B. Jin, N. W. Zhu, H. P. Yuan and J. B. Ruan, Bioresour. Technol., 2015, 175, 120–127.
6. N. B. Jin, Y. W. Shao, H. P. Yuan, Z. Y. Lou and N. W. Zhu, Chem. Eng. J., 2015, 265, 9–15.
7. N. B. Jin, Y. W. Shao, J. Zhu, Z. Q. Shou, H. P. Yuan and N. W. Zhu, RSC Adv., 2015, 5, 43949–43955.
8. Y. G. Chen, S. Jiang, H. Y. Yuan, Q. Zhou and G. W. Gu, Water Res., 2007, 41, 683–689.
9. D. Cysneiros, C. J. Banks, S. Heaven and K. A. Karatzas, Bioresour. Technol., 2012, 123, 263–271.
10. X. G. Liu, B. Dong and X. H. Dai, Bioresour. Technol., 2013, 148, 461–466.
11. W. G. Jie, Y. Z. Peng, N. Q. Ren and B. K. Li, Bioresour. Technol., 2014, 152, 124–129.
12. G. H. Yu, P. J. He, L. M. Shao and P. P. He, Water Res., 2008, 42, 4637–4644.
13. J. W. Zhao, Q. Yang, X. M. Li, D. B. Wang, H. X. An, T. Xie, Q. X. Xu, Y. C. Deng and G. M. Zeng, Int. Biodeterior. Biodegrad., 2015, 104, 283–289.
14. X. Yang, C. L. Wan, D. J. Lee, M. A. Du, X. L. Pan and F. Wan, Bioresour. Technol., 2014, 168, 173–179.
15. H. Liu, J. Wang, X. L. Liu, B. Fu, J. Chen and H. Q. Yu, Water Res., 2012, 46, 799–807.
16. D. J. Kosman, Coord. Chem. Rev., 2013, 257, 210–217.
17. S. G. Liu, N. W. Zhu and L. Y. Li, Bioresour. Technol., 2012, 104, 266–273.
18. A. V. Piterina, J. Bartlett and J. T. Pembroke, Water Res., 2012, 46, 2488–2504.
19. F. M. Espinoza-Escalante, C. Pelayo-Ortíz, J. Navarro-Corona, Y. González-García, A. Bories and H. Gutiérrez-Pulido, Biomass Bioenergy, 2009, 33, 14–20.
20. Y. G. Chen, N. D. Xiao, Y. X. Zhao and H. Mu, Bioresour. Technol., 2012, 114, 349–356.
21. M. A. Dareioti, A. I. Vavouraki and M. Kornaros, Bioresour. Technol., 2014, 162, 218–227.
22. S. Gustin and R. Marinsek-Logar, Process Saf. Environ. Prot., 2011, 89, 61–66.
23. M. A. Latif, C. M. Mehta and D. J. Batstone, Water Res., 2015, 81, 288–293.
24. Y. F. Xu, H. Hu, J. Y. Liu, J. H. Luo, G. R. Qian and A. J. Wang, Chem. Eng. J., 2015, 267, 260–265.
25. H. P. Yuan, C. W. Xu and N. W. Zhu, Bioresour. Technol., 2014, 169, 686–691.
26. APHA, AWWA, WEF, American Public Health Association/American Water Works Association/Water Environment Federation, Washington, D.C., USA, 2005.
27. J. G. Caporaso, J. Kuczynski, J. Stombaugh, K. Bittinger, F. D. Bushman, E. K. Costello, N. Fierer, A. G. Pena, J. K. Goodrich and J. I. Gordon, Nat. Methods, 2010, 7, 335–336.
28. Y. Yuan, Y. Z. Peng, Y. Liu, B. D. Jin, B. Wang and S. Y. Wang, Bioresour. Technol., 2014, 174, 1–5.
29. USEPA, Environmental Protection Agency, Washington, DC, 1990.
30. A. Chu, D. S. Mavinic, W. D. Ramey and H. G. Kelly, Water Res., 1996, 30, 1759–1770.
31. K. Wang, J. Yin, D. S. Shen and N. Li, Bioresour. Technol., 2014, 161, 395–401.
32. H. Wu, A. Ikeda-Ohno, Y. Wang and T. D. Waite, Water Res., 2015, 76, 213–226.
33. Y. M. Kim, H. U. Cho, K. Y. Park, H. Park and K. H. Cho, Int. Biodeterior. Biodegrad., 2014, 92, 66–70.

---