Phosphorus Competition in Bioinduced Vivianite Recovery from

Article Cite This: Environ. Sci. Technol. 2018, 52, 13863−13870

pubs.acs.org/est

Phosphorus Competition in Bioinduced Vivianite Recovery from Wastewater Shu Wang,† Jingkun An,† Yuxuan Wan,‡ Qing Du,‡ Xin Wang,‡ Xiang Cheng,§ and Nan Li*,† †

Environ. Sci. Technol. 2018.52:13863-13870. Downloaded from pubs.acs.org by IOWA STATE UNIV on 01/14/19. For personal use only.

Academy of Eco-Environmental Science, School of Environmental Science and Engineering, Tianjin University, No. 135 Yaguan Road, Jinnan District, Tianjin 300350, China ‡ MOE Key Laboratory of Pollution Processes and Environmental Criteria, Tianjin Key Laboratory of Environmental Remediation and Pollution Control, Nankai University, No. 38 Tongyan Road, Jinnan District, Tianjin 300350, China § Beijing Key Lab Source Control Technology Water Pollution, Beijing Forestry University, Beijing 100083, China S Supporting Information *

ABSTRACT: Phosphorus undergoes a one-way flow from minerals to soil to water, which creates a phosphorus crisis as well as aquatic eutrophication. Dissimilatory metal reduction bacterial (DMRB)induced vivianite recovery from wastewater is a promising route to solve these problems synthetically. In this study, phosphorus competition between biomass growth and bioinduced vivianite mineralization was investigated at the batch scale. Biomass growth leads to phosphorus utilization over vivianite mineralization. Geobacter was selected as the main functional microorganism and presented higher vivianite recovery rates (20−48%) than sewage biomass (7−33%). An optimal Fe/P stoichiometric ratio of 1:1 was observed for both sewage biomass and Geobacter-inoculated batches. The highest vivianite yield of 4.3 mM was obtained in Geobacter-inoculated batches at a Fe:P of 1:1, with values 59% higher than those at a Fe:P of 1:0.67 (equal to the Fe/P molar ratio in vivianite). Sufficient PO43− stimulated cell growth and yielded a higher Fe3+ reduction rate and vivianite yield. Nevertheless, excessive PO43− facilitated the precipitation of KFe3 (PO4)2(OH)· 8H2O and Fe7 (PO4)6, which inhibited vivianite synthesis. In the optimal Geobacter batch, the μ−S curve indicated a mixed order reaction (0 < x < 1) for both vivianite formation and biomass growth. The vivianite growth series proceeded as follows: tiny blue particles, plain pieces, dark blue nodules, and large spherical crystals.

■

INTRODUCTION Phosphorus (P), a required nutrient, is involved in almost all major biochemical reactions. Accounting for approximately 2− 4% of the dry weight of most cells, phosphorus plays an essential role in inheritance (as a component of DNA and RNA) and energy metabolism.1 Unlike another essential nutrient element, nitrogen, which can be captured by nitrogen-fixing microorganisms from the atmosphere, phosphorus is primarily obtained from the artificial mining of phosphate rocks. However, only a very small proportion of phosphorus (0.007%) can be extracted economically and sustainably from phosphate rocks.2 Consensus has been reached among some experts that clean phosphate rocks are a limited resource, with the only disagreement of when this resource will be exhausted.3 Simultaneously, excessive phosphorus is discharged into bodies of water, resulting in eutrophication as well as the degradation of aquatic ecosystems. Ecological, geopolitical, and economic concerns call for phosphorus recovery.4 Li et al. proposed the concept of “wastewater-resource factories” to recover carbon, nitrogen, and phosphorus from wastewater.5 By June 2015, the 3802 wastewater treatment plants (WWTPs) in cities in China achieved a total sewage treatment capacity of over 1.6 × 108 m3/ d.6 Generally, the phosphorus content is approximately 5−20 © 2018 American Chemical Society

mg/L in domestic wastewater in the form of phosphates and organic phosphorus compounds,5 accounting for 1−2% of dry matter in urban sludge.7 A rough calculation from statistical data suggests that 800−3200 tons of phosphorus per day is treated in Chinese WWTPs.6 If promising phosphorus capture technologies are applied, WWTPs can serve as enormous urban phosphorus reservoirs. As a storable slow-release fertilizer, struvite has attracted a great deal of attention because it is recycled in some WWTPs.3 However, the limitations of the process (a combination of enhanced biological phosphorus removal (EBPR) and sludge digestion) and the less than ideal operational conditions (high alkalinity and magnesium concentration) inhibit the wider application of struvite recycling technology. Furthermore, the low phosphorus recovery efficiency of struvite decreases its application potential.4 New options are needed to obtain storable, transferable, and high-value phosphorus products. Received: Revised: Accepted: Published: 13863

June 13, 2018 September 26, 2018 November 9, 2018 November 9, 2018 DOI: 10.1021/acs.est.8b03022 Environ. Sci. Technol. 2018, 52, 13863−13870

Article

Environmental Science & Technology Iron is omnipresent in WWTPs and is either contained in influent (typically at a concentration of 0.5−1.5 mg Fe·L−1) or is used in the sewage treatment process (for phosphorus removal, sludge flocculation, filamentous bulking control, and odors and corrosion elimination).4 Generally, iron-bound phosphorus in sewage and sludge is less transferable, making it difficult to release the phosphorus from iron−phosphorus compounds and limiting phosphorus recovery. However, natural phosphorus mobilization from Fe−P complexes in aquatic and terrestrial ecosystems is very efficient.8−11 Wilfert et al. reviewed the relevance of iron and phosphorus chemistry to determine a new route for phosphorus recovery. They proposed the development of a new phosphorus recovery process based on biological, chemical oxidation, and reduction of FePs.4 Vivianite (Fe3 (PO4)2·8H2O) is an authigenic ferrous iron phosphate mineral that is commonly found in aquatic systems, terrestrial systems, and wastewater sludges.12 Vivianite is a biogenic mineral product of dissimilatory metal reduction bacteria (DMRB). Though the existence of vivianite in sludge has been known for a long time,13−15 its potential in phosphorus recovery was only recognized recently and accounts for 40−50% in iron-based phosphorus removal.4,16 As a dominant component of sludge phosphorus with a higher P content (a Fe:P molar ratio of 1:0.67 compared to 1:0.4 for ferric FeP precipitates in wastewater experiments), vivianite is believed to be the most important phosphorus pool and a promising route for phosphorus recovery in WWTPs.4 However, since the growth of DMRB also consumes phosphate, chemical vivianite formation is usually a competition between phosphorus and bacteria, which is a significant problem in real systems. Wastewater phosphorus recovery is commonly performed in anaerobic supernatants (PO43− concentration of 65−96 mg·L−1)17,18 or digested sludge supernatants (PO43− concentration of 137−177 mg·L−1).18,19 Because of the wide range of PO43− concentrations for phosphorus recovery, it is necessary to understand the phosphorus competition between bacteria growth and mineral growth in vivianite recovery, which has not yet been investigated. The main objective of this study is to explore the phosphorus competition mechanism of vivianite formation mediated by DMRB. Fe(III) reduction and vivianite formation were investigated at the batch scale via inoculation with sewage biomass or typical DMRB. The optimal Fe:P in vivianite formation was investigated by identifying the fates of Fe and P in liquid, mineral precipitant, and biomass (bio-P). The dynamics of biomineralization were also investigated in an optimal pure culture batch.

injected into a sterilized basal medium (100 mL) at the late-lag phase (OD600 = 0.2).20 All reagents in this research were analytical pure grades. The basal medium for both inocula contained the following: sodium acetate (20 mM), ferric citrate (9 mM), NH4Cl (1.5 g·L−1), CaCl2 (0.075 g·L−1), MgCl2·6H2O (0.1 g·L−1), KCl (0.1 g·L−1), 5 mL·L−1 vitamin solution, and 12.5 mL·L−1 mineral solution.21 The phosphate concentration in the medium was controlled by adding 1.8, 3, 4.5, 6.0, 9.0, and 18 mM (Fe:P = 1:0.2, 1:0.33, 1:0.5, 1:0.67, 1:1, and 1:2, respectively) KH2PO4, marked as M1−M6 for sewage biomass inoculated in basal media groups, W1−W6 for raw sewage directly inoculated groups, S1−S6 for sludge groups, and P1−P6 for G. sulfurreducens PCA-inoculated groups, respectively. It should be noted that Fe/P ratios of M4, W4, S4, and P4 are the same with the standard stoichiometric Fe/P ratios in vivianite (1:0.67). Each sample had three trials conducted in parallel. All experiments were performed continuously at 30 °C on the shaker (120 rpm) for 30 days. The samples were collected at specific time points using a sterile syringe. Abiotic experiments were executed by dosing iron-based chemical (9 mM) into 1.8, 3, 4.5, 6, 9, and 18 mM KH2PO4 to investigate the chemically induced precipitation of vivianite. Ferric chloride (FeCl3·6H2O), ferrous sulfate (FeSO4·7H2O), and ferric citrate were feeding into abiotic batches under the same temperature (30 °C) with biotic batches. Growth of G. sulfurreducens PCA. The G. sulf urreducens PCA (ATCC 51573) was taken from laboratory frozen stocks. The growth medium was prepared with the following constituents: NH4Cl (1.5 g·L−1), KCl (0.1 g·L−1), CaCl2 (0.075 g·L−1), MgCl2·6H2O (0.1 g·L−1), KH2PO4 (0.6 g·L−1), 5 mL·L−1 vitamin solution, and 12.5 mL·L−1 mineral solution.21 Sodium acetate (20 mM) was used as the electron donor, while ferric citrate (20 mM) served as the electron acceptor.22,23 The culture medium was adjusted to pH 7.8−8.0 with NaOH and flushed with oxygen-free N2−CO2 (80:20 [vol/vol]) prior to sealing with butyl rubber stoppers and autoclaving.24 Before inoculation, the medium was autoclaved at 120 °C for 20 min and then cells were transferred to a fresh medium using a sterile syringe.22,25 The cell density was monitored by measuring the optical density (OD) at 600 nm every 6 h using a micro plate reader (SPARK 10M, TECAN, Switzerland).20 The entire experiment was conducted in an anaerobic glovebox (Thermo Scientific 1209, USA). Chemical, Material, and Microbial Analyses. The concentrations of Chemical Oxygen Demand (COD), PO43−, Fe2+, TFe, Total Suspended Solids (TSS), and Volatile Suspended Solids (VSS) were determined according to standard methods.26,27 The Fe2+ and Fe3+ concentrations were measured by phenanthroline spectrophotometry.27 The PO43− concentration was determined by a modified molybdenum antimony anti-spectrophotometry method described by Uhlmann et al.26 To measure the contents of P, Fe(II), and Fe(III) in the mineral phase, precipitants were pretreated in 0.5 M HCl overnight, and then chemical analyses of the corresponding ions were performed according to the methods described above.28 The solution pH was measured with a pH meter (PHS-3E, INESA, China). All measurements were performed in triplicate, and the results are presented as the mean values. The morphology of the samples and their elemental compositions were examined by scanning electron microscopy-energy dispersive X-ray (SEM-EDX) (LEO 1530VP, Germany). The precipitates were observed under a reflected-

■

MATERIALS AND METHODS Batch Experiments for Anaerobic Phosphorus Recovery. Anaerobic phosphorus recovery was carried out using either sewage biomass or pure culture. For the wastewater biomass group, raw sewage and anaerobic sludge were collected from a local municipal WWTP (Tianjin University, Tianjin, China), and the characteristics are listed in Table S1. All samples were placed in a glass bottle (1 L) overnight to remove oxygen. For raw sewage, one group was inoculated to anaerobic bottle (100 mL) directly, another was centrifuged at 4000 rpm for 3 min, then biomass was resuspended in a sterilized basal medium, and samples were transferred to an anaerobic bottle (100 mL). For the pure culture group, the harvested Geobacter sulfurreducens PCA cells (NCBI BioProject number: PRJNA57743) were 13864

DOI: 10.1021/acs.est.8b03022 Environ. Sci. Technol. 2018, 52, 13863−13870

Article

Environmental Science & Technology

Figure 1. Variation of (a) the Fe(II) content, (b) Fe2+ concentration, (c) Fe(III) content, (d) Fe3+ concentration, (e) P content, and (f) PO43− concentration in basal media sewage batch experiments (Fe:P = 1:0.2, 1:0.33, 1:0.5, 1:0.67, 1:1, and 1:2 were marked as M1−M6).

inhibited at a low initial PO43− concentration (1.8−4.5 mM). A sharp increase of solid Fe(III) was observed in M5 and M6 from day 19 to day 25, which was indicative of Fe(III) precipitation formation (Figure 1c) that was possibly due to the bioconsumption of citrate (the complexing agent of Fe3+). At the end of these batches, M6 still had 5.29 mM residual PO43− in the solution, which was a value that was 2−52 times higher than that of other samples (0.10−1.58 mM). The Fe(III) reduction rates (R) of M1−M6 were calculated as follows,

light microscope (Olympus, Japan), and their composition was analyzed by an X-ray diffractometer (XRD) (Bruker AXS GmbH, Germany). The diffraction patterns were obtained for 10° < 2θ < 90° with a step size of 10°. Jade 6.0 (Materials Data, Inc. USA) was used to analyze the mineral phases from the PDFdatabase licensed by ICDD (International Centre for Diffraction Data, 30-0662: vivianite Fe3(PO4)2·8H2O).29,30 DNA was extracted from wastewater samples using a Soil Genomic DNA Kit (CW2091S, 132 ComWin Biotech 133 Co., Ltd., China) according to standard protocols.20,31 To investigate the bacterial community, amplicons were quantified and sequenced on the MiSeq Illumina sequencing platform by Novogene (Beijing).

R=

■

d(C Fe(II) + C Fe 2 +) dt

(1)

where CFe(II) is the solid Fe(II) content (mM), CFe2+ is the concentration of Fe2+ (mM), and t is the inoculation time (d). Linear increases of the total Fe(II) contents with time were observed in all sewage batches. The average Fe(III) reduction rates of M5 and M6 reached 0.16 and 0.18 mM·d−1, respectively. These values increased by a factor of 1−5 compared to the other samples, which ranged from 0.03 to 0.07 mM·d−1. Combined with the low total Fe(II) contents at initial PO43− concentrations from 1.8 to 4.5 mM, the relatively low overall Fe(III) reduction rates of M1−M4 re-emphasized that the lack of PO43− has a negative impact on the Fe(III) reduction. It is very interesting that M4 was under the exact Fe/P stoichiometric ratio of vivianite, indicating that there was competition for the use of P by other consumers. At the end of sewage directly inoculated

RESULTS AND DISCUSSION Anaerobic Fe(III) Reduction and Phosphorus Recovery with Sewage Biomass. The fate of Fe and P in the liquid and solid phases is critical for P recovery. The Fe(II) content in the solid phase obviously increased with the increase of the initial PO43− (Figure 1a), accompanied by the decrease of soluble Fe3+ and PO43−. At the end of the basal medium sewage batches, M1 had the lowest Fe(II) content of 0.56 mM. The highest Fe(II) content was 5.3 mM in M6, followed by 4.5 mM in M5 and 1.9 mM in M4, all of which increased by a factor of 2−8 compared to that in M1. The soluble Fe2+ concentration slightly varied from 0.19 mM (M3) to 0.32 mM (M6) (Figure 1b). The low total Fe(II) contents in M1−M3 indicated that Fe3+ reduction was 13865

DOI: 10.1021/acs.est.8b03022 Environ. Sci. Technol. 2018, 52, 13863−13870

Article

Environmental Science & Technology groups, Fe(II) content increased from 0.73 mM (W1) to 3.07 mM (W5) and then decreased to 1.72 mM in W6 (Figure S1), and the soluble Fe2+ concentration varied from 0.13 to 0.26 mM. Though the R of W5 increased by 1−3 times compared to others (0.03−0.06 mM·d−1) and reached 0.11 mM·d−1, it was slightly lower than that in M5 (0.16 mM·d−1). For sludge groups, the change trend was consistent with sewage groups; S5 had the highest Fe(II) content of 2.45 mM (Figure S2), which was 25% and 46% lower than W5 and M5. The Fe2+ concentrations varied from 0.14 to 0.27 mM, and the highest R of S5 reached 0.09 mM· d−1, followed by 0.07 mM·d−1 in S6. The XRD patterns of the precipitant collected from M3−M6 demonstrated the existence of vivianite as well as other FePs, such as KFe3(PO4)2(OH)·8H2O and Fe7(PO4)6 (Figure S3), with the highest vivianite peak observed in M5. However, these peaks almost disappeared in M3, which further confirmed that the low PO43− feeding resulted in a limited mineral content in precipitates. In abiotic controls, FePs precipitate was observed when dosing FeCl3 concentration higher than 6 mM (Table S2). However, stable Fe(H2PO4)3 precipitate was hard for bioutilization. No vivianite was found in batches with FeCl3 and KH2PO4. The output of phosphorus in FeCl3 chemical precipitation was divided into two parts: liquid−P and Fe(III)−P. With the Fe/P = 1:0.67−1:2, only 2−7% of TP was precipitated and more than 90% of TP remained in solution. When the dosing FeCl3 increased to 18 mM, liquid−P and Fe(III)−P took almost an equal share of TP in A6 (Table S3). The same concentration (9 mM) of FeSO4 and ferric citrate cannot react with 1.8−18 mM KH2PO4 completely under abiotic conditions. Vivianite production hardly occurred through chemically induced precipitation under Fe/P ratio of 1:0.2 to 1:2, which confirmed that biological-induced precipitation was the main mechanism for vivianite production in wastewater treatment process. Combining the XRD results with the variation of solid Fe(II) and P, the P recovery rate (RP) and vivianite recovery rate (RV) were calculated as follows: RP =

CSP × 100% C0

Figure 2. Phosphorus and vivianite recovery rate in basal media sewage batch experiments (Error bars represent the standard deviations of three repeated analyses).

was obtained in S5, while the highest RV was only 19 ± 2% (Figure S4b), 42% lower than that of M5 (33 ± 2%). A higher content of P-precipitates was generated in W5 and S5 because of the complicated components of sewage and sludge, while a higher biomass of complex communities in activated sludge competes for more nutrients, which inhibited Fe3+ reduction and vivianite production. Although the microbial community varies with the initial PO43− concentration, Proteobacteria dominated (>50%) in M1, M4, and M5 after 30 days (Figure 3a). At the genus level (Figure

(2)

where CSP is the final solid P content and C0 is the initial PO43− concentration. RV =

CV × 100% C0

(3)

where CV is the final solid P in the form of vivianite. As shown in Figure 2, RP increased from 31 ± 1% in M1 to 50 ± 2% in M5 and then decreased to 37 ± 1% in M6. M5 had the highest RP of 50 ± 2%, as well as the highest RV of 33 ± 2%, in which vivianite accounted for 67% of the total solid phosphate. Though M6 was fed the highest PO43− concentration, RP and RV decreased to 37 ± 1% and 18 ± 1% of those in M5, respectively. Excess PO43− was observed to limit the formation of vivianite, but enhance precipitation of other phosphates, such as KFe3(PO4)2(OH)·8H2O and Fe7(PO4)6 (Figure S3). In sewage directly inoculated groups, the highest RP reach 57 ± 3% (W5), slightly higher than that of sewage inoculated in basal media (50 ± 2% in M5), while the highest RV of 23 ± 2% in W5 (Figure S4a) was 10% lower than that in M5 (33 ± 2%). Higher RV of M5 confirmed that the basal media was more suitable for DMRB growth. Similarly, the highest RP of 52 ± 2%

Figure 3. Typical microbial community analysis of M1, M4, M5, and sewage biomass. Relative abundance (a) at the phylum level; (b) at the genus level.

3b), Geobacter accounted for 66−69% in M4 and M5, but was relatively low in M1 (30%). It should be noted that the highest Geobacter proportion of 69% was observed in M5, in accordance with the highest RV, which indicated that Geobacter was the key microorganism for vivianite formation. 13866

DOI: 10.1021/acs.est.8b03022 Environ. Sci. Technol. 2018, 52, 13863−13870

Article

Environmental Science & Technology

lower Fe(II) content of 2.7 mM. Simultaneously, the Fe3+ concentrations of P5 and P6 decreased by 85% and 92%. The Fe(II) precipitant accumulated quickly from day 7 to 17 in P5 and P6, correlated to the exponential growth of G. sulf urreducens PCA (Figure 4a and Figure S5), while the Fe(III) precipitant appeared after day 17 (Figure 4c). From 7 to 17 days, the Fe(III) reduction rates reached 0.27 mM·d−1 in P5 and 0.28 mM·d−1 in P6, with values of 69% and 56% higher than those in the corresponding sewage-inoculated systems, showing that pure Geobacter exhibited a higher Fe(III) reduction activity than mixed consortia. Sufficient and excessive phosphate-stimulated cell growth in P5 (6 × 108 CFU/L) and P6 (5 × 108 CFU/L) (Figure S5), leading to higher Fe3+ reduction rates. Clean and short-rod cells with little extracellular products and precipitates were observed at low PO43− concentrations, such as P1 and P2, in SEM images (Figure 5a,b). Secretion of extracellular polymeric substances (EPS) seemed to be inhibited under these conditions. Meanwhile, precipitant was barely observed on the cell surface. EPS appeared on SEM images when the Fe:P was higher than 1:0.5 (Figure 5c). Cells were tightly aggregated with minerals by EPS when Fe:P further increased to 1:1 and 1:2 (Figure 5d−f). When Fe:P decreased from 1:0.2 to 1:1, RP increased from 33% (P1) to 54% (P5) and then decreased to 46% in P6 (Figure 6b). Similar to the sewage-inoculated system, the highest RV of 48% was obtained in P5, which was 81% higher than that in P6 (26%). P in the form of vivianite accounted for 89% of the total P in the precipitate of P5. The relative share of vivianite decreased in the XRD spectrum of P6, indicating excessive PO43−, which facilitated the precipitation of KFe3(PO4)2(OH)·8H2O and Fe7(PO4)6 (Figure S6). Particularly, the RV values of pure culture batches (20−48%) were significantly higher than those of the parallel sewage batches (7−33%). The highest RV (33%, M5) in the sewage batch was 31% lower than that of pure Geobacter cultures (48%, P5), which further confirmed that Geobacter played a key role in vivianite formation. The mass balance equation in pH stable batches (Table S4) can be written as TP = Fe(III)−P + Vivi−P + Liq−P + Bio−P. When feeding with insufficient PO43− (1.8 and 3 mM), nearly 60% of fed PO43− was utilized by bacteria, while 30−40% was precipitated as minerals (Figure 6a). Vivi−P accounted for 60− 85% of P-precipitate in P1 and P2 (Table S5). However, with higher phosphate dosages (4.5, 6, 9, and 18 mM), only 26−38% of the total phosphorus was used by bacteria and 46−54% was precipitated as minerals (Figure 6a); over 80% of mineral−P existed as the form of Vivi−P (Table S5). Although the fraction of Bio−P in P1−P3 varied from 38% to 58%, the total moles of phosphorus consumed by bacteria was similar, in a narrow range of 1.1−1.7 mmol. DMRB has priority for phosphate utilization for biomass growth. Phosphate can be used for bioinduced vivianite recovery, only after the phosphate requirement for DMRB growth is fully met. At the end of the experiments, 28% of PO43− was retained in the liquid phase in P6, implying that the phosphorus content at a Fe:P ratio of 1:2 was excessive for bioutilization and bioinduced vivianite recovery. Vivianite Growth Dynamics Induced by G. sulfurreducens PCA. According to the performance of the vivianite recovery, P5 (Fe/P = 1:1, the highest RV of 48%) was selected to investigate vivianite growth using a reflected-light microscope (Figure S7). Several small blue particles (2−10 μm in length) gradually formed on day 6 (Figure S7a) and then gathered to form plain pieces (10−20 μm in diameter) on day 10 (Figure S7b). Dark blue nodules (20−30 μm in diameter) were

In general, it is believed that a higher Fe content is favorable for P removal and vivianite formation. A ferrous Fe:P molar ratio of 1:0.67 was reported to reach a phosphorus removal efficiency as high as 98% in batches with secondary effluents. Heiberg et al.32 believed it was necessary to provide excess Fe for efficient P removal, and the lowest proper Fe:P molar ratio was supposed to be 2 (1:0.5). Fulazzaky et al.33 and Luedecke et al.34 suggested that a higher Fe:P ratio tended to form vivianite as a net sink for P. In this study, however, the Fe/P molar ratio in vivianite was 1:0.67 and the highest RP and RV was obtained in M5 (Fe/P = 1:1) rather than M4 (Fe/P = 1:0.67). Relatively more phosphorus was required for maximum vivianite production, probably because of the growth consumption of DMRB. To confirm that phosphorus was used by DMRB but no other microorganisms, these tests were repeated in a DMRB pure cultured system. G. sulfurreducens PCA-Mediated Anaerobic Phosphorus Recovery. To detail the mechanism of how initial PO43− concentration affects the phosphorus competition between bioutilization and bioinduced vivianite recovery, G. sulf urreducens PCA was selected as the model strain according to community analysis. P1−P4 shared similar trends of Fe and P variation (Figure 4), which were comparable to that in sewage-

Figure 4. Variation of (a) the Fe(II) content, (b) Fe2+ concentration, (c) Fe(III) content, (d) Fe3+ concentration, (e) P content, and (f) PO43− concentration in G. sulf urreducens PCA-inoculated batch experiments (Fe:P = 1:0.2, 1:0.33, 1:0.5, 1:0.67, 1:1, and 1:2 were marked as P1−P6).

inoculated systems. In these samples, solid Fe(II) and P as well as soluble Fe2+ increased slowly, accompanied by a mild decrease of soluble Fe3+ and PO43−. At the end of the pure culture batches, 0.4, 1.0, and 2.0 mM solid Fe(II) was detected in P1−P3, respectively. Feeding with sufficient or excessive phosphate, the solid Fe(II) contents in P5 and P6 were 11 and 12 times higher (4.3 mM in P5 and 4.8 mM in P6) than that of P1 on day 31. P4, which had a Fe/P stoichiometric ratio of 1:0.67, exhibited a 13867

DOI: 10.1021/acs.est.8b03022 Environ. Sci. Technol. 2018, 52, 13863−13870

Article

Environmental Science & Technology

Figure 5. SEM results of the precipitant in (a) P1, (b) P2, (c) P3, (d) P4, (e) P5, and (f) P6 (all samples were observed at a magnification of 10 000×).

Figure 6. Phosphorus distribution (a) and vivianite recovery rate (b) of G. sulfurreducens in PCA-inoculated batch experiments.

assembled (days 14 and 18) from these plain pieces (Figure S7c,d) and then grew into larger spherical crystals; platy and needle-shaped crystal aggregates were formed (40−50 μm in diameter, days 25 and 30) (Figure S7e,f). As observed in the SEM images, rough surface and flowerlike glimmering crystals appeared in the precipitant on day 25 (Figure S8). With subtraction of the background value of glassy carbon, O, Fe, and P were the major elements of the mineral. EDX analysis revealed that the Fe:P ratio of the mineral was 1:0.67 (determined from at %, Figure S8), which was the same as the ratio of vivianite. Vivianite formation accompanied biomass growth in P5 (Figure 7a). The logarithmic growth of bacteria started from day 12, and the maximum growth rate appeared on day 17, both of which were 2 days earlier than the start of vivianite formation; the bacteria growth rate (VB) and vivianite production rate (VV) added weight to the discovery. (Figure 7b).

VB =

dNB dt

(4)

VV =

d NV dt

(5)

where NB and NV are the biomass and vivianite production, respectively. The specific bacterial growth (μB) and vivianite yield rate (μV) were calculated as

13868

μB =

VB NB

(6)

μV =

VV NV

(7) DOI: 10.1021/acs.est.8b03022 Environ. Sci. Technol. 2018, 52, 13863−13870

Article

Environmental Science & Technology

Figure 7. Vivianite production, phosphate concentration, and CFU counts (a); growth rates of bacteria and vivianite (b); and specific growth rate with the PO43− concentration (c) of the P5 group.

which showed that an online dosing system is necessary to control the Fe:P ratio in the narrow window of 1−2. By combing the mechanism with the engineering process, we expected to implement that all dosed iron would be recovered as various FePs (ideally vivianite). The efficient bioinduced vivianite recovery will bring net benefits instead of additional chemical costs. Overall, our findings provide fundamental data for determining the fate of phosphorus during bioinduced vivianite recovery from wastewater, which has broad implications in phosphorus cycling and vivianite mineralization.

In addition, the Monod Equation (eq 8) was applied to simulate the specific growth rate (μ) using the PO 4 3− concentration (S). μ=

μmax S Ks + S

(8)

where μ is the specific growth rate, which is equivalent to μB and μV; μmax is the maximum specific growth rate; and Ks is a constant. In the optimal condition (Fe/P = 1:1), the μB−S and μV−S curves both indicated a mixed-order reaction (0< x < 1), and the trends of the two curves were consistent; μB was higher than μV throughout the whole stage, which indicated that bacteria growth occurred prior to mineral formation. In addition, when the PO43− concentration was close to 0, the μ−S curve could be reflected as a first-order reaction with a gray dotted line (Figure 7c). Implications. In the vivianite-based biological phosphorus recovery system, we found a phosphorus competition between biomass growth and vivianite production, where the highest vivianite recovery was obtained at an Fe:P ratio of 1:1 rather than 1:0.67. Geobacter was found to act as the main functional microorganism in sewage biomass batch experiment, and had higher phosphorus and vivianite recovery capacities than those of sewage biomass. The vivianite growth series was small blue particles, plain pieces, dark blue nodules, and large spherical crystals. DMRB has priority for phosphate utilization for biomass growth. Phosphate can be used for bioinduced vivianite recovery, but only after the phosphate requirement for DMRB growth is fully met. Therefore, it is very important to ensure a sufficient content of phosphate for DMRB growth when vivianite-based biological phosphorus recovery technology is utilized in wastewater. However, excessive phosphate resulted in the formation of KFe3(PO4)2(OH)·8H2O and Fe7(PO4)6,

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.8b03022. Additional tables, details on characteristics of raw sewage and activated sludge, chemically-induced precipitation, and mass balance of abiotic test and pure culture batches. Additional figures, details on different form of Fe and P variation in sewage directly batches and activated sludge batches, XRD of precipitates in sewage biomass batches, RP and RV in raw sewage directly batches and activated sludge batches, growth of Geobacter. Sulf urreducens PCA, XRD of precipitates in pure culture batches, reflectedlight microscope images of mineral growth, and SEMEDX results of blue mineral (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Fax: 86-022-2789 2622. Phone: 86022-8740 2072. ORCID

Xin Wang: 0000-0002-3522-5627 13869

DOI: 10.1021/acs.est.8b03022 Environ. Sci. Technol. 2018, 52, 13863−13870

Article

Environmental Science & Technology

(19) Marti, N.; Bouzas, A.; Seco, A.; Ferrer, J. Struvite precipitation assessment in anaerobic digestion processes. Chem. Eng. J. 2008, 141 (1), 67−74. (20) Zhou, L.; Li, T.; An, J.; Liao, C.; Li, N.; Wang, X. Subminimal inhibitory concentration (Sub-MIC) of antibiotic induces electroactive biofilm formation in bioelectrochemical systems. Water Res. 2017, 125, 280. (21) Lovley, D. R.; Phillips, E. J. Novel mode of microbial energy metabolism: organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl. Environ. Microbiol. 1988, 54 (6), 1472−1480. (22) An, J.; Li, N.; Wan, L.; Zhou, L.; Du, Q.; Li, T.; Wang, X. Electric field induced salt precipitation into activated carbon air-cathode causes power decay in microbial fuel cells. Water Res. 2017, 123, 369−377. (23) Leang, C.; Coppi, M. V.; Lovley, D. R. OmcB, a c-type polyheme cytochrome, involved in Fe(III) reduction in Geobacter sulfurreducens. J. Bacteriol. 2003, 185 (7), 2096−2103. (24) Zhang, F.; Yu, S.; Li, J.; Li, W.; Yu, H. Mechanisms behind the accelerated extracellular electron transfer in Geobacter sulfurreducens DL-1 by modifying gold electrode with self-assembled monolayers. Front. Environ. Sci. Eng. 2016, 10 (3), 531−538. (25) Lin, W. C.; Coppi, M. V.; Lovley, D. R. Geobacter sulfurreducens Can Grow with Oxygen as a Terminal Electron Acceptor. Appl. Environ. Microbiol. 2004, 70 (4), 2525−2528. (26) Uhlmann, D.; Röske, I.; Hupfer, M.; Ohms, G. A simple method to distinguish between polyphosphate and other phosphate fractions of activated sludge. Water Res. 1990, 24 (11), 1355−1360. (27) Gilcreas, F. W. Standard methods for the examination of water and waste water. Am. J. Public. Health. Nations. Health. 1966, 56 (3), 387−388. (28) Fredrickson, J. K.; Zachara, J. M.; Kennedy, D. W.; Dong, H.; Onstott, T. C.; Hinman, N. W.; Li, S. M. Biogenic iron mineralization accompanying the dissimilatory reduction of hydrous ferric oxide by a groundwater bacterium. Geochim. Cosmochim. Acta 1998, 62 (19−20), 3239−3257. (29) Zhang, Q.; Li, Y.; Yang, Q.; Chen, H.; Chen, X.; Jiao, T.; Peng, Q. Distinguished Cr(VI) capture with rapid and superior capability using polydopamine microsphere: Behavior and mechanism. J. Hazard. Mater. 2018, 342, 732−740. (30) Wang, B.; Xia, J.; Mei, L.; Wang, L.; Zhang, Q. Highly Efficient and Rapid Lead (II) Scavenger by Natural Artemia Cyst-shell with Unique Three-Dimension Porous Structure and Strong Sorption Affinity. ACS Sustainable Chem. Eng. 2018, 6 (1), 1343−1351. (31) Li, N.; An, J.; Zhou, L.; Li, T.; Li, J.; Feng, C.; Wang, X. A novel carbon black graphite hybrid air-cathode for efficient hydrogen peroxide production in bioelectrochemical systems. J. Power Sources 2016, 306, 495−502. (32) Heiberg, L.; Koch, C. B.; Kjaergaard, C.; Jensen, H. S.; Hans Christian, B. H. Vivianite precipitation and phosphate sorption following iron reduction in anoxic soils. J. Environ. Qual. 2012, 41 (3), 938−949. (33) Fulazzaky, M. A.; Salim, N. A. A.; Abdullah, N. H.; Yusoff, A. R. M.; Paul, E. Precipitation of iron-hydroxy-phosphate of added ferric iron from domestic wastewater by an alternating aerobic−anoxic process. Chem. Eng. J. 2014, 253, 291−297. (34) Luedecke, C.; Hermanowicz, S. W.; Jenkins, D. Precipitation of Ferric Phosphate in Activated Sludge: A Chemical Model and Its Verification. Water. Pollut. Res. Control. Brighton. 1988, 21 (4−5), 325− 337.

Xiang Cheng: 0000-0001-9019-7246 Nan Li: 0000-0002-5852-2325 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was financially supported by the National Natural Science Foundation of China (Nos. 51778408 and 21577068).

■

REFERENCES

(1) Karl, D. M. Phosphorus, the staff of life. Nature 2000, 406, 31−32. (2) Cordell, D.; White, S. Peak Phosphorus: Clarifying the Key Issues of a Vigorous Debate about Long-Term Phosphorus Security. Sustainability 2011, 3 (10), 2027−2049. (3) Gilbert, N. Environment: The disappearing nutrient. Nature 2009, 461 (7265), 716−718. (4) Wilfert, P.; Kumar, P. S.; Korving, L.; Witkamp, G. J.; van Loosdrecht, M. C. M. The Relevance of Phosphorus and Iron Chemistry to the Recovery of Phosphorus from Wastewater: A Review. Environ. Sci. Technol. 2015, 49 (16), 9400−9414. (5) Li, W. W.; Yu, H. Q.; Rittmann, B. E. Chemistry: Reuse water pollutants. Nature 2015, 528 (7580), 29−31. (6) Ministry of Housing and Urban-Rural Development of the People’s Republic of China. Report of the ministry of housing and urban-rural development on the construction and operation of urban sewage treatment facilities in the second quarter of 2015. http://www. mohurd.gov.cn/wjfb/201508/t20150817_223298.html (accessed 27 July 2015). (7) Zhang, Q. H.; Yang, W. N.; Ngo, H. H.; Guo, W. S.; Jin, P. K.; Dzakpasu, M.; Yang, S. J.; Wang, Q.; Wang, X. C.; Ao, D. Current status of urban wastewater treatment plants in China. Environ. Int. 2016, 92− 93, 11−22. (8) Geddeda, Y. I. Effects of vesicular-arbuscular mycorrhizae and phosphorus on apple seedlings. J. Am. Soc. Hort. Sci. 1984, 109 (1), 24− 27. (9) Roden, E. E.; Edmonds, J. Phosphate mobilization in iron-rich anaerobic sediments: microbial Fe(III) oxide reduction versus ironsulfide formation. Archiv. Fur. Hydrobiol. 1997, 139 (3), 347−378. (10) Hinsinger, P. Bioavailability of soil inorganic P in the rhizosphere as affected by root-induced chemical changes: a review. Plant Soil 2001, 237 (2), 173−195. (11) Chacon, N.; Silver, W. L.; Dubinsky, E. A.; Cusack, D. F. Iron Reduction and Soil Phosphorus Solubilization in Humid Tropical Forests Soils: The Roles of Labile Carbon Pools and an Electron Shuttle Compound. Biogeochemistry 2006, 78 (1), 67−84. (12) Rothe, M.; Kleeberg, A.; Hupfer, M. The occurrence, identification and environmental relevance of vivianite in waterlogged soils and aquatic sediments. Earth-Sci. Rev. 2016, 158, 51−64. (13) Ghassemi, M.; Recht, H. L. Phosphate precipitation with ferrous iron. Water. Pollut. Control Res. Series. 1971. (14) Frossard, E.; Bauer, J. P.; Lothe, F. Evidence of vivianite in FeSO4 -flocculated sludges. Water Res. 1997, 31 (10), 2449−2454. (15) Seitz, M. A.; Riedner, R. J.; Malhotra, S. K.; Kipp, R. J. Ironphosphate compound identification in sewage sludge residue. Environ. Sci. Technol. 1973, 7 (4), 354−357. (16) Wilfert, P.; Mandalidis, A.; Dugulan, A. I.; Goubitz, K.; Korving, L.; Temmink, H.; Witkamp, G. J.; Van Loosdrecht, M. C. Vivianite as an important iron phosphate precipitate in sewage treatment plants. Water Res. 2016, 104, 449−460. (17) Battistoni, P.; De Angelis, A.; Prisciandaro, M.; Boccadoro, R.; Bolzonella, D. P removal from anaerobic supernatants by struvite crystallization: long term validation and process modelling. Water Res. 2002, 36 (8), 1927−1938. (18) Martí, N.; Pastor, L.; Bouzas, A.; Ferrer, J.; Seco, A. Phosphorus recovery by struvite crystallization in WWTPs: influence of the sludge treatment line operation. Water Res. 2010, 44 (7), 2371−2379. 13870

DOI: 10.1021/acs.est.8b03022 Environ. Sci. Technol. 2018, 52, 13863−13870