Glucose and Applied Voltage Accelerated p-Nitrophenol Reduction in Biocathode of Bioelectrochemical Systems

p-Nitrophenol (PNP) is common in the wastewater from many chemical industries. In this study, we investigated the effect of initial concentrations of PNP and glucose and applied voltage on PNP reduction in biocathode BESs and open-circuit biocathode BESs (OC-BES). The PNP degradation efficiency of a biocathode BES with 0.5 V (Bioc-0.5) reached 99.5 ± 0.8%, which was higher than the degradation efficiency of the BES with 0 V (Bioc-0) (62.4 ± 4.5%) and the OC-BES (59.2 ± 12.5%). The PNP degradation rate constant (kPNP) of Bioc-0.5 was 0.13 ± 0.01 h-1, which was higher than the kPNP of Bioc-0 (0.024 ± 0.002 h-1) and OC-BES (0.013 ± 0.0005 h-1). PNP degradation depended on the initial concentrations of glucose and PNP. A glucose concentration of 0.5 g L-1 was best for PNP degradation. The initial PNP increased from 50 to 130 mg L-1 and the kPNP decreased from 0.093 ± 0.008 to 0.027 ± 0.001 h-1. High-throughput sequencing of 16S rRNA gene amplicons indicated differences in microbial community structure between BESs with different voltages and the OC-BES. The predominant populations were affiliated with Streptococcus (42.7%) and Citrobacter (54.1%) in biocathode biofilms of BESs, and Dysgonomonas were the predominant microorganisms in biocathode biofilms of OC-BESs. The predominant populations were different among the cathode biofilms and the suspensions. These results demonstrated that applied voltage and biocathode biofilms play important roles in PNP degradation.

INTRODUCTION p-Nitrophenol (PNP), a priority pollutant listed by the United States Environmental Protection Agency, is widely used for the synthesis of industrial products, and it is generated by the degradation of pesticides in the environment (Kowalczyk et al., 2015). In addition, PNP is known to have high toxicity, which can threaten ecosystem and human health if released directly into the environment (Chen et al., 2016). Hence, it is necessary to treat PNP-containing wastewater. The nitro group, with its high electron-withdrawing effect, and the benzene ring are resistant to oxidative degradation ; thus, an easy reduction of PNP is essential. Some physicochemical technologies have been applied to clean PNP-containing wastewater (Liu et al., 2010;An et al., 2012;Arbab Zavar et al., 2012;Yarlagadda et al., 2012;Zhang et al., 2012a;Zhou et al., 2016), which require high costs, extreme pH conditions, high power input, or a long processing time. Therefore, it is necessary to develop effective methods for PNP removal.
Some previous studies have shown that nitrophenols (such as PNP) were reduced to aminophenols (PAP) with less toxicity and easier mineralization (Wang et al., 2011Shen et al., 2012;Jiang et al., 2016). The use of biocathode BESs has been reported as a low energy and sustainable method for metals remediation and nitrate remediation (Huang et al., 2015). A previous investigation reported that the cathode biofilms of biocathode BESs played an important role in PNP degradation when sodium bicarbonate was used as sole carbon source in the cathodic chamber . The ecological conditions affect PNP degradation and microbial community structure of the biocathode biofilm in BESs. Optimizing the operation conditions of BESs is necessary to accelerate PNP reduction velocity. Nevertheless, the functional microbial community for PNP reduction has not been fully investigated. Moreover, it is important to investigate PNP degradation efficiency and stability with different concentrations of glucose and initial PNP because PNP-containing wastewater usually has variables carbon source concentrations and PNP content.
In this study, we investigated the kinetics of PNP degradation and PAP formation in biocathode BESs. We analyzed the effects of applied voltage and the initial concentrations of glucose and PNP on PNP degradation. We also explored the biocathode microbial communities using high-throughput sequencing of 16S rRNA gene amplicons.

Reactor Setup
We used two-chamber BES reactors consisting of glass bottles separated by a cationic exchange membrane (Ultrex CMI7000, Membranes International, Inc., United States). The working volume of each chamber was 300 mL. Carbon brushes (5 cm in diameter and 7 cm long, fiber type: T700-12K, Toray Industries, CO., Ltd.) were used as the electrodes. Prior to use, the membrane and electrode brushes were pretreated as previously described . The Ag/AgCl reference electrodes [0.247 V vs. standard hydrogen electrode (SHE), model-217, Shanghai Precision Scientific Instrument Co., Ltd., China] were inserted into the cathode chambers for measuring cathode potentials and for electrochemical analysis. The anode, cathode, and reference electrodes were connected to a data acquisition system (Keithley 2700, Keithley, Co., Ltd., United States) with high-precision external resistance (10 ). All electric potentials reported here were already against the SHE.

Inoculation and Operation
The reactors with biocathode were inoculated in fed-batch mode as previously described . PNP degradation was operated under three modes: (I) biocathode with closed circuit and 0.5 V of applied voltage (Bioc-0.5), (II) biocathode with closed circuit with 0 V of applied voltage (Bioc-0), and (III) OC-BES (as a control test). The anode anolyte culture medium contained 1.67 g L −1 of NaAC, trace minerals, vitamins, and 50 mM of phosphate buffer solution (PBS) (Lovley and Phillips, 1988;Lu et al., 2012). The cathode was fed with 30 mg L −1 PNP, trace minerals, vitamins, and 0.5 g L −1 glucose mixed with 50 mM PBS. We adjusted the glucose to 0.1, 0.3, 0.5, 0.8, and 1 g L −1 with 50 mg L −1 PNP to investigate the influence of different glucose concentrations. We adjusted the PNP concentration to 50, 70, 90, 110, and 130 mg L −1 with 0.5 g L −1 glucose to investigate the effects of different initial PNP concentrations. The biocathodes were replaced with new sterile carbon brushes (121 • C, 30 min) to determine the impact of biocathode microbial communities on PNP degradation. The reactors and the medium were autoclaved at 121 • C for 15 min. All experiments were operated in replicated cycles for consistency, and all experiments were conducted at 25 ± 2 • C.

Chemical Analyses and Calculations
The concentrations of PNP and PAP were determined, as previously reported, and the production was analyzed by a high performance liquid chromatography mass spectrometer (HPLC-MS) . We used a gas chromatograph (Agilent, 4890D; J&W Scientific, United States) with a flame ionization detector and an appropriate column (19095N-123HP-INNOWAX, 30 m × 0.530 mm × 1.00 µm, J&W Scientific, United States) to analyze the concentrations of volatile fatty acids (VFAs), including acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid, and valeric acid . The glucose concentration of the cathode effluent was analyzed with a glucose determination kit (RSBIO, Shanghai). We measured cell biomass of the cathode effluent with a Modified BCA Protein Assay Kit (Sangon Biotech). Before the protein tests, the effluent samples were frozen to −20 • C for 24 h, then thawed and boiled for 10 min.
Current density (Am −3 ) was calculated based on the cathode volume (300 mL). PNP degradation efficiency (DE PNP ) was calculated based on the difference between the influent and effluent PNP concentrations. The kinetics of the PNP reduction and PAP formation were assumed to follow the first-order reaction models C = C 0 e −kt and C = C 0 (1-e −kt ), respectively (C represents the PNP or PAP concentration (mg L −1 ) at time (h) and C 0 is the initial PNP concentration or maximum PAP concentration); the rate constant k (h −1 ) of PNP and PAP was calculated by Origin 8.0 software. The half-life time (t 1/2 ) of PNP was calculated using the equation t 1/2 = 0.693/k.

Electrochemical Analysis
We conducted cyclic voltammetry (CV) on the cathode using an electrochemical workstation (WMPG1000K8 multichannel potentiostat, WonATech, Co., Ltd., South Korea); the anode was the counter electrode, and Ag/AgCl was the reference electrode (+0.197 V vs. SHE). We measured CV for 30 mg L −1 of PNP and 0.5 g L −1 of glucose at a scan rate of 5 mV/s. All CV tests were operated at 25 • C with a scan range from −1.0 to +1.0 V. We used electrochemical impedance spectroscopy (EIS) with the same instrument as the CV tests with a frequency range from 100 KHz to 10 mHz using a 10 mV sine wave.

Microbial Community Analysis
Samples of cathodic biofilms (Bioc-0.5-C, Bioc-0-C, and OC-BES-C) and suspended growth cultures (Bioc-0.5-S, Bioc-0-S, and OC-BES-S) were removed aseptically from the corresponding reactors. DNA was extracted using the PowerSoil DNA Isolation Kit (MO BIO, Carlsbad, CA, United States). DNA samples were stored at −20 • C before analysis. Polymerase chain reaction (PCR) amplifications of bacteria were sequenced using the universal primers 8F (5 -AGAGTTTGATCCTGGCTCAG-3 ) and 533R (5 -TTACCGCGGCTGCTGGCAC-3 ) for the 16S rRNA gene V1-V3 region (length of approximately 455 bp). We conducted 454 GS-FLX pyrosequencing using the method described previously . The abundance of a given phylogenetic group was defined by the proportion of the number of sequences affiliated to that group to the total number of sequences obtained, and we conducted 454 pyrosequencing data analyses using the methods detailed in our previous study .
In all experiments, glucose was consumed quickly and the concentrations were less than 11.4 ± 2.2 mg L −1 within 6 h (Figure 2A). The concentration of VFAs of Bioc-0.5 was lower than the VFA concentration of OC-BES and higher than the VFA concentration of Bioc-0. Abundant cell biomass was present in the suspension of cathodic chamber (Figure 2B), and the cell biomass of the effluent was different among three modes. The cell biomass of Bioc-0.5 increased in the first 12 h, then decreased from 12 to 24 h, and the cell biomass was stable at the final stage (from 24 to 36 h). The cell biomass of Bioc-0 and OC-BES increased gradually, and the cell biomass was greatest in the OC-BES.

Electrochemical Properties of Biocathode BESs for PNP Removal
The cathode potential decreased and stabilized after 20 h at approximately −1.0 V for Bioc-0.5, which was lower than the cathode potential of Bioc-0 (−580 mV) and OC-BES (−493 mV) (Figure 3). The current density of Bioc-0.5 reached 2.49-2.26 Am −3 , which was much higher than the current density of Bioc-0 (approximately 0.057 Am −3 ). After reaching the maximum, the current density of Bioc-0.5 decreased and stabilized at approximately 0.5 Am −3 after 20 h; the current density of Bioc-0 was approximately 0.028 Am −3 . These results showed that applied voltage can significantly increase the absolute cathodic potential and the current density.   No redox peak was observed from the cathode of Bioc-0.5 and Bioc-0, but the polarization currents changed significantly (Supplementary Figure S1A). Compared to the CV curve of Bioc-0, the cathodic current of Bioc-0.5 was enhanced, and the onset potential had a positive shift with an application of 0.5 V. The EIS analysis indicated that the internal resistance of Bioc-0.5 was 391 , which was 23% less than the internal resistance of Bioc-0 (508 ) (Supplementary Figure S1B).
Current and cathode potential increased when the biofilms were eliminated from the cathode (Supplementary Figure S2). The PNP degradation rate (k PNP ) decreased from 0.098 ± 0.011 to 0.022 ± 0.001 (Supplementary Figure S3). PNP removal efficiency decreased from 98.4 ± 1.1 to 40.1 ± 2.8%, and the PAP concentration of the effluent decreased from 26.5 ± 0.4 to 4.5 ± 1.0 mg L −1 at 24 h with no biofilm (Supplementary Figure S4). These results showed that the electrode biofilm contributed to PNP removal.

Effect of Initial PNP Concentration on PNP Reduction
Five initial concentrations of PNP were used to assess PNP reduction with 0.5 g L −1 glucose ( Figure 5A). We found that k PNP decreased from 0.093 ± 0.008 to 0.027 ± 0.0001, k PAP decreased from 0.086 ± 0.00914 to 0.0231 ± 0.005, and DE PNP decreased from 99.8 ± 0.35 to 64.1 ± 2.4% when the initial concentration of PNP increased from 50 to 130 mg L −1 ( Figure 5A). As the PNP concentration increased from 50 to 130 mg L −1 , the cathode potential and the current density increased and then decreased (Figures 5B,C). A lower initial concentration of 30 mg/L showed a higher k PNP (0.13 ± 0.01 h −1 ) (Figure 1). These results showed that the initial PNP concentration influenced the rate of PNP reduction and its efficiency.
Most populations of bacteria in Bioc-0.5-C were affiliated with Streptococcus (42.7%), Lactococcus (15.6%), and Bacteroides FIGURE 6 | Relative abundance of predominant phylum (A) and genera (B) in the cathode suspension and biofilm of BES. The "others" was the phyla and genera less than 1% of the total summarized. The Bioc-0.5 was the biocathode BES with 0.5 V voltage, Bioc-0 was the biocathode BES with 0 V voltage, OC-BES was BES with open circuit. C represents cathode and S represents suspension.

DISCUSSION
This study proved BESs substantially enhanced PNP degradation. Compared to a previous study, k PNP (0.093 ± 0.008 h −1 ) of BESs fed with glucose was approximately three times greater than the k PNP of BESs fed with sodium bicarbonate (0.02978 ± 0.00339 h −1 ) . More importantly, electrodes of the biocathode were not the sole electron donor to PNP, glucose was also an electron donor for PNP degradation. In the anodic chamber, H + and electron were generated by NaAC and were transferred to the cathodic chamber as the electron for PNP reduction. The glucose in the cathodic chamber could also generate e − , and PNP was reduced to PAP in the cathodic chamber.
In the cathodic chamber, glucose was first transformed by the bacteria to relatively lower molecular VFAs through a fermentation process (Figure 2A). The VFAs were then used as the electron donors for PNP degradation after the glucose was consumed (6 h), which was called a syntrophic interaction in a previous study (Zeng et al., 2015). Syntrophic relationships between fermentative and PNP-reducing bacteria were essential in the biocathode when the glucose was fully consumed within 6 h (for an initial glucose concentration of 0.5 g L −1 ). PNP degradation in the OC-BES further indicated that partial PNP reduction was by glucose and not by the cathode.
A previous study indicated that PNP can inhibit PNP biodegradation because it is toxic to bacteria (Carrera et al., 2011). A high concentration of PNP (over 90 mg/L) could depress bacteria biofilms and the performance of BES. In our study, PNP reduction occurred from 30 to 130 mg L −1 , further demonstrating the advantage of PNP tolerance in biocathode BES for PNP degradation. The biofilm removed from the biocathode of BESs, the PNP degradation and the PAP formation decreased (Supplementary Figures S3, S4), implied that biofilms on the cathode and the applied voltage influenced PNP degradation. The current of the BES with an abiotic cathode was lower than the BES with a biotic cathode, indicating that the biocathode can supply electrons for PNP reduction.
The relative abundance of Streptococcus was highest in the biocathode biofilms of Bioc-0.5-C. Streptococcus, a biofilmforming pathogen, has been studied (Loo et al., 2000;Moscoso et al., 2006), but its capability of PNP degradation is not known. Dysgonomonas was present in the suspension and the OC-BES cathode, and it was the predominant genus detected in the BESs (Watanabe et al., 2011;Kodama et al., 2012). The electroactive Comamonas, enriched in the biofilms and suspension of OC-BESs, can use phenol, 4-nitrobenzoate, 4-chlorophenol, and nitrobenzene (Groenewegen and Debont, 1992;Hollender et al., 1997;Arai et al., 2000;Wu et al., 2006). The PNP degradation of Streptococcus, Dysgonomonas, and Comamonas should be investigated in the future. Revealing functional genes related with PNP degradation and syntrophic interaction between different populations using metagenomic technology is still important to understand PNP degradation in future study on BES.
CONCLUSION p-Nitrophenol degradation was enhanced in the biocathode BESs with glucose and an applied voltage of 0.5 V. PNP degradation efficiency of the BES was much higher than that of OC-BES. The initial concentrations of glucose and PNP influenced PNP degradation and PAP formation. The microbial communities of the biocathode biofilm and suspension were different in OC-BES and BES with different voltages, implying that the differences in microbial communities and BES resulted in different PNP degradation. These results demonstrated that voltage and biocathode biofilms contribute to PNP degradation.

AUTHOR CONTRIBUTIONS
DX and NR designed the experiments. XW performed the experiments. XW, DX, XM, and BL contributed the data analysis and wrote sections of the manuscript. All authors contributed to manuscript revision and approved the submitted version.