Effect of External Aeration on Cr (VI) Reduction in the Leersia hexandra Swartz Constructed Wetland-Microbial Fuel Cell System

: Cr (VI) is hazardous to humans and our environment. Leersia hexandra Swartz ( L. hexandra ) is the ﬁrst wet chromium hyperaccumulator found in China. This study constructed the L. hexandra constructed wetland-microbial fuel cell (CW-MFC) system to treat Cr (VI) wastewater. It also determined the effects of different dissolved oxygen (DO) concentrations on power generation, pollutant removal, and Cr (VI) reduction. Cathode aeration promoted the voltage output and pollutant removal of the L. hexandra CW-MFC when the DO concentration was 4.5 mg · L − 1 : the highest voltage was 520 mV, the chemical oxygen demand (COD) removal rate was 93.73%, and the Cr (VI) removal rate was 97.77%. Moreover, the increase in the DO concentration improved the absorption of heavy metal Cr by the substrate and L. hexandra , and promoted the transformation from Cr (VI) to Cr (III). Chromium mostly exists as a residue with low toxicity and low mobility in L. hexandra and the substrate. This proves that the increased DO concentration promotes the redox reaction in the system and plants, reducing Cr (VI) to Cr (III). At the same time, the key micro-organism Geobacter that enhances the performance of the system and Cr (VI) reduction was found. The research results can provide a reference for the subsequent CW-MFC treatment of actual Cr-containing wastewater.


Introduction
With rapid industrial development, a large amount of Cr-containing wastewater is being produced in the leather, mining, electroplating, pigment, and other industries [1].Chromium is present in water bodies mainly in the forms of Cr (VI) and Cr (III); however, Cr (VI) toxicity is approximately 100 times greater than that of Cr (III) [2].Cr (VI) has high solubility, mobility, and toxicity (mutagenicity, carcinogenicity, and teratogenicity) [3].Therefore, it is listed as a priority pollutant in many countries worldwide [4].Chromium (VI) is non-biodegradable and can be stored in the environment for many years [5] and reduced to Cr (III) under certain conditions.The majority of the chromium (III) is transformed into a solid phase by the bottom sludge in water, with very little remaining dissolved [6].Chemically reducing Cr (VI) to Cr (III) is therefore a viable and useful technology.However, current strategies involving direct reduction [7], catalytic reduction [8], microbial reduction [9], and other methods result in insufficient treatment despite a high investment and operation cost [10,11].
A constructed wetland (CW) and microbial fuel cell (MFC) can remove chromium from water [12,13].Leersia hexandra Swartz (L.hexandra) CW removes Cr (VI) from water, and L. hexandra CW detoxifies Cr (VI) by reducing Cr (VI) to Cr (III) [14].Some MFCs can remove Cr (VI) from water by generating large amounts of electrons [15,16].The CW and MFC systems may be coupled because the redox gradient of the natural layer of the CW system is highly consistent with the MFC operating conditions [17].In recent years, the coupled constructed wetland-microbial fuel cell (CW-MFC) system has received extensive attention [18][19][20].For example, the optimal removal rate of Cr (VI) by CW-MFC was approximately 91% [21].However, current research has only aimed to improve the treatment capacity of Cr (VI) [20], and the removal process of Cr (VI) in water by the CW-MFC system is unclear.
Oxygen as an electron acceptor is considered to be a key parameter for optimizing CW-MFC [22,23].When the system cathode is in an aerobic environment, aromatic amine mineralization [24], nitrification [25,26], and anode unused organic matter [27] will be further degraded under the action of micro-organisms, which is conducive to improving the removal efficiency of pollutants in the system [26,28].Intermittent aeration (IA) can promote the wastewater treatment and power generation of the Canna indica L. CW-MFC system [22].The overall performance of CW-MFC is improved by adding aeration devices or cultivating plants in the cathode area [29].The aeration process consumes more organic pollutants by inhibiting the activity of anaerobic heterotrophic micro-organisms [30].Therefore, the cathode dissolved oxygen (DO) concentration plays an important role in improving the system performance.Cr (VI) is reduced to Cr (III) after receiving electrons from the anode at the cathode as an electron acceptor [31].At the same time, the cathode O 2 acts as another electron acceptor to combine with electrons to form water and another reducing compound [23].In this process, both Cr (VI) and O 2 act as electron acceptors to consume electrons from the anode and promote the degradation of anode organic matter.The introduction of MFC provides an effective way for electron conduction to the cathode [22].Therefore, based on the above analysis, we speculate that there may be a competitive relationship between O 2 and Cr (VI) in the CW-MFC system, which affects the valence state conversion of Cr (VI) to Cr (III).Nevertheless, the role of O 2 in the reduction process of Cr (VI) needs to be further explored and clarified.
The first wet chromium hyperaccumulator discovered in China was L. hexandra [32] and this species was selected as a wetland plant in this study.The study analyzed the effect of an increasing DO concentration on the reduction ability of Cr (VI) in the L. hexandra CW-MFC system based on accumulation by L. hexandra, the contribution of micro-organisms, together with the physical adsorption and bioelectrochemical reduction.In addition, the relationship between Cr (VI) as an electron acceptor and O 2 was determined.The results of this study can provide a reference for the CW-MFC treatment of Cr-containing wastewater under practical conditions.

Construction of the CW-MFC System
The unplasticized polyvinyl chloride tube was used to construct the up-flow L. hexandra CW-MFC system with a diameter of 30 cm and an effective height of 60 cm (Figure 1).The bottom layer of the device was 8-16 mm gravel (15 cm), the middle layer was 5-8 mm ceramsite (15 cm), and the anode and cathode layers were 3-5 mm activated carbon (20 cm and 10 cm, respectively).The outside of the anode and cathode were wrapped with 100 mesh stainless steel mesh, and connected to the resistance box (1000 Ω, Fuyang Precision Instrument Factory, Hangzhou, China) by a copper wire to form a closed loop.The voltage information was collected by a data acquisition card (NI USB-6009, Suzhou Fuyutong Electronics Co., Ltd., Nanjing, China).The system uses synthetic Cr (VI) wastewater with L. hexandra as wetland plants, and a peristaltic pump connects the inlet reservoir with the reactor (BT100-2J, Baoding Lange Constant Flow Pump Co., Ltd., Baoding, China).The aeration device was arranged on the surface of the system, and the aeration rate (0-8 L•min −1 ) was changed by the gas flowmeter.Five devices were operated under the conditions of aeration rates of 0 L Baoding, China).The aeration device was arranged on the surface of the sy aeration rate (0-8 L•min −1 ) was changed by the gas flowmeter.Five devices w under the conditions of aeration rates of 0 L•min −1 , 2 L•min −1 , 4 L•min −1 , 6 L•min −1 , respectively, until the water quality and voltage data were stable a The substrate and plant samples of each device were collected for the follow Each device is placed in the artificial climate chamber of the school, and the temperature was 20-35 °C.The whole study lasted about 7 months.

Sludge Inoculation and Wastewater Composition
Anaerobic sludge was collected from Guilin Liquan Brewery.The orig sludge was pre-cultured in the laboratory for a week and then inoculated in for 2 months until the system was stable.A preliminary simulation experim formed to determine the optimal conditions for the measurements (result The basic component concentration of synthetic Cr (VI) wastewater in th ments was 300 mg•L −1 chemical oxygen demand (COD), 80 mg•L −1 Cr (VI), 10 nitrogen (TN), 5 mg•L −1 total phosphorus (TP), and 6.5-8.0 pH.

Water Quality and Electrochemical Analysis Index
Samples were taken regularly from each reactor outlet after stabilizat tem.DO concentration was measured in situ using a portable water qu (Smart Roll portable multi-parameter water quality detector, Pro1030, Springs Instruments, Yellow Springs, OH, USA)).The COD was measured ter quality determination of the COD-dichromate method (HJ828-2017).C in the water was measured using diphenylcarbohydrazide spectrophotom (GB7467-87).The pollutant removal efficiency RE was calculated using the f mula (1):

Sludge Inoculation and Wastewater Composition
Anaerobic sludge was collected from Guilin Liquan Brewery.The original anaerobic sludge was pre-cultured in the laboratory for a week and then inoculated into the reactor for 2 months until the system was stable.A preliminary simulation experiment was performed to determine the optimal conditions for the measurements (results not shown).The basic component concentration of synthetic Cr (VI) wastewater in the final experiments was 300 mg•L −1 chemical oxygen demand (COD), 80 mg•L −1 Cr (VI), 100 mg•L −1 total nitrogen (TN), 5 mg•L −1 total phosphorus (TP), and 6.5-8.0 pH.

Water Quality and Electrochemical Analysis Index
Samples were taken regularly from each reactor outlet after stabilization of the system.DO concentration was measured in situ using a portable water quality analyzer (Smart Roll portable multi-parameter water quality detector, Pro1030, Visa, (Yellow Springs Instruments, Yellow Springs, OH, USA)).The COD was measured using the water quality determination of the COD-dichromate method (HJ828-2017).Chromium (VI) in the water was measured using diphenylcarbohydrazide spectrophotometric method (GB7467-87).The pollutant removal efficiency R E was calculated using the following Formula (1): where R E is the pollutant removal efficiency, C in is the L. hexandra CW-MFC pollutant influent concentration, and C out is the L. hexandra CW-MFC pollutant effluent concentration.The unit of COD is mg•L −1 .
Voltage collection: After the system was stabled, the data acquisition card was used to measure the system voltage in real time and record the voltage information regularly.

Scanning Electron Microscopy
The original activated carbon, anode, and cathode samples after system stabilization were dried at 75 • C, and the dried samples were ground through a 200 mesh-sized sieve.The morphology of biofilm pattern was observed by scanning electron microscope (SEM, ZEISS Sigma 300/ZEISS Gemini SEM 300/TESCAN MIRA LMS, Taufkirchen, Germany).The specific steps are as follows: Take a trace solid sample and place it directly on the conductive glue.Then, using the Oxford Quorum SC7620 (Oxford, UK) sputtering coating apparatus, spray gold for 45 s at a 10 mA rate.Finally, use an SEM to photograph the sample's morphology.During the topographical shooting, the acceleration voltage was 3 kV.The component is the SE2's second electronic detector.

Sample Collection and Determination of Chromium
The substrate samples were collected by five-point sampling method.The collected samples were dried in an oven at 75 • C to constant weight.The roots, stems, and leaves of the plants were washed, deactivated in an oven at 105 • C for 30 min, and dried at 75 • C to constant weight.The samples were grated and passed through a 100 mesh-sized sieve.Cr (VI) was extracted using alkali digestion method, and the content was determined using diphenylcarbohydrazide spectrophotometric method.Total Cr was extracted using acid digestion method [14], and the content was determined using flame atomic absorption spectrometry.The experimental results were repeated 3 times and averaged.

Chemical Speciation Analysis of Cr in Plants
The chemical forms of L. hexandra roots, stems, and leaves of 0.4 g plant powder was analyzed following a stepwise treatment: (1) 80% ethanol, (2) deionized water, (3) 1 mol•L −1 sodium chloride, (4) 2% acetic acid, (5) 0.6 mol•L −1 hydrochloric acid, and (6) residual chromium.Ten mL of extractant was added to the sample in turn, oscillated at room temperature for 2 h, centrifuged, and separated, and repeated twice.All three centrifugal liquids were collected in a 50 mL triangular flask and placed on an electric heating plate at 120-180 • C.After heating and concentrating, 2 mL of concentrated nitric acid was added and capped for refluxing for 2 h, steamed to nearly dry, and repeated refluxing once.The plant residues were digested using 10 mL HNO 3 + 2 mL H 2 O 2 + 2 mL HF microwave closed digestion method (Shanghai Xinyi Microwave Chemical Technology Co., Ltd., MDS-8 (Shanghai, China)).The Cr content in the digestion solution was determined using a graphite furnace/flame atomic absorption spectrometer (900 T).

Chemical Speciation Analysis of Cr in the Substrate
The improved BCR continuous extraction method was used to extract the form of substrate Cr [33,34].The 0.2 g substrate sample was weighed, placed in a 100 mL polytetrafluoroethylene centrifuge tube, and three different extractants were added in turn for extraction (each extraction was performed for 16 h).The addition of the extractant and the corresponding chemical form of Cr are as follows: (3) Oxidizable state: 10 mL 30% H 2 O 2 (pH 2), then 10 mL 30% H 2 O 2 (pH 2)-cool, and add 50 mL 1 mol•L −1 NH 4 OAc (pH 2); (4) Residue state: The oxidizable extract residue was dried, filtered through a 100-mesh soil sample sieve, and digested with the substrate Cr total determination method.

Microbial Analysis Methods
The 16S rRNA gene was sequenced while the experimental system was running, and 454 high-throughput sequencing technology was utilized to evaluate the microbial community structure.System filler was also sampled in accordance with the experimental criteria.During sampling, the filler samples were taken out and put in sampling bottles.
They were then transported to the lab and kept chilled (−80 • C) until Sangon Biotech (Shanghai) Co., Ltd. could sequence them.The kit was used to extract DNA from the samples, amplify and purify the 16S RNA gene fragment in the V3-V4 region, and then perform 454 high-throughput sequencing.

Quality Control
During the analysis, plant standard samples (citrus leaves, GBW10020 (GSB-11), Institute of Geophysical and Geochemical Exploration) and blanks were added for quality control (one quality control sample was added for every 10 samples).The recovery rate is between 90-110% within the acceptable range of EPA.

Scanning Electron Microscopy Measurement before and after System Stabilization
The surface structure of the original activated carbon, and the cathode and anode activated carbon after hanging film are shown in Figure 2. The original activated carbon surface was relatively rough, with numerous layers of uneven pores creating favorable living conditions for bacterial growth, and increasing the possibility of adsorbing heavy metals [35,36].The activated carbon surface formed a dense biofilm compared with the original activated carbon.Micro-organisms are biocatalysts that can oxidize carbon sources and promote electricity production [21].Additionally, metal ions cover the surface of the activated carbon, and the pores are generally smooth.This might be caused by a decrease in the adsorbent surface's non-uniformity.This means that the metal is adsorbing on the functional groups in the pores [36].The L. hexandra CW-MFC system is steady throughout the experiment as a result.community structure.System filler was also sampled in accordance with the experimental criteria.During sampling, the filler samples were taken out and put in sampling bottles.They were then transported to the lab and kept chilled (−80 °C) until Sangon Biotech (Shanghai) Co., Ltd. could sequence them.The kit was used to extract DNA from the samples, amplify and purify the 16S RNA gene fragment in the V3-V4 region, and then perform 454 high-throughput sequencing.

Quality Control
During the analysis, plant standard samples (citrus leaves, GBW10020 (GSB-11), Institute of Geophysical and Geochemical Exploration) and blanks were added for quality control (one quality control sample was added for every 10 samples).The recovery rate is between 90-110% within the acceptable range of EPA.

Scanning Electron Microscopy Measurement before and after System Stabilization
The surface structure of the original activated carbon, and the cathode and anode activated carbon after hanging film are shown in Figure 2. The original activated carbon surface was relatively rough, with numerous layers of uneven pores creating favorable living conditions for bacterial growth, and increasing the possibility of adsorbing heavy metals [35,36].The activated carbon surface formed a dense biofilm compared with the original activated carbon.Micro-organisms are biocatalysts that can oxidize carbon sources and promote electricity production [21].Additionally, metal ions cover the surface of the activated carbon, and the pores are generally smooth.This might be caused by a decrease in the adsorbent surface's non-uniformity.This means that the metal is adsorbing on the functional groups in the pores [36].The L. hexandra CW-MFC system is steady throughout the experiment as a result.

Effects of Different DO Concentration on Pollutant Removal and Electricity Generation
With an increase in the aeration rate, the DO concentration of the system gradually stabilized at 60 days.When the aeration rates were 0 L•min −1 , 2 L•min −1 , 4 L•min −1 , 6 L•min −1 , and 8 L•min −1 , the average cathode DO concentrations were 3.0 mg•L −1 , 3.5 mg•L −1 , 4.0 mg•L −1 , 4.5 mg•L −1 , and 5.0 mg•L −1 , respectively (Figure 3a).The subsequent research analysis will be based on these DO concentrations.Figure 3b-d shows the relationship of pollutant removal and electricity generation with the DO concentration in the L. hexandra CW-MFC system.The pollutant removal rate and power generation gradually stabilized after 45 d in the L. hexandra CW-MFC system (Figure 3).The COD removal rate increased from 81.56% to 93.73%, and the Cr (VI) removal rate increased from 84.82% to 97.77% when the DO concentration increased from 3.0 to 4.5 mg•L −1 .The maximum voltage of the system was 520 mV at 4.5 mg•L −1 .The pollutant removal rate and electricity production decreased when the DO concentration increased to 5.0 mg•L −1 .

Effects of Different DO Concentration on Pollutant Removal and Electricity Generation
With an increase in the aeration rate, the DO concentration of the system gradually stabilized at 60 days.When the aeration rates were 0 L  3a).The subsequent research analysis will be based on these DO concentrations.Figure 3b-d shows the relationship of pollutant removal and electricity generation with the DO concentration in the L. hexandra CW-MFC system.The pollutant removal rate and power generation gradually stabilized after 45 d in the L. hexandra CW-MFC system (Figure 3).The COD removal rate increased from 81.56% to 93.73%, and the Cr (VI) removal rate increased from 84.82% to 97.77% when the DO concentration increased from 3.0 to 4.5 mg•L −1 .The maximum voltage of the system was 520 mV at 4.

Effect of Different DO Concentrations on Cr Content in the Substrate
The Cr content in the substrate initially increases, then decreases with an elevated DO concentration, and the Cr (III) content is higher than that of Cr (VI) (Table 1).The Cr (III) content in the cathode is higher than that in the anode under each DO concentration.The highest total Cr content in the substrate was 31,295.83mg•kg −1 , wherein the anode content was 14,945.83mg•kg −1 .The cathode content was 16,350.00mg•kg −1 when the DO concentration was 4.5 mg•L −1 .The highest Cr (III) content was 30,733.75mg•kg −1 , with an anode content of 14,587.92mg•kg −1 and a cathode content of 16,145.83mg•kg −1 .The lowest Cr (VI) content was 562.08 mg•kg −1 ; this comprised 357.92 mg•kg −1 and 204.17 mg•kg −1 at the anode and cathode, respectively.The Cr content in the substrate initially increases, then decreases with an elevated DO concentration, and the Cr (III) content is higher than that of Cr (VI) (Table 1).The Cr (III) content in the cathode is higher than that in the anode under each DO concentration.The highest total Cr content in the substrate was 31,295.83mg•kg −1 , wherein the anode content was 14,945.83mg•kg −1 .The cathode content was 16,350.00mg•kg −1 when the DO concentration was 4.5 mg•L −1 .The highest Cr (III) content was 30,733.75mg•kg −1 , with an anode content of 14,587.92mg•kg −1 and a cathode content of 16,145.83mg•kg −1 .The lowest Cr (VI) content was 562.08 mg•kg −1 ; this comprised 357.92 mg•kg −1 and 204.17 mg•kg −1 at the anode and cathode, respectively.The Cr content in L. hexandra initially increased, then decreased with an elevated DO concentration.The Cr (III) content was higher than that of Cr (VI) (Table 2).The chromium (III) content was higher in the roots than that in the stems and leaves under each DO concentration.The highest total Cr content in the L. hexandra root, stem, and leaf was 18,869.79mg•kg −1 , 3642.71 mg•kg −1 , and 3070.83mg•kg −1 , respectively, at a DO concentration of 4.5 mg•L −1 .The highest Cr (III) content was 18,228.13mg•kg −1 , 3223.96 mg•kg −1 , and 2685.42 mg•kg −1 , respectively.5).The proportion of the residual state increased to the maximum, accounting for 55% of the total.This indicated that the increased DO concentration contributed to the transformation of Cr to a stable state.The content of chemically extracted Cr in the stems of L. hexandra reached the maximum residual state (2024.06mg•kg −1 ), HCl extraction state (480.00mg•kg −1 ), H 2 O extraction state (350.69mg•kg −1 ), ethanol extraction state (220.13mg•kg −1 ), HAc extraction state (311.25 mg•kg −1 ), and NaCl extraction state (324.25 mg•kg −1 ) at 4.5 mg•L −1 (Figure 6).Chromium mainly existed as the residual form, the oxalate form, and as other bound chromium in the stems of L. hexandra.The chemical form of Cr in L. hexandra leaves is mainly in the residual state.The content of residual Cr increased, then decreased as the DO concentrations were raised.A maximum of 1303.75 mg•kg −1 was observed at a DO concentration of 4.5 mg•L −1 ; this accounted for 43% of the total Cr amount (Figure 7).This indicated that the increased DO concentration reduced the Cr bioavailability.The chemical form of Cr in L. hexandra leaves is mainly in the residual state.The content of residual Cr increased, then decreased as the DO concentrations were raised.A maximum of 1303.75 mg•kg −1 was observed at a DO concentration of 4.5 mg•L −1 ; this accounted for 43% of the total Cr amount (Figure 7).This indicated that the increased DO concentration reduced the Cr bioavailability.The chemical form of Cr in L. hexandra leaves is mainly in the residual state.The content of residual Cr increased, then decreased as the DO concentrations were raised.A maximum of 1303.75 mg•kg −1 was observed at a DO concentration of 4.5 mg•L −1 ; this accounted for 43% of the total Cr amount (Figure 7).This indicated that the increased DO concentration reduced the Cr bioavailability.

Community Diversity Analysis
Higher Chao1 and Ace indices reflect a more abundant microbial community.Higher community diversity is observed through a greater Shannon index and a lower Simpson index.The community distribution diversity of the L. hexandra CW-MFC system was higher with aeration than that without aeration (Table 3).Aeration improved the devel-

Community Diversity Analysis
Higher Chao1 and Ace indices reflect a more abundant microbial community.Higher community diversity is observed through a greater Shannon index and a lower Simpson index.The community distribution diversity of the L. hexandra CW-MFC system was higher with aeration than that without aeration (Table 3).Aeration improved the development and proliferation of micro-organisms in the L. hexandra CW-MFC system, and positively affected the increase of community diversity.

Effect of Cathode DO Concentration on System Performance
Dissolved oxygen is required for the growth and reproduction of cathode aerobic micro-organisms that affect biocathode performance [28].The increase of cathode DO concentration significantly promoted the voltage output and pollutant removal of L. hexandra CW-MFC.These findings are similar to previous studies showing that aeration promotes pollutant removal and bioelectricity production [37,38].The increased DO concentration in the cathode region accelerates the transfer of electrons from the anode to the cathode by increasing the number of terminal electron acceptors (O2) [39].Sufficient DO concentration can increase the reduction reaction rate at the cathode and increase the cathode potential, thereby improving the bioelectricity generation performance and pollutant removal rate [28].

Effect of Cathode DO Concentration on the Migration and Distribution of Cr
Substrate adsorption, plant absorption, and degradation (photodegradation, hydrolysis, and biodegradation) are considered the main pathways for Cr removal in CW-MFC [40].Increasing the DO concentration can improve the absorption of Cr by the substrate and L. hexandra, and promote the conversion of Cr (VI) to Cr (III).In this study, when the DO concentration was 4.5 mg•L −1 , the Cr (III) concentration in the substrate was the highest (Table 1), because the increase in the DO concentration could improve the activity of cathode micro-organisms (such as electrochemically active bacteria, EAB) and produce more bioelectricity [41], thus, promoting the cathodic electrochemical reduction of Cr (VI) to Cr (III) and total Cr removal.The substrate contains 59.5% of the total Cr in the wetland [42,43].This indicates that the substrate type and chemical composition may be a key factor in improving the metal-removal efficiency of the wetland [44].The accumulation of Cr (III) in L. hexandra reached a maximum value at 4.5 mg•L −1 (Table 2), indicating that the increase in the O2 content improved the enrichment ability and adsorption capacity of Cr in L. hexandra [45], and the Cr content in the plant roots was greater than that in the stems and leaves [46], indicating that the roots played an important role in absorbing heavy metal ions.On the one hand, the increase in the DO concentration promoted the

Effect of Cathode DO Concentration on System Performance
Dissolved oxygen is required for the growth and reproduction of cathode aerobic micro-organisms that affect biocathode performance [28].The increase of cathode DO concentration significantly promoted the voltage output and pollutant removal of L. hexandra CW-MFC.These findings are similar to previous studies showing that aeration promotes pollutant removal and bioelectricity production [37,38].The increased DO concentration in the cathode region accelerates the transfer of electrons from the anode to the cathode by increasing the number of terminal electron acceptors (O 2 ) [39].Sufficient DO concentration can increase the reduction reaction rate at the cathode and increase the cathode potential, thereby improving the bioelectricity generation performance and pollutant removal rate [28].

Effect of Cathode DO Concentration on the Migration and Distribution of Cr
Substrate adsorption, plant absorption, and degradation (photodegradation, hydrolysis, and biodegradation) are considered the main pathways for Cr removal in CW-MFC [40].Increasing the DO concentration can improve the absorption of Cr by the substrate and L. hexandra, and promote the conversion of Cr (VI) to Cr (III).In this study, when the DO concentration was 4.5 mg•L −1 , the Cr (III) concentration in the substrate was the highest (Table 1), because the increase in the DO concentration could improve the activity of cathode micro-organisms (such as electrochemically active bacteria, EAB) and produce more bioelectricity [41], thus, promoting the cathodic electrochemical reduction of Cr (VI) to Cr (III) and total Cr removal.The substrate contains 59.5% of the total Cr in the wetland [42,43].This indicates that the substrate type and chemical composition may be a key factor in improving the metal-removal efficiency of the wetland [44].The accumulation of Cr (III) in L. hexandra reached a maximum value at 4.5 mg•L −1 (Table 2), indicating that the increase in the O 2 content improved the enrichment ability and adsorption capacity of Cr in L. hexandra [45], and the Cr content in the plant roots was greater than that in the stems and leaves [46], indicating that the roots played an important role in absorbing heavy metal ions.On the one hand, the increase in the DO concentration promoted the conversion of Cr (VI) to Cr (III), thereby reducing Cr toxicity, which was more conducive to plant growth and Cr absorption [47].On the other hand, the promotion effect of plants was mainly due to its indirect effect.The increase in the cathodic DO concentration improves the microbial environment of the plant rhizosphere, enhances biodegradation [48], and accelerates the metabolism of EAB [49,50].Therefore, an appropriate increase in the DO concentration can improve the absorption of heavy metal Cr by the substrate and L. hexandra, and promote the conversion of Cr (VI) to Cr (III).The level of chemically extracted Cr was higher in the cathode than that in the anode (Figure 4).This may be because the reduction of Cr (VI) mainly occurred in the cathode [47], and the increase of the DO concentration increased the electron acceptor of the cathode.The oxidation potential of Cr (VI) (1.33 V) is higher than that of O 2 (1.23 V) [51].Therefore, Cr (VI) is a better oxidant than O 2 .The Cr (VI) electron acceptor in the cathode is first reduced when the cathode O 2 content increased.This proves that the increased DO concentration promotes Cr (VI) reduction.However, further increases in the DO concentration results in the O 2 cathode penetrating into the anode area and destroying the anaerobic environment of the anode; this is not conducive to the electricity generation and pollutant degradation of the system [37,52].The reducing ability of Cr (VI) is reduced when there is insufficient electron transfer to the cathode.The forms of Cr in the cathode and anode are mainly residual and oxidizable; this may be one of the reasons why most of the Cr remains in the substrate [43].
Increasing the DO concentration results in Cr mainly existing in the bound form with pectin, protein, heavy metal phosphate, and oxalate in L. hexandra (Figures 5-7).These forms reduce the toxicity and mobility of Cr [53][54][55].Plant cell walls (including intercellular spaces and vacuoles) mainly contain Cr in bound forms to pectinate, protein, heavy metal phosphate, and oxalate; this reduces the levels of free Cr, thereby reducing cellular toxicity [56,57].This suggests that the change in O 2 content affects the electricity generation of the system, indirectly changing the metabolic state of plants, and the redox reaction in plants.The chemical form of Cr changes in plants due to aeration.

Changes to the Microbial Community in the CW-MFC System under Aeration
Micro-organisms are the main components of the CW-MFC system that significantly affect its purification and power-generation efficiency [58].Increasing the concentration of cathode DO improves the performance of the system [59,60] and contributes to the reduction and sedimentation of Cr (VI) [61].Current research increased the cathode DO concentration by planting plants in the cathode [62,63] or providing greater aeration of the device [64].Increasing the DO concentration by artificial aeration improves heterotrophic bacterial activity without reducing denitrification and contributes to COD and nitrogen removal [65].Chromium removal efficiency is improved by altering microbial activity with aeration [66].In this study, the relative abundance of Geobacter in non-aerated and aerated systems was 2.09% and 3.66%, respectively (Figure 9).This indicated that aeration significantly enriched the Geobacter in the system, enhanced system performance, and facilitated the reduction of highly toxic Cr (VI) to less toxic Cr (III).The ability of L. hexandra to absorb Cr (III) is stronger than that of Cr (VI) [14].Geobacter are the main exoelectrogens and dissimilatory metal-reducing bacteria; their enrichment in the system promotes electricity production and Cr (VI) reduction [67].Therefore, the enrichment of Geobacter in the CW-MFC system may be the key micro-organism for Cr detoxification and super-enrichment.

Conclusions
The results of this study showed that the appropriate increase of DO concentration could improve the power generation and pollutant removal rate of the L. hexandra CW-MFC system.When the aeration rate was 6 L•min −1 , the DO concentration was 4.5 mg•L −1 , and 8 L•min −1 , respectively, until the water quality and voltage data were stable after 3 months.The substrate and plant samples of each device were collected for the following research.Each device is placed in the artificial climate chamber of the school, and the day and night temperature was 20-35 • C. The whole study lasted about 7 months.

Figure 3 .
Figure 3. Removal rates of pollutants and electricity production in the L. hexandra CW-MFC system at different dissolved oxygen (DO) concentrations: (a) DO concentration at different aeration rates; (b) voltage at different DO concentrations; (c) chemical oxygen demand (COD) removal rate at different DO concentrations; and (d) Cr (VI) removal rate at different DO concentrations.

Figure 3 .
Figure 3. Removal rates of pollutants and electricity production in the L. hexandra CW-MFC system at different dissolved oxygen (DO) concentrations: (a) DO concentration at different aeration rates; (b) voltage at different DO concentrations; (c) chemical oxygen demand (COD) removal rate at different DO concentrations; and (d) Cr (VI) removal rate at different DO concentrations.3.3.Effect of Different DO Concentrations on Cr Content in the L. hexandra CW-MFC System 3.3.1.Effect of Different DO Concentrations on Cr Content in the Substrate

Figure 4 .
Figure 4.The content and proportion of Cr following chemical extraction from the L. hexandra CW-MFC system under different DO concentrations: (a) Cr content of the anode; (b) Cr proportion of the anode; (c) Cr content of the cathode; and (d) Cr proportion of the cathode.

Figure 4 .Figure 5 .
Figure 4.The content and proportion of Cr following chemical extraction from the L. hexandra CW-MFC system under different DO concentrations: (a) Cr content of the anode; (b) Cr proportion of the anode; (c) Cr content of the cathode; and (d) Cr proportion of the cathode.

Figure 5 .
Figure 5.The content and proportion of chemical-extracted Cr in the roots of L. hexandra under different DO concentrations: (a) Cr content in L. hexandra roots; and (b) Cr proportion in L. hexandra roots.

Figure 6 .
Figure 6.The content and proportion of chemically extracted Cr in the stems of L. hexandra under different DO concentrations: (a) Cr content in L. hexandra stems; and (b) Cr proportion in L. hexandra stems.

Figure 7 .
Figure 7.The content and proportion of chemically extracted Cr in the leaves of L. hexandra under different DO concentrations: (a) Cr content in L. hexandra leaves; and (b) Cr proportion in L. hexandra leaves.

Figure 6 .
Figure 6.The content and proportion of chemically extracted Cr in the stems of L. hexandra under different DO concentrations: (a) Cr content in L. hexandra stems; and (b) Cr proportion in L. hexandra stems.

Figure 6 .
Figure 6.The content and proportion of chemically extracted Cr in the stems of L. hexandra under different DO concentrations: (a) Cr content in L. hexandra stems; and (b) Cr proportion in L. hexandra stems.

Figure 7 .
Figure 7.The content and proportion of chemically extracted Cr in the leaves of L. hexandra under different DO concentrations: (a) Cr content in L. hexandra leaves; and (b) Cr proportion in L. hexandra leaves.

Figure 7 .
Figure 7.The content and proportion of chemically extracted Cr in the leaves of L. hexandra under different DO concentrations: (a) Cr content in L. hexandra leaves; and (b) Cr proportion in L. hexandra leaves.

Figure 8 .
Figure 8. Relative abundance map of the L. hexandra CW-MFC microbial community composition at the phylum level.

Figure 8 .
Figure 8. Relative abundance map of the L. hexandra CW-MFC microbial community composition at the phylum level.

Figure 9 .
Figure 9. Relative abundance map of the L. hexandra CW-MFC microbial community composition at the genus level.

4. 3 .
The Effect of Cathode DO Concentration on the Valence Change and Morphological Characteristics of Cr

Table 1 .
Cr content in the substrate at different DO concentrations.

Table 1 .
Cr content in the substrate at different DO concentrations.

Table 2 .
Cr content in L. hexandra at different DO concentrations.Chemical Speciation Analysis of Cr in the SubstrateThe contents of the BCR-extracted Cr residual state, oxidizable state, reducible state, and weak acid extractable state reached the highest levels when the DO concentration was 4.5 mg•L −1 (Figure4).The anode values were 6447.01 mg•kg −1 , 3795.45 mg•kg −1 , 2385.42 mg•kg −1 , and 2304.21mg•kg−1 , respectively.The cathode values were 6668.86 mg•kg −1 , 4301.67 mg•kg −1 , 2786.99 mg•kg −1 , and 2230.58mg•kg−1 , respectively.The effective Cr state decreases to a minimum of 31% and the stable state is 69%.This indicated that the increased DO concentration affected the transformation of various forms of Cr, and helped the transformation of Cr from an effective state to a stable state.3.4.2.Chemical Speciation Analysis of Cr in L. hexandra Total Chromium (mg•kg−1 ) Cr (VI) (mg•kg −1 ) Cr (III) (mg•kg −1 )

Table 3 .
Statistics of the alpha diversity index.

Table 3 .
Statistics of the alpha diversity index.