Set-up of MDC-CW in laboratory
A column (with a length of 1.2 m and a diameter of 0.3 m) was established in the laboratory. Quartz sand was added in the column as the substrate in three layers (the 50-cm thick upper layer of quartz sand with the particle diameter of 4 mm ~ 6 mm; the 50-cm thick middle layer of quartz sand with the particle diameter of 2 mm ~ 3 mm; the 10-cm thick bottom layer of quartz sand with the particle diameter of 4 mm ~ 6 mm). Electrodes (graphite felt, 0.15 m × 0.15 m) were put in the column at the middle between the middle and upper layers. The vertical distance from the anode to the bottom of the column is 35 cm and the vertical distance from the cathode to the bottom of the column is 85 cm. The vertical distance between the anode and cathode is 50 cm.
A circuit was set to connect the external resistance with the voltage collector in order to monitor the power generation efficiency of the column device in real time. The ion exchange membrane (HACH USA, 0.15 m × 0.15 m) was put between the two electrodes. The anion exchange membrane (AEM) was close to the cathode, which was 61 cm away from the bottom of the column. The cation exchange membrane (CEM) was close to the anode, which was 48 cm away from the bottom of the column. The vertical distance between the two ion exchange membranes was 24 cm. Plants (canna indica and Acorus calamus) were planted in the column in such a way that the roots were arranged at a position close to the cathode as possible.
Water inlet was set on one side at the height 115 cm above the bottom of the column. Water outlet was set on the other side at the height 5 cm above the bottom of the column. Five sampling ports were set at the horizontal positions of anode, cathode, AEM and CEM, as well as the junction between the upper layer and the middle matrix. The device is shown in Fig. 1.
Six devices were established. Three devices with canna were set and named MDC-CW-P1 and the other three devices with calamus was named MDC-CW-P2. In addition, six vertical CWs with the same size were established in triplicate and named CK-CW-P1 and CK-CW-P2.
System operation
Environmental conditions were described below. The experimental period was from April to August, 2019. The average temperature was 20 °C ~ 40 °C and the relative humidity was 60% ~ 95%. The experiments were launched at Shanghai Ocean University.
Microbial inoculation acclimation was performed as follows. The inoculum of the device was the mixture taken from the natural water body along the Dongtan Coast in Shanghai, the sludge of treatment for sea foods breeding and wastewater of brine products. The inoculum was added into the devices according to the ratio of 4:1 (v/v, inoculum/tap water) for 3 times within 60 days.
According to Pollutant Discharge Standard of Municipal Wastewater Treatment (GB18918-2016), the simulated sewage was prepared with 500 mg/L COD (glucose), TN 60 mg/L (30 mg/L NH4+-N and 30 mg/L NO3--N) in tap water and stored in a storage tank (200 L). The salinity of sewage was 5 g/L (sodium chloride) [9]. Simulated sewage was simultaneously pumped into each experimental device.
After the inoculation period, the influent and effluent were continuously sampled for the determination of water quality indicators and the devices were discharged once every 7 days within 70 days.
Sampling and testing
Water samples
Influent and effluent were collected from the sampling port of each device in triplicate and stored into 50-mL bottles. After sampling, the physical and chemical indicators (temperature, oxidation reduction potential (ORP), dissolved oxygen (DO) and pH) were tested immediately with multiple parameter instrument (HACH, USA). Water samples were filtered (0.22-μm filter membrane) and then COD, TN, NH4+-N, NO3--N and NO2--N concentrations were determined according to the Standard Determination Method of Water and Wastewater by America health and Public Association (AHPA) in 2012. The performance of power generation was monitored by a data collector. The salinity was also determined with tap water as correcting samples. The desalination performance was calculated with salt removal rate.
Microbial samples
A total of 20 g of quartz sand samples were collected at the bottom of each device. Plant roots in triplicate devices were harvested and mixed. The electrodes in triplicate devices were collected as well. The surface biofilm of quartz sand, roots and electrodes were exfoliated into the solution by ultrasonic vibration. Then 0.22-μm filtration membrane was used to filter the microbial solution and the microbial DNA information on the membrane was extracted, amplified by PCR, and then measured. Operational taxa unit (OTU) was used to express the microbial abundances.
Electrode and membrane samples
The electrode and ion exchange membrane were treated for the observation under a scanning electron microscope (SEM). The electrode sample was placed in a glutaraldehyde solution (2.5%, pH 6.8) and allowed to stand for 12 h at 4 °C. Then, the samples were washed for 3 times (15 min each time) with PBS buffer solution (0.1 M, pH 6). Dehydration was performed for 15 min with 25%, 50%, 70%, 85%, 95% and 100% ethanol, respectively. The samples were treated with the mixture of 100% ethanol and isoamyl acetate (v/v, 1:1) for 30 min and then displaced with pure isoamyl acetate for 2 h. Finally, the samples were dried in a freeze-dryer.
Data analysis
The data of water quality samples measured in the experiments were analyzed in SPSS and Prism (ANOVA test with P ≤ 0.05). The experimental data were the mean values and the error bars were plotted with the water quality data from repeated experiments.
Polarization curve and power density curves were plotted according to the following steps. Firstly, the MDC-CW devices were set as the open circuit voltage mode, so that open circuit potential (OCV) was obtained after stabilization. The resistor was then plugged in and the resistance was incrementally increased from 5 to 10000 Ω (each resistance value was maintained for 30 min). Then, current (I) was calculated as.
where U indicates voltage (v) and R indicates resistance (Ω).
Finally, the polarization curve was obtained with the values of voltage and current and the linear part of the curve was fitted. The linear slope represents the total internal resistance of the MDC-CW devices.
where P indicates the power density; V indicates the effective working area of electrode (m3).