A Bio-Electro-Fenton System Employing the ... - Semantic Scholar

materials Article

A Bio-Electro-Fenton System Employing the Composite FePc/CNT/SS316 Cathode Yi-Ta Wang * and Ruei-Shiang Wang Department of Mechanical and Electro-Mechanical Engineering, National Ilan University, Yilan 26047, Taiwan; [email protected] * Correspondence: [email protected]; Tel.: +886-3-931-7457 Academic Editor: Douglas Ivey Received: 11 November 2016; Accepted: 8 February 2017; Published: 13 February 2017

Abstract: Bio-electro-Fenton microbial fuel cells generate energy through the decomposition of organic matter by microorganisms. The generated electricity drives a Fenton reaction in a cathode chamber, which can be used for the decolorization of dye wastewater. Most of the previous works added expensive platinum catalyst to improve the electrical property of the system. In this research, aligned carbon nanotubes (CNTs) were generated on the surface of SS316 stainless steel by chemical vapor deposition, and an iron phthalocyanine (FePc) catalyst was added to fabricate a compound (FePc/CNT/SS316) that was applied to the cathode electrode of the fuel cell system. This was expected to improve the overall electricity generation efficiency and extent of decolorization of the system. The results showed that the maximum current density of the system with the modified electrode was 3206.30 mA/m2 , and the maximum power was 726.55 mW/m2 , which were increased by 937 and 2594 times, respectively, compared to the current and power densities of a system where only the SS316 stainless steel electrode was used. In addition, the decolorization of RB5 dye reached 84.6% within 12 h. Measurements of the electrical properties of bio-electro-Fenton microbial fuel cells and dye decolorization experiments with the FePc/CNT/SS316 electrode showed good results. Keywords: bio-electro-Fenton microbial fuel cells; iron phthalocyanine; carbon nanotubes

1. Introduction A microbial fuel cell is a cell module that converts chemical energy to electrical energy with the help of microbial strains [1]. A typical double-chamber microbial fuel cell contains an anode chamber and a cathode chamber, and employs a proton exchange membrane between the anode and cathode for separation. Air flowing into the cathode chamber is blocked by this proton exchange membrane and does not affect the electricity generation of the microorganisms at the anode. The protons generated by the anode can be transmitted through the proton exchange membrane located in the middle to the cathode, and the cathode chamber can generate water by combining the electrons and protons transmitted from the external circuit with oxygen. The electrochemical reactions that occur during the operation of the microbial fuel cell are similar to those that occur in common methanol fuel cells, and are shown in Equations (1) and (2), with glucose as the nutrient [2]: C6 H12 O6 + 6H2 O → 6CO2 + 24H+ + 24e− ,

(1)

O2 + 4H+ + 4e− → 2H2 O,

(2)

Bio-electro-Fenton microbial fuel cells integrate microbial fuel cell and E-Fenton technologies, where the anode chamber generates electrons and hydrogen ions by the decomposition of organic matter with microbial strains to drive the cathode Fenton reaction, thus degrading sewage and

Materials 2017, 10, 169; doi:10.3390/ma10020169

www.mdpi.com/journal/materials

Materials 2017, 10, 169

2 of 15

wastewater solutions that the microorganisms cannot degrade. The electron-Fenton reaction pathway is described in Equations (3)–(6): Fe2+ + H2 O2 → ·HO + OH− + H2 O2 → Fe3+ + ·HO + OH− ,

Fe2+

(3)

HO + Fe2+ → Fe3+ + OH− ,

(4)

Fe3+ + e− → Fe2+ ,

(5)

2H+ + O2 + 2e− → H2 O2 ,

(6)

A study by Feng showed that the energy generated by the Shewanella anode of a microbial fuel cell enabled the cathode to generate ferrous ions and hydrogen peroxide. The generated hydroxyl radical degraded methyl orange dye, with 100% degradation being achieved within 14 h [3]. Fernández de Dios studied the decolorization of five different types of dye wastewater using a microbial fuel cell, and found that the decolorization percentages of four of the five types were greater than 88% [4]. Wang treated toxic sewage using bio-electro-Fenton microbial fuel cells that generated hydrogen peroxide and ferrous ions. The results showed that the highly toxic trivalent arsenic ion was converted into the pentavalent arsenic ion, which has lower toxicity [5]. Zhang applied bio-electro-Fenton microbial fuel cells to the macromolecular aqueous solutions generated during dye degradation and various pharmaceutical processes, and the degradation efficiency after 9 h was 70% [6]. This showed that bio-electro-Fenton microbial fuel cells do not need continuous additions of the Fenton reagent, and the iron source in the system has a self-sustaining mechanism, which is conducive to the decolorization and degradation of various types of dye wastewater, toxic waste solutions, and polymer solutions. When selecting the electrode materials for bio-electro-Fenton microbial fuel cells, the anode materials should have (1) good conductivity and low impedance; (2) excellent biocompatibility; (3) stable chemical properties; (4) good corrosion resistance; and (5) good mechanical properties [7–10]. The cathode electrode needs to have good electrochemical properties because a strong oxidation-reduction ability may be conducive to electron transport. The best result is usually achieved when platinum is added as the catalyst. This is mainly because platinum can reduce the cathode activation energy and improve the reaction rate [11]. In Moon’s research, the density of the power generated by a graphite electrode coated with platinum was 150 mW/m2 , which was three times the energy generated by a pure graphite electrode. This proved that platinum can improve the electrocatalytic activity and reduce the barrier caused by the activation of the over-potential [12]. However, platinum is a precious metal, which may increase the fabrication cost of a microbial fuel cell. Many studies have been conducted with the goal of reducing the proportion of platinum, or finding other catalysts as substitutes for platinum. Transition metallic macrocycles are often used as catalysts in studies on the cathode electrode of a microbial fuel cell because they have good oxidation-reduction abilities, which improve the efficiency and lower the cost compared to the use of platinum [13–15]. Macrocyclic metal complexes have good oxidation–reduction activities and chemical stabilities. The planar structure of a macrocycle increases the electron density of the central atom and improves the conductivity [16]. Therefore, most studies have employed iron phthalocyanine (iron (II) phthalocyanine, FePc) and cobalt tetramethoxyphenylporphyrin (CoTMPP) as catalysts. Zhao et al. coated carbon felt with FePc and CoTMPP as catalysts to produce a cathode electrode, and found that under a neutral pH, the current density increased from 0.40 to 0.97 mA/cm2 , which proved that transition metallic macrocycles have good oxidation–reduction abilities [17]. Deng et al. employed a combination of CoTMPP and FePc as a macrocycle catalyst, and found that the power density with the Co/Fe/macrocycles was 751 mW/m2 . This was 1.5 times the power density with the Pt/C catalyst under the same conditions, which showed that transition metallic macrocycles have excellent electrocatalytic activities [18].

Materials 2017, 10, 169

3 of 15

Carbon nanotubes (CNTs) have properties such as a high specific surface area, high mechanical strength, and ductility, as well as excellent chemical stability and conductivity, which make them a promising pad material. Wang used Mo2 C/CNTs as the catalyst when investigating the performance of a microbial fuel cell system. When using a carbon felt electrode modified by Mo2 C/CNTs with 16.7 wt % Mo, the density of the power generated by the system was 1050 mW/m2 , which was six times the density when employing only CNTs or Mo2 C [19]. In a study by Ghasemi et al. a polymer polypyrrole/CNT composite material on carbon paper generated power at 113.5 mW/m2 , which was higher than that with unmodified carbon paper (69.1 mW/m2 ) [20]. This proved that CNTs have good electrocatalytic performances in microbial fuel cells. In comparison with carbon nanotubes, it is found that carbon black particles easily cluster, which reduces the utilization rate of the platinum catalyst. In addition, the association of platinum and carbon black is unstable. Gradual separation may occur during the operation of the fuel cell system over a long period, which reduces the overall performance [21]. Shaijumon compared the performances of proton exchange membrane fuel cells (PEMFCs) using carbon black and CNTs as carriers, with the same weight percentage of platinum catalyst. The results showed that when the platinum catalyst was combined with the CNTs, the density of the current generated by the system was 535 mA/m2 , which was higher than that obtained (310 mA/m2 ) when the platinum catalyst was loaded on carbon black. This proved that the platinum catalyst was distributed more uniformly on the surface of the CNTs, making their oxidation–reduction ability better than that of the catalyst obtained by combining carbon black and platinum [21]. In this research, it was expected that the utilization rate of the serialized CNT catalyst in a bio-electro-Fenton microbial fuel cell and the overall efficiency of the system could be improved using an iron phthalocyanine compound with aligned CNTs on the surface of a stainless steel (SS316) electrode. 2. Materials and Methods In this research, we employed FePc/CNTs on an SS316 pad as the cathode electrode in a bio-electro-Fenton microbial fuel cell, and the overall electrical properties of the system and the dye decolorization were investigated. 2.1. Electrode Production A 30 mm × 50 mm × 5 mm piece of carbon felt (CeTech Co., Ltd., Taichung, Taiwan) was employed as the electrode for the anode. Hydrogen peroxide and deionized water (in a 1:9 ratio) were poured into a beaker and heated to 90 ◦ C, and a hydrophilic treatment was performed for 3 h. SS316 (Qunlong Steel Industry Co., Ltd., Yilan, Taiwan) was employed as the cathode electrode. The composition of SS316 is provided in Table 1. The electrode size was 20 mm × 30 mm × 2 mm. The impurities on the surface of the stainless steel were removed with various grades of sandpaper used in the following sequence: 100, 240, 400, and 600. The ground stainless steel electrode was then placed in an acetone solution, removed after 30 min of ultrasound cleaning, wiped, and dried. Table 1. Analysis results for SS316 components. Component

Si

Mo

Cr

Mn

Content (%)

0.86

2.23

17.78 1.09

Fe

Ni

67.41 10.63

2.2. The Fabricaiton Process of Composite FePc/CNT/SS316 Cathode A uniform 0.5 mL drop of iron pentacarbonyl (Sigma-Aldrich, Saint Louis, MO, USA) on the bare SS316 plate was placed by pipet. The iron pentacarbonyl serves as the catalyst for CNT synthesis. Thermal pyrolysis chemical vapor deposition (TPCVD) was then employed to deposit CNT on the SS316 plate. Prior to the CNT synthesis process, the TPCVD furnace was heated and vacuumed to 750 ◦ C and 50 mTorr, respectively, and followed by purging with nitrogen gas of 10 sccm. Then, the furnace was filled with acetylene gas of 50 sccm and/or argon gas of different flow rate, 100, 500, or 800 sccm,

Materials 2017, 10, 169

4 of 15

to start the CNT synthesis process. The samples were then naturally cooled to room temperature after the previous CNT synthesis process. A composite CNT/SS316 cathode electrode was completed at this stage. A stirred solution of 0.018 g iron phthalocyanine (FePc) and 1 mL tetrahydrofuran (THF) was then dropped on the composite CNT/SS316 cathode plate and subjected to ultrasonic oscillation for 30 min to form a uniform layer on the composite CNT/SS316 cathode plate. After baking in an oven of 80 ◦ C for 24 h, the composite FePc/CNT/SS316 cathode electrode was completed. 2.3. Construction of Bio-Electro-Fenton Microbial Fuel Cell System Materials 2017, 10, 169

4 of 15

The cathode and anode chambers of the double-chamber body employed in this system had 2.3. Construction of Bio-Electro-Fenton Microbial Fuel Cell System a total capacity of 130 mL, with a DuPont™ Nafion® PFSA N-117 (DuPont Limited, Wilmington, DE, The cathode and anode chambers of the double-chamber body employed in this system had a USA) proton exchange membrane placed between them to perform the mass transfer reaction in the total capacity of 130 mL, with a DuPont™ Nafion® PFSA N-117 (DuPont Limited, Wilmington, DE, microbial fuel cell, as shown Figureplaced 1. The anode solution wastewater fromin dairy products USA) proton exchange in membrane between them to performwas the mass transfer reaction the microbial fuel cell, as shown Figure 1. The anode was solution wastewater was from dairy (Zhong Shumao Laboratory, Yilan,inTaiwan), and thesolution cathode a 0.3products M iron (II) sulfate (Zhong Shumao Laboratory, Yilan, Taiwan), and the cathode solution was a 0.3 M iron (II) sulfate (FeSO4 ·7H2 O, Japan Reagent, Tokyo, Japan) solution mixed with active carbon black dye at 1000 ppm. (FeSO4·7H2O, Japan Reagent, Tokyo, Japan) solution mixed with active carbon black dye at 1000 ppm.

Figure 1. Experimental setup (left: anode chamber, right: cathode chamber).

Figure 1. Experimental setup (left: anode chamber, right: cathode chamber). 2.4. Measurement of Electrical Quantities study, theQuantities electrical quantities were measured using an electrochemical workstation (Zive 2.4. MeasurementInofthis Electrical SP1, WonATech, Seoul, Korea) and Zive SM (WonATech). The cathode and anode of the microbial fuel cell were to the electrochemical workstation to measure voltage, current, and the workstation In this study, the connected electrical quantities were measured usingthean electrochemical cell polarization curve of the cell. The changes in these parameters were simultaneously recorded (Zive SP1, WonATech, Seoul, Korea) and Zive SM (WonATech). The cathode and anode of the microbial using a data acquisition system (Jiehan-5020, Jiehan Technology Corp., Taichung, Taiwan). fuel cell were connected the electrochemical workstation measure voltage, current, and the The constantto current measurement was used to calculate theto impedance lossthe within the microbial fuel cell by controlling current, the overall of the cell polarization potential (Ecell) cell polarization curve of the the cell. The and changes in calculation these parameters were simultaneously recorded shown in Equation (7): using a data isacquisition system (Jiehan-5020, Jiehan Technology Corp., Taichung, Taiwan). Ecell =used En − ηohm ηact − ηconc, the impedance loss within (7) the microbial The constant current measurement was to −calculate Here, ηohm the ohmic and polarization potential loss, ηact is the polarization potential fuel cell by controlling theis current, the overall calculation of activation the cell polarization potential (Ecell ) is loss, and ηconc is the concentration polarization potential loss. shown in Equation (7): To calculate the power density, the power (P) can be obtained from the relationship between the =shown En − inηohm − η(8). ηconc (7) voltage (Vcell) and the currentE(I), Equation When an ,external load resistance (Rext) act − cellas was added to the cell, the power (P) and current (I) could be obtained from the measured voltage

Here, ηohm theEquations ohmic (9)–(11): polarization potential loss, ηact is the activation polarization potential (Vcell)is using loss, and ηconc is the concentration polarization Ppotential loss.  IVcell , (8) To calculate the power density, the power (P) can be obtained from the relationship between the (9) Vcell R ext , voltage (Vcell ) and the current (I), as shown inI Equation (8). When an external load resistance (Rext ) was added to the cell, the power (P) and currentP(I) (10) voltage (Vcell )  I 2could  R ext , be obtained from the measured using Equations (9)–(11): 2 P  Vcell R ext , (11) P = IVcell , (8) The power calculation was analyzed based on the polarization curve, with the system potential measured by controlling the current andI calculating the (P) output of the system using the = Vcell /R , extpower current-voltage correlation. Cyclic voltammetry was conducted using three electrodes. The auxiliary electrode was made of platinum, and the reference an, Ag/AgCl electrode. The working electrode P = I2was ·Rext was placed at the location of the electrochemical reaction of the electrode, and the scanning rate was 10 mV. P = V2cell /Rext ,

(9) (10) (11)

The power calculation was analyzed based on the polarization curve, with the system potential measured by controlling the current and calculating the power (P) output of the system using the

Materials 2017, 10, 169

5 of 15

current-voltage correlation. Cyclic voltammetry was conducted using three electrodes. The auxiliary electrode was made of platinum, and the reference was an Ag/AgCl electrode. The working electrode was placed at the location of the electrochemical reaction of the electrode, and the scanning rate was 10 mV. The anolyte solution contained the mixed solution of dairy wastewater, including yogurt and milk.Materials The anolyte solution contained microorganisms. We filled the anode chamber with half 2017, 10, 169 5 of 15of the anolyte solution and half of glucose. Then we connected the anode electrode with a 1 kΩ resistor The anolyte solutiondrop contained thethe mixed solution of dairy yogurt and and monitored the voltage across resistor. After a fewwastewater, hours, theincluding voltage started to drop. milk. The anolyte solution contained microorganisms. We filled the anode chamber with half of the Once it fell to approximately 5 mV, we replaced the anolyte solution completely. We repeated the anolyte solution and half of glucose. Then we connected the anode electrode with a 1 kΩ resistor and above processes until the voltage reached at least 100 mV, at which point the microorganism incubation monitored the voltage drop across the resistor. After a few hours, the voltage started to drop. Once it was complete. After that,5 we the resistor andsolution startedcompletely. up the bio-electro-Fenton fell to approximately mV,removed we replaced the anolyte We repeated the system above for one-day discharging. The start-up wasmV, completed at this stage. We reconnected thewas resistor processes until the voltage reachedprocess at least 100 at which point the microorganism incubation and continued monitoring voltage during the operation of the bio-electron-Fenton system. complete. After that, we the removed the drop resistor and started up the bio-electro-Fenton system for onedischarging. The start-up process completed at this We reconnected theglucose. resistor and Onceday it fell to approximately 50 mV, we was partially replaced thestage. anolyte solution with continued monitoring the voltage drop during the operation of the bio-electron-Fenton system. Once

2.5. Preparation for Dye Decolorization it fell to approximately 50 mV, we partially replaced the anolyte solution with glucose. The dye wastewater employed in the cathode chamber in this research was active carbon black 2.5. Preparation for Dye Decolorization (RB5). It had an absorption wavelength of 595 nm, a molecular formula of C26 H25 N5 O19 S6 ·4Na, The dye wastewater employed in the cathode chamber in this research was active carbon black and a molecular weight of 991.82. It contained a nitrogen-nitrogen double bond (N=N), which could (RB5). It had an absorption wavelength of 595 nm, a molecular formula of C26H25N5O19S6·4Na, and a be broken by the electron-Fenton reaction, resulting in a decolorization effect after the reduction of the molecular weight of 991.82. It contained a nitrogen-nitrogen double bond (N=N), which could be chroma of RB5 within the cathode chamber. This in proved that a continuous electron-Fenton broken by the electron-Fenton reaction, resulting a decolorization effect after the reduction of reaction the occurred. RB5 is a commonly used dye in the experimental detection of dye in wastewater. chroma of RB5 within the cathode chamber. This proved that a continuous electron-Fenton reactionIn the experiment, comparison of theused concentration and absorbance values RB5 was performed occurred.a RB5 is a commonly dye in the experimental detection of of dye in wastewater. In theusing a spectrophotometer (SH-U880, Shishin Technology Taipei, Taiwan), variation of the experiment, a comparison of the concentration andCorp., absorbance values of RB5and wasthe performed using a RB5 spectrophotometer (SH-U880, ShishininTechnology Corp., Taipei, Taiwan), and the variation of the by concentration vs. absorbance is shown Figure 2. The decolorization experiment was performed RB5 concentration absorbance is shown 2. The decolorization experiment was performed (D) diluting 1.5 mL of vs. electrolyte and 1.5 mLinofFigure distilled water in the sampling chamber and collecting by (D) diluting 1.5 mL of electrolyte and 1.5 mL of distilled water in the sampling chamber was and used samples on an hourly basis for 12 h. Before and after the experiment, the spectrophotometer collecting samples on an hourly basis for 12 h. Before and after the experiment, the spectrophotometer to make comparisons between the visible light and the absorbance value, as shown in Equation (12): was used to make comparisons between the visible light and the absorbance value, as shown in Equation (12): Ai − At

D= D

AiAAt Ai

i

× 100%

100 %,

(12)

(12)

Ai = A Initial absorbance value, absorbance value. i = Initial absorbance value,AA Measured absorbance value t t== Measured

Figure 2. Plot of RB5 concentrationvs. vs. absorbance solution analyzed usingusing Figure 2. Plot of RB5 concentration absorbancevalue valueininaqueous aqueous solution analyzed a spectrophotometer. a spectrophotometer.

The experimental results of this research, including analyses of the cyclic voltammograms,

The experimental results ofdischarge, this research, including analyses of the voltammograms, analyses of the fixed resistance the electrochemical analysis, and the cyclic decolorization, were analyses of with the standard fixed resistance discharge, the electrochemical analysis, and theand decolorization, plotted graphics software (Kaleida Graph, K-graph, Synergy Software), the error in the values was in the range of ±5%.

Materials 2017, 10, 169

6 of 15

were plotted with standard graphics software (Kaleida Graph, K-graph, Synergy Software), and the error in the values was in the range of ±5%. Materials 2017, 10, 169

6 of 15

3. Results and Discussion 3. Results and Discussion Materials 2017, 10, 169

6 of 15

3.1. Morphology of CNT Formation 3.1. Morphology of CNT Formation 3. Results and Discussion

In the In present workwork thermal pyrolysis chemical vapor deposition to deposit depositCNT CNTonon the the present thermal pyrolysis chemical vapor deposition(TPCVD) (TPCVD) to the Morphology SS316steel stainless steel plate was employed. The TPCVD furnace was heated vacuumed SS316 3.1. stainless was employed. The TPCVD furnace was heated andand vacuumed to to 750 ◦ C ofplate CNT Formation °C and 50 mtorr, respectively, followed by purging with nitrogen gas10 at sccm. 10 sccm. Then,the the furnace furnace was and 50 750 mTorr, respectively, by purging with vapor nitrogen gas at Then, In the present workfollowed thermal pyrolysis chemical deposition (TPCVD) to deposit CNT on was filled with acetylene gas at 50 sccm to start the CNT deposition process for 20 min. Figure 3a filled with acetylene gas steel at 50plate sccmwas to start the CNT process forheated 20 min. Figure 3a shows the SS316 stainless employed. Thedeposition TPCVD furnace was and vacuumed to the shows the field-emitted SEM image, showing that the resulting CNT is randomly alignment, referred field-emitted SEM showing that the resulting CNT is randomly as RACNT 750 °C and 50 image, mtorr, respectively, followed by purging with nitrogen gasalignment, at 10 sccm. referred Then, theto furnace to as RACNT in the following text. The X-ray diffraction, Figure 3b, shows that two intensity was filled with acetylene gas at 50 sccm to start the CNT deposition process for 20 min. Figure in the following text. The X-ray diffraction, Figure 3b, shows that two intensity absorption3apeaks absorption peaks occurred at 2θ of 26.0° and 43.0°, respectively, which were exactly the absorption ◦ shows the field-emitted SEM◦ image, showing that the resulting CNT is randomly alignment, referred occurred at 2θ 26.0 peaks of of CNT [22].and 43.0 , respectively, which were exactly the absorption peaks of CNT [22]. to as RACNT in the following text. The X-ray diffraction, Figure 3b, shows that two intensity absorption peaks occurred at 2θ of 26.0° and 43.0°, respectively, which were exactly the absorption peaks of CNT [22].

Figure 3. (a) The field emitted SEM images of carbon nanotubes is random alignment magnified

Figure 3. (a) The field emitted SEM images of carbon nanotubes is random alignment magnified 30,000 30,000 times; (b) XRD patterns of RACNT. times; (b) XRD patterns of RACNT. Figure 3. (a) The field emitted SEM images of carbon nanotubes is random alignment magnified Figure 4a shows the surface morphology of the CNTs generated on the surface of the stainless 30,000 times; (b) XRD patterns of RACNT.

steel within 20 min, observed using SEM, when flow generated velocity of the gas wasof100 Figure 4a shows theassurface morphology of thethe CNTs on argon the surface thesccm, stainless the acetylene flow velocity was 10 sccm, andwhen the temperature was 750 of °C.the Thisargon showsgas thatwas there100 are steel within 20 min, as observed using SEM, the flow velocity Figure 4a shows the surface morphology of the CNTs generated on the surface of the stainless sccm, agglomerates of CNTs on the stainless steel surface, with no CNTs generated in other forms. A SEM ◦ the acetylene flow was 10 sccm, and the temperature was 750 Thisgas shows that there are steel within 20velocity min, as observed using SEM, when the flow velocity of theC. argon was 100 sccm, image at a magnification of 15,000 times is shown in Figure 4b, where it is seen that impurities of agglomerates of CNTs the stainless steeland surface, with no CNTs generated in other A SEM the acetylene flow on velocity was 10 sccm, the temperature was 750 °C. This shows thatforms. there are non-tubular carbon appear in the spaces between the CNT agglomerates. These impurities were agglomerates of CNTs on the stainless steel surface, with no CNTs generated in other forms. A SEM image generated at a magnification of the 15,000 times is shown Figure 4b,a portion where it that impurities of because after decomposition at high in temperature, of is theseen carbon ions within image at a magnification of the 15,000 times between is shown in Figure 4b, where it is seenThese that impurities of were non-tubular carbon appear in spaces the CNT agglomerates. impurities the quartz tube clustered together, giving an amorphous carbon deposition in parallel with the non-tubular carbon appear in the spaces between the CNT agglomerates. These impurities were generated because decomposition high temperature, a portion of the carbonstuck ions in within generation ofafter CNTsthe with the catalyst onat the surface of the stainless steel, with particles the the generated because after the decomposition at high temperature, a portion of the carbon ions within between the nanotubes [23].an amorphous carbon deposition in parallel with the generation quartz spaces tube clustered together, giving the quartz tube clustered together, giving an amorphous carbon deposition in parallel with the of CNTs with the catalyst on the surface with particles stuck in thestuck spaces between generation of CNTs with the catalyst of onthe thestainless surface ofsteel, the stainless steel, with particles in the the nanotubes [23]. spaces between the nanotubes [23].

Figure 4. SEM images of carbon nanotubes prepared using argon gas at a flow rate of 100 sccm: (a) Magnified 3000 times; (b) Magnified 15,000 times. Figure 4. SEM images of carbon nanotubes prepared using argon gas at a flow rate of 100 sccm:

Figure 4. SEM images of carbon nanotubes prepared using argon gas at a flow rate of 100 sccm: (a) Magnified 3000 times; (b) Magnified 15,000 times. (a) Magnified 3000 times; (b) Magnified 15,000 times.

Materials 2017, 10, 169 Materials 2017, 10, 169 Materials 2017, 10, 169

7 of 15 7 of 15 7 of 15

The morphology observed by SEM (Figure 5a) showed that CNTs with impurities were generated The morphology observed SEM (Figure 5a) showed that CNTs with impurities of the stainless steelby electrode within 20 min when the argon gas flowwere was 500 sccm on the surface rate generated on the surface of the stainless steel electrode within 20 min when the argon gas flow rate was 500 sccm and the acetylene acetylene flow rate was 10 sccm at 750 ◦°C. In addition to the CNTs and impurities, sparse C. and impurities, a sparse and the acetylene flow rate was 10 sccm at 750 °C. In addition to the CNTs and impurities, a sparse distribution of individual tubes was found. The The agglomeration agglomeration was was relatively relatively low, and disordered distribution of individual tubes was found. The agglomeration was relatively low, and disordered CNTsCNTs werewere formed, as shown in Figure 5b. CNTs were formed, as shown in Figure 5b. formed, as shown in Figure 5b.

Figure 5. SEM graphs carbon nanotubes generated using argon gas flow velocity of 500 sccm:

Figure 5. SEM graphs carbon nanotubes generated using argon gas flow velocity of 500 sccm: (a) 5. Magnified 5000 times; (b) Magnified 50,000 times. using argon gas flow velocity of 500 sccm: Figure SEM graphs carbon nanotubes generated (a) Magnified 5000 times; (b) Magnified 50,000 times. (a) Magnified 5000 times; (b) Magnified 50,000 times. Figure 6a shows the morphological observation with SEM of the CNTs generated on the stainless steel surface within min when the argon gas flowwith rate SEM was 800 sccm and the acetylene rate Figure 6a shows the20morphological morphological observation of the the CNTs generated on the stainless stainless Figure 6a shows the observation with SEM of CNTs generated onflow the was 10 sccm at 20 750min °C when Figurethe 6. It is clearly seenrate that,was with ansccm argonand flow rate of 800 sccm, the was steel surface within argon gas flow 800 the acetylene flow rate steel surface within 20 min when the argon gas flow rate was 800 sccm and the acetylene flow rate generated CNTs are widely distributed on the surface of the stainless steel without agglomeration. ◦ 10 sccm at 750 C Figure 6. It is clearly seen that, with an argon flow rate of 800 sccm, the generated was 10 sccm at 750 °C Figure 6. It is clearly seen that, with an argon flow rate of 800 sccm, the Moreover, it can be observed that the CNTs were generated in an upward direction. An image with CNTs are widely distributed on the surface of the stainless steel without agglomeration. Moreover, generated CNTs are widely on the surface the stainless steel without a magnification of 15,000 distributed times is shown in Figure 6b,ofand no impurity particles are agglomeration. seen in the it can be observed that the CNTs were in an upward direction. An image with aAn magnification Moreover, it canbetween be observed that thegenerated CNTs were generated in an upward direction. image interspaces the nanotubes. The generated CNTs are arranged in an orderly way, with their with of 15,000 times is shown in Figure 6b, and no impurity particles are seen in the interspaces between the a magnification of 15,000 times is shown in Figure andduring no impurity particles are the seen in the tips slightly bent. This was because of the influence of 6b, gravity their generation. Thus, tops nanotubes. The generated CNTs are arranged in an orderly way, with their tips slightly bent. This was of the CNTs provided mutual support each other. interspaces between the nanotubes. The for generated CNTs are arranged in an orderly way, with their

because of the influence gravityof during their generation. Thus, the tops of the CNTs tips slightly bent. This wasofbecause the influence of gravity during their generation. Thus,provided the tops mutual support for each other. of the CNTs provided mutual support for each other.

Figure 6. SEM images of carbon nanotubes formed using argon gas flow rate of 800 sccm: (a) Magnified 2000 times; (b) Magnified 15,000 times.

The above results show that impurities may hinder the generation of CNTs [18], and the flow Figure 6. SEM carbon formed upon using argon gas flow of rate of 800 sccm: the (a) Magnified velocity theimages argon of gas has annanotubes absolute impact generation CNTs. argon gas Figure 6. of SEM images of carbon nanotubes formed using the argon gas flow rate of 800When sccm: (a) Magnified 2000 times; (b) Magnified 15,000 times.CNT structure had a lower probability of curling, along with flow rate was higher, the generated 2000 times; (b) Magnified 15,000 times. aligned growth. This result showed that the CNTs were similar to those observed in previous The above results showwere thatgenerated impurities may hindergas the generation CNTs and flow the flow research. When the CNTs with an argon flow rate of 800of sccm and [18], acetylene The above results show that impurities may hinder the generation of CNTs [18], and the rateof of the 10 sccm, yield CNTs could be observed [23]. In addition, nanotubes an flow velocity argona high gas has anofabsolute impact upon the generation ofthe CNTs. Whenshowed the argon gas velocity of the argon gas has an absolute impact upon the generation of CNTs. When the argon gas generation because the lowerhad concentration of impuritiesof[9]. flow aligned rate was higher, phenomenon the generated CNT of structure a lower probability curling, along with

flow rategrowth. was higher, generated a lower probability curling,in along with aligned This the result showedCNT that structure the CNTshad were similar to those of observed previous aligned ThisCNTs resultwere showed that thewith CNTs to those previous research. research.growth. When the generated an were argonsimilar gas flow rate ofobserved 800 sccminand acetylene flow When the CNTs were generated with an argon gas flow rate of 800 sccm and acetylene flow rate an of rate of 10 sccm, a high yield of CNTs could be observed [23]. In addition, the nanotubes showed 10 sccm, a high yield of CNTs could be observed [23]. In addition, the nanotubes showed an aligned aligned generation phenomenon because of the lower concentration of impurities [9]. generation phenomenon because of the lower concentration of impurities [9].

Materials 2017, 10, 169 Materials 2017, 10, 169

Materials 2017, 10, 169

8 of 15 8 of 15

8 of 15

ofofIron Phthalocyanine Component 3.2.3.2. Analyses of Iron Phthalocyanine 3.2.Analyses Analyses Iron PhthalocyanineComponent Component TEM was used the formof ofthe theCNTs CNTsand andcatalyst, catalyst, shown 7.7.Many black TEM was used to to observe the and catalyst, shown inFigure Figure Many black TEM was used toobserve observe theform form of the CNTs asas shown in in Figure 7. Many black spots bebeobserved, which were uniformly distributed on the surface the CNTs, ashighlighted highlightedby spots cancan be which were uniformly distributed on the surface ofof the CNTs, spots canobserved, observed, which were uniformly distributed on the surface of the CNTs, asashighlighted bythe thecircles. circles. the by circles.

Figure 7. TEM image of iron phthalocyanine on the surface of carbon nanotubes.

Figure 7. 7. TEM image of carbon carbonnanotubes. nanotubes. Figure TEM imageofofiron ironphthalocyanine phthalocyanineon on the the surface of

An EDS analysis of the CNT component was conducted, and the results are shown in Figure 8.

EDS analysis the CNTcomponent component wasamong conducted, the are inin Figure 8. 8. An EDS analysis ofof the CNT conducted, and theisresults results areshown shown Figure TheAn CNTs contain four elements (C, O, Pt, andwas Fe), whichand carbon the majority element, with The CNTs contain four elements (C, O, Pt, and Fe), among which carbon is the majority element, with the highest peak. Theelements oxygen peak the Fe), oxygen present in the instrument and aelement, Pt The CNTs contain four (C, isO,due Pt,toand among which carbon is thecavity, majority the highest peak. The oxygen peak is due to the oxygen present in the instrument cavity, and a Pt coating was used to facilitate observation with the electron microscope. The trace iron element with the highest peak. The oxygen peak is due to the oxygen present in the instrument cavity, and a Pt coating was used to facilitate observation with the electron microscope. The trace iron element originated from the added iron phthalocyanine. coating was used to facilitate observation with the electron microscope. The trace iron element originated from the added iron phthalocyanine. originated from the added iron phthalocyanine.

Figure 8. EDS analysis results for the carbon nanotube-iron phthalocyanine composite.

Figure 8. EDS analysisresults resultsfor forthe thecarbon carbonnanotube-iron nanotube-iron phthalocyanine 3.3. Cyclic Voltammetry Analyses Figure 8. EDS analysis phthalocyaninecomposite. composite.

The electrochemical activity generated by the electrode in the cell system was characterized 3.3. Cyclic Voltammetry Analyses 3.3. Cyclic Analyses method. When the redox activity of the electrode was stronger, the using Voltammetry the cyclic voltammetry The electrochemical activity generated electrode in the cell systemanalyses was characterized measured peak was more pronounced. In by thisthe research, cyclic voltammetry of four The electrochemical activity generated by the electrode in the cell system was characterized using using the cyclic activity of theare electrode stronger, electrodes were voltammetry performed at amethod. scanningWhen rate ofthe 10 redox mV, and the results shown inwas Figure 9. The the the cyclic voltammetry method. When the redox activity of the electrode was stronger, the measured reduction peak observed on the SS316 electrode is at the position with a reference voltage of 0.035 measured peak was more pronounced. In this research, cyclic voltammetry analyses of Vfour peakand wascurrent more of pronounced. In this research, cyclic voltammetry analyses of four electrodes were mA. The CNT/SS316 wasshown less than that of the electrodes were −0.37 performed at areduction scanningpeak ratefor of the 10 mV, and theelectrode results are in Figure 9. The performed atpeak a scanning rate 10 mV, electrode andofthe shown inatFigure 9. The reduction stainless steel electrode, andof the position theresults peak was voltage ofvoltage −0.047 V 0.035 and peak reduction observed on the SS316 isreduction at theare position with aareference of V observed on the SS316 electrode is at the position with a reference voltage of 0.035 V and current current of −12 mA. The of the reduction of the FePc/RACNT/SS316 electrode (where and current of −0.37 mA.position The reduction peak for peak the CNT/SS316 electrode was less than that of the of − 0.37 mA. The reduction peak the CNT/SS316 electrode was less than that of the stainless RACNT represents random carbon nanotube) at reduction a voltage ofpeak −0.022 V and of of −8.6 mA. The stainless steel electrode, and the for position of is the was at acurrent voltage −0.047 V and reduction peak of the FePC/CNT/SS316 electrode with the added iron phthalocyanine is at a voltage steel electrode, and the position of the reduction peak was at a voltage of − 0.047 V and current current of −12 mA. The position of the reduction peak of the FePc/RACNT/SS316 electrode (whereof of 0.018 V position and current mA. The reduction peaks observed in the V four samples near theThe −12RACNT mA. The of of the−21 reduction peak ofisthe RACNT represents random carbon nanotube) at aFePc/RACNT/SS316 voltage of −0.022 andelectrode currentwere of(where −8.6 mA. reference potential of 0 V, and no oxidation peak reaction was observed because the electrodic represents random carbon nanotube) is at aelectrode voltage of −0.022 V andiron current of −8.6 mA.isThe reduction peak of the FePC/CNT/SS316 with the added phthalocyanine at areduction voltage peak the V FePC/CNT/SS316 the added phthalocyanine at a voltage 0.018 of of 0.018 and current of −21electrode mA. The with reduction peaksiron observed in the four is samples were of near the V potential of The 0 V, reduction and no oxidation peak reaction because thethe electrodic andreference current of −21 mA. peaks observed in the was four observed samples were near reference potential of 0 V, and no oxidation peak reaction was observed because the electrodic reaction was

Materials 2017, 10, 169 Materials 2017, 10, 169

9 of 15 9 of 15

reaction wasThe irreversible. addition of CNTsthe could improve theactivity electrochemical activity ofiron the irreversible. addition ofThe CNTs could improve electrochemical of the electrode, and electrode, and iron a goodincatalyst, as shown in et theal.studies byPacios Deng et al. al. [24], [18] phthalocyanine wasphthalocyanine a good catalyst,was as shown the studies by Deng [18] and and Pacios et al. [24], it improved the electrochemical Iniron thisphthalocyanine research, CNTswere and because it improved thebecause electrochemical activity. In this research, activity. CNTs and iron phthalocyanine were combined oncatalyst, the SS316 electrode the catalyst, could greatly combined on the SS316 electrode as the which could as greatly improvewhich the electrochemical improveofthe activity of the stainless steel electrode. activity theelectrochemical stainless steel electrode.

Figure Figure 9. 9. Measurement chart of cyclic voltammetry current-voltage. current-voltage.

In the the study study of In of the the effect effect of of the the generation generation direction direction of of the the CNTs CNTs on on the the electrochemical electrochemical activity, activity, the observations shown in Figure 9 indicate that the reduction peaks generated by the the observations shown in Figure 9 indicate that the reduction peaks generated by the SS316 SS316 electrode electrode for different different argon argongas gasflows flowswere were significant, measured electrochemical activity for notnot significant, andand the the measured electrochemical activity was was obviously worse than that for the CNT/SS316 electrode. Although the reduction peak could be obviously worse than that for the CNT/SS316 electrode. Although the reduction peak could be clearly clearly observed for the FePc/RACNT/SS316 the FePc was added, it wasthan not the better than the observed for the FePc/RACNT/SS316 after the after FePc was added, it was not better CNT/SS316 CNT/SS316 electrochemical activity generated with aligned nanotubes. In addition, it can be observed electrochemical activity generated with aligned nanotubes. In addition, it can be observed that the that the FePc/CNT/SS316 has a reduction andelectrochemical the best electrochemical activity. FePc/CNT/SS316 electrodeelectrode has a reduction peak and peak the best activity. Compared Compared with discussed the resultsin discussed in Section 3.1 forgas an argon gas flowofvelocity of 800 the with the results Section 3.1 for an argon flow velocity 800 sccm, the sccm, direction direction of the generated CNTs was easily discerned. The excessive impurities on the CNTs of the generated CNTs was easily discerned. The excessive impurities on the CNTs generated generated withargon different gas flow are attached to the which CNTs, which may affect their surface with different gas argon flow rates are rates attached to the CNTs, may affect their surface area area generation and generation direction [25,26]. Thus, they would result a lowredox redoxactivity activitymeasured measured for for the the and direction [25,26]. Thus, they would result inin a low randomly-oriented CNTs. randomly-oriented CNTs. Resistance 3.4. Analyses of Discharge with Constant Resistance the long-term long-term discharge discharge of of the the present present bio-electro-Fenton bio-electro-Fenton system system employing employing the In Figure 10 the FePc/CNT/SS316 composite FePc/CNT/SS316 compositecathode cathodeisisreported. reported.We We conducted conducted four four discharge discharge cycles, and each cycle atat thethe start-up process mentioned at the endend of the lasts about four four days. days.Each Eachdischarge dischargecycle cyclebegins begins start-up process mentioned at the of Section 2.4 and ends when reaching a maximum output voltage. Then, the chambers are cleaned up the Section 2.4 and ends when reaching a maximum output voltage. Then, the chambers are cleaned newnew anolyte andand catholyte solutions areare added, up anolyte catholyte solutions added,and andthe thenext nextdischarge dischargecycle cycle is is started. started. Figure 10 shows that each discharge cycle can reach a stable maximum voltage level of 0.8 V, which verifies the of the the present presentbio-electro-Fenton bio-electro-Fentonsystem systememploying employingthe thecomposite compositeFePc/CNT/SS316 FePc/CNT/SS316 reproducibility of cathode. Additionally, Additionally,the the maximum output voltage (0.8agrees V) agrees with all open MFCs’ openvoltages circuit cathode. maximum output voltage (0.8 V) with all MFCs’ circuit voltagesV) (0.7–0.8 V) mentioned byInRen [9]. In summary, the long-term discharge that the (0.7–0.8 mentioned by Ren [9]. summary, the long-term discharge proves thatproves the composite composite FePc/CNT/SS316 suitable for cathode being theofcathode of a bio-electro-Fenton FePc/CNT/SS316 is suitableisfor being the a bio-electro-Fenton system. system.

Materials 2017, 10, 169 Materials 2017, 10, 169

10 of 15 10 of 15

Figure Figure10. 10.Plot Plotofofdischarge dischargevoltage voltagevs. vs.time timefor forFePc/CNT/SS316. FePc/CNT/SS316.

3.5. 3.5. Analysis Analysis of of Electrical Electrical Quantities Quantities The The electrical electrical quantity quantity of of aa bio-eletron-Fenton bio-eletron-Fenton system system is is characterized characterized by by the the three three physical physical quantities: current density, and and power density. The anode electrode of the present quantities:open-circuit open-circuitvoltage, voltage, current density, power density. The anode electrode of the bio-electron-Fenton system is system carbon felt, while its is made of a SS316 stainless steel present bio-electron-Fenton is carbon felt,cathode while electrode its cathode electrode is made of a SS316 on whichsteel CNTon and FePcCNT are deposited carbon felt was commonly employed as stainless which and FePc sequentially. are depositedSince sequentially. Since carbon felt was commonly the electrode of microbial fuel cells in fuel many literatures, it is not within thewithin investigation scope of employed as the electrode of microbial cells in many literatures, it is not the investigation the present work. The present workpresent is the first thatisemploys composite of FePc/CNT/SS316 scope of the present work. The work the firstthethat employs the composite of as a cathode electrode. Hereelectrode. the influence FePc/CNT/SS316 cathode oncathode the electrical FePc/CNT/SS316 as a cathode Here of thethe influence of the FePc/CNT/SS316 on the properties of a bio-electro-Fenton system is discussed. Figure 11 compares electrical electrical properties of a bio-electro-Fenton system is discussed. Figure 11the compares theproperties electrical of a bio-electro-Fenton system employing different treatments cathode electrodes, SS316, properties of a bio-electro-Fenton system four employing four different of treatments of cathodebare electrodes, SS316 covered by aligned CNT, SS316 covered by FePc and random CNT, SS316 CNT, covered bySS316 FePc bare SS316, SS316 covered by aligned CNT, SS316 covered by FePc andand random and and aligned CNT.and Its abscissa is current density,isthe left-ordinate voltage, and theisright-ordinate is covered by FePc aligned CNT. Its abscissa current density,isthe left-ordinate voltage, and the power density. is The I–V curves (V The vs. CD) to (V thevs. left-ordinate, the power curves (PD CD) right-ordinate power density. I–V refer curves CD) refer towhile the left-ordinate, while thevs. power refer to (PD the right-ordinate. case of bare SS316, open-circuit voltage intersection point(the of curves vs. CD) refer toFor thethe right-ordinate. For theitscase of bare SS316, its (the open-circuit voltage 2 I–V curve and left-ordinate) is and 0.6 V, the maximum current is 3.42 mA/m (theisintersection intersection point of I–V curve left-ordinate) is 0.6 V, thedensity maximum current density 3.42 mA/m2 point of I–V curvepoint and abscissa), which too tiny to readisintoo thistiny figure, and in thethis maximum power (the intersection of I–V curve andisabscissa), which to read figure, and the 2 2 maximum power density is 0.29 For the case its of CNT/SS316, open-circuit voltage is 0.74 V, density is 0.29 mW/m . For the mW/m case of .CNT/SS316, open-circuitits voltage is 0.74 V, the maximum 2 . FormW/m 2, and the 2. For the maximum 2497the mA/m maximum power density is 588.96 current densitycurrent is 2497density mA/m2is, and maximum power density is 588.96 mW/m the case of 2, the case of FePc/RACNT/SS316, its open-circuit voltage 0.64 V, the maximum current FePc/RACNT/SS316, its open-circuit voltage is 0.64 V, theismaximum current density is 2365density mA/mis 2, and the 2365the mA/m maximum density is 2540.44 mW/m . For the case of FePc/CNT/SS316, its and maximum power densitypower is 540.44 mW/m . For the case2of FePc/CNT/SS316, its open-circuit 2 , and 2, and the maximum open-circuit is 0.7 V, the maximum current density is 3206 mA/m power voltage is 0.7voltage V, the maximum current density is 3206 mA/m the maximum power density is 2 . mW/m2. densitymW/m is 726.55 726.55 Bare SS316 SS316 shows shows the the worst worst electrical electrical quantity, quantity, both both of of I–V I–V and and power power density density curves, curves, which which is is Bare due totoitsits poor electron transport property [27]. CNT/SS316 showselectrical better electrical due poor electron transport property [27]. CNT/SS316 shows better propertiesproperties than that than thatwhich of SS316, which that CNT can promotetransport the electron transport in efficiently stainless steel of SS316, reveals that reveals CNT can promote the electron in stainless steel [18]. efficiently [18]. FePc/CNT/SS316 further better electrical than that ofwhich CNT/SS316, FePc/CNT/SS316 shows further shows better electrical properties thanproperties that of CNT/SS316, is due which due to FePc being its planar structure can to easily to CNT promote to FePcisbeing a catalyst and aitscatalyst planar and structure can easily adhere CNTadhere to promote thetoreduction the reduction reaction [17,28]. at the cathode It should be that mentioned here that theCNTs direction of the reaction at the cathode It should[17,28]. be mentioned here the direction of the are aligned CNTs are the aligned theduring flow rate argon during the chemical vaporofdeposition of by tuning flow by ratetuning of argon theof chemical vapor deposition process CNTs. To process verify the CNTs. To verify the influence the alignment of CNTs, another case of random CNT, influence of the alignment of CNTs, of another case of random CNT, FePc/RACNT/SS316, is compared FePc/RACNT/SS316, is compared here again. As shown in Figure FePc/RACNT/SS316 here again. As shown in Figure 11, FePc/RACNT/SS316 reveals worse11, electrical properties thanreveals those worse electrical properties than those of CNT/SS316 and FePc/CNT/SS316. This is because the

Materials 2017, 10, 169

11 of 15

Materials 2017, 10, 169 11 of 15 of CNT/SS316 and FePc/CNT/SS316. This is because the RACNT were intertwined, which leads to a lower specific surface area than that of aligned CNTs. Low specific surface area limits the utilization RACNT were intertwined, which leads to a lower specific surface area than that of aligned CNTs. of theLow catalyst FePc [29].area In limits summary, CNTs can effectively promote thesummary, conductivity SS316 specific surface the utilization of the catalyst FePc [29]. In CNTsof can stainless steel, the catalyst FePc can easily adhere to CNTs to promote the reduction reaction effectively promote the conductivity of SS316 stainless steel, the catalyst FePc can easily adhere toat the cathode, and alignment of CNTs is critical to cathode, the utilization FePc dueoftoCNTs its high specific surface CNTs to the promote the reduction reaction at the and theofalignment is critical to the of[16] FePcemployed due to itsahigh specificnano-Mo surface area. Zagal et al. [16] employed a composite area. utilization Zagal et al. composite as the anode electrode of 2 C/CNT/carbon-felt nano-Mo 2C/CNT/carbon-felt the anode electrode of a microbial fuel cell. The present reaches a microbial fuel cell. The presentas work reaches a maximum open-circuit voltage 0.74 Vwork which is slightly 2 a maximum open-circuit voltage 0.74 V which is slightly higher than that of it (0.702 V). higher than that of it (0.702 V). Our maximum current density reaches about 3.2 A/m , which isOur slightly density reaches about 3.2 A/m2, which is slightly lower than that of Zagal et al.’s 2 ). Our lowermaximum than thatcurrent of Zagal et al.’s electrode (11.2 A/m maximum power density reaches about electrode (11.2 A/m2). Our maximum power density reaches about 0.73 W/m2 which a little lower 2 ).isThe 0.73 W/m2 which is a little lower than that of Zagal et al.’s electrode (1.05 W/m performance than that of Zagal et al.’s electrode (1.05 W/m2). The performance of electricity generation in the of electricity generation in the present work slightly lower thatwork of [16]. the present present work is slightly lower than that of is [16]. However, thethan present is a However, bio-electro-Fenton worksystem is a bio-electro-Fenton system whose electricity generation is to support the Fenton reaction at whose electricity generation is to support the Fenton reaction at the cathode. The performance the cathode. The performance of the Fenton reaction is characterized by the dye declorization of the Fenton reaction is characterized by the dye declorization at the cathode, which will beat the cathode, whichinwill be discussed in the Section 3.6. discussed the Section 3.6.

Figure 11. Plot showing comparisonofofelectrical electrical properties properties of current, andand power on the Figure 11. Plot showing comparison ofvoltage, voltage, current, power onfifth the fifth day after initiation of the bio-electro-Fenton system. day after initiation of the bio-electro-Fenton system.

3.6. Decolorization Analyses

3.6. Decolorization Analyses

The measurement results for the decolorization of the RB5 dye within 12 h using the different

The measurement results forinthe decolorization the RB5 within 12achieved h using with the different cathode electrodes are shown Figure 12. In these of figures, the dye decolorization the cathode electrodes are shownis in Figure 12.isIn figures, theofdecolorization achieved with the FePc/CNT/SS316 electrode 84.6%, which thethese maximum degree decolorization among the four cathode electrodes. The electrodes can be listed in the order of decreasing activity as the FePc/CNT/SS316 electrode is 84.6%, which is the maximum degree ofdecolorization decolorization among CNT/SS316 The electrode, the FePc/RACNT/SS316 andof thedecreasing SS316 electrode, which four follows: cathodethe electrodes. electrodes can be listed inelectrode, the order decolorization hadas decolorization of 74.0%, 42.1%, and 28%, The color ofelectrode, RB5 dye has different activity follows: therates CNT/SS316 electrode, the respectively. FePc/RACNT/SS316 and the SS316 chromaphore and concentration values due to the functional group, and the electron-Fenton reaction electrode, which had decolorization rates of 74.0%, 42.1%, and 28%, respectively. The color of breaks the bond of the functional group with •OH, resulting in the reduction of the chroma of the RB5 dye has different chromaphore and concentration values due to the functional group, and the dye. In the bio-electro-Fenton system, the maximum number of hydroxyl radicals was generated electron-Fenton reaction breaks the bond the functional with ·OH, resulting in the reduction when FePc/CNT/SS316 was used as the of cathode electrode.group Thus, the most significant decolorization of theofchroma of the dye. In the bio-electro-Fenton system, the maximum number of hydroxyl radicals the RB5 dye was achieved with this electrode. In terms of the electrical properties of the bio-electrowas generated when fuel FePc/CNT/SS316 was used as the cathode Thus, the most significant Fenton microbial cell system, the current density generated byelectrode. the system had a close relationship decolorization of the RB5 dye with this In terms ofimproved, the electrical properties with the decolorization rate was of theachieved RB5 dye. When this electrode. electrical property was the relative of the bio-electro-Fenton microbial fuel cell system, the current density generated by the system had a close relationship with the decolorization rate of the RB5 dye. When this electrical property was

Materials 2017, 10, 169

12 of 15

Materials 2017, 10, 169 demand for hydroxyl radicals was high, and the decolorization of the 12 RB5 of 15 dye improved, the relative Materials 10, 169 12 of 15 was fast. The2017, decolorization effect with the FePc/CNT/SS316 electrode, which had the maximum demand for hydroxyl radicals was high, and the decolorization of the RB5 dye was fast. The current density, was the best. This result showed that the advantages and disadvantages of the decolorization effect with the FePc/CNT/SS316 which had of thethe maximum density, demand for hydroxyl radicals was high, andelectrode, the decolorization RB5 dyecurrent was fast. The electrical properties of result the bio-electro-Fenton microbialand fueldisadvantages cell system were related toproperties the degree of was the best. This showed that the advantages of the electrical decolorization effect with the FePc/CNT/SS316 electrode, which had the maximum current density, dye decolorization. of the bio-electro-Fenton microbial fuel cell system were related to the degree of dye decolorization.

was the best. This result showed that the advantages and disadvantages of the electrical properties of the bio-electro-Fenton microbial fuel cell system were related to the degree of dye decolorization.

12. Plot showing rates decolorizationfor for different different electrode (concentration of RB5: FigureFigure 12. Plot showing rates of of decolorization electrodesystems systems (concentration of RB5: 50 ppm). Figure 12. Plot showing rates of decolorization for different electrode systems (concentration of RB5: 50 ppm). 50 ppm).

The decolorization rates for the bio-electro-Fenton microbial fuel cell with the FePc/CNT/SS316 The decolorization rates for thethe bio-electro-Fenton fuel cellshown withthe the FePc/CNT/SS316 electrode and different RB5 concentrations of 20, 50, 70,microbial and 100 ppm in Figure 13. These The decolorization rates for bio-electro-Fenton microbial fuelare cell with FePc/CNT/SS316 results show that the decolorization degree had a boundary around the concentration of 50Figure ppm. 13. electrode and different RB5 concentrations of 20, 50, 70, and 100 ppm are shown in electrode and different RB5 concentrations of 20, 50, 70, and 100 ppm are shown in Figure 13. These the concentration was greater than 50 ppm, the decolorization degree within 12 h was of TheseWhen results show that the decolorization degree had a boundary around the concentration results show that the decolorization degree had a boundary around the concentration of 50 ppm. approximately 60%–70%. When the concentration was less than 50 ppm, the decolorization degree When the the concentration waswas greater thanthan 50 ppm, the the decolorization degree within 12 h 12 was 50 ppm. When concentration greater 50 ppm, decolorization degree within h was could reach more than 80%. Thus,the the RB5 concentration was thedecolorization required for the approximately 60%–70%. When concentration was less less than 50 ppm, degree approximately 60%–70%. When the when concentration was than 50higher, ppm,the thetime decolorization degree decolorization by the bio-electro-Fenton longer. could reach more than 80%. Thus, whensystem the RB5was concentration was higher, the time required for the

could reach more than 80%. Thus, when the RB5 concentration was higher, the time required for the decolorization by the bio-electro-Fenton system was longer. decolorization by the bio-electro-Fenton system was longer.

Figure 13. Chart of absorbance value vs. time for different concentrations of RB5. Figure 13. Chart of absorbance valuevs. vs.time time for for different of of RB5. Figure 13. Chart of absorbance value differentconcentrations concentrations RB5.

Materials 2017, 10, 169

13 of 15

4. Conclusions In this research, the electrical properties of bio-electro-Fenton microbial fuel cells with FePc/CNT/SS316 electrode were investigated, and the conclusions are as follows. 1.

2.

3.

4.

5.

CNTs were fabricated using TPCVD, and it was found that the argon gas flow velocity had an absolute impact upon the generation of CNTs. When the argon gas flow velocity was higher, the probability of curling in the generated CNT structure was lower, with aligned generation. The CNTs were generated with good alignment under an argon gas flow rate of 800 sccm and acetylene flow rate of 10 sccm. In the measurements using the cyclic voltammetry method, the measured position of the reduction peak of the SS316 electrode had values of 0.035 V and −0.37 mA; the position of the reduction peak of CNT/SS316 had values of −0.047 V and −12 mA; the position of the reduction peak of the FePc/RACNT/SS316 electrode had values of −0.022 V and −8.6 mA; and the position of the reduction peak of FePc/CNT/SS316 had values of 0.018 V and −21 mA. This showed that the iron phthalocyanine and serialized CNT catalysts were conducive to improving the redox activity of the electrode, while the FePc/CNT/SS316 electrode could improve the redox activity of the SS316 electrode. The current density and power density of the system with the FePc/CNT/SS316 electrode were 3206.30 mA/m2 and 726.55 mW/m2 , respectively; while the current density and power density generated by the SS316 electrode were 3.42 mA/m2 and 0.28 mW/m2 , respectively. These results showed that the current density and power density of the modified FePc/CNT/SS316 electrode were 937 and 2594 times higher than those of the system with only the SS316 electrode. In addition, in the experiment on the fixed resistance discharge of the bio-electro-Fenton microbial fuel cells, after exchanging the chamber solution over four cycles, the FePc/CNT/SS316 electrode could still reach a stable voltage output of 0.8 V. In the decolorization experiment, the electricity generated by the anode chamber of the bio-electro-Fenton microbial fuel cell system could independently promote the initiation of the electron-Fenton system reaction of the cathode chamber. The decolorization rates of the SS316 electrode, FePc/RACNT/SS316 electrode, CNT/SS316 electrode, and FePc/CNT/SS316 electrode within 12 h were 28%, 42.1%, 74%, and 84.6%, respectively, among which the decolorization efficiency of the modified FePc/CNT/SS316 electrode was the best. In this research on a bio-electro-Fenton microbial fuel cell system, the anode was employed for electricity generation and the cathode was employed for the Fenton reaction, achieving the effects of energy generation and sewage disposal at the same time. In addition, an iron phthalocyanine and serialized CNT composite was adopted as the catalyst, which could improve the electrical properties and decolorization effect, and has great potential for improving bio-electro-Fenton microbial fuel cells in the future.

Acknowledgments: The financial support provided by the National Science Council of the Republic of China (Taiwan) through the MOST 104-2221-E-197-025- projects is greatly appreciated. Author Contributions: Yi-Ta Wang conceived the study, designed the experiments, analyzed the data, and wrote the paper; Ruei Shiang Wang performed the experiments. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3.

Allen, R.M.; Bennetto, H.P. Microbial fuel-cells electricity production from carbohydrates. Appl. Biochem. Biotechnol. 1993, 39, 27–40. [CrossRef] Chaudhuri, S.; Lovely, D.R. Electricity generation by direct oxidation of glucose in mediatorless microbial fuel cells. Nat. Biotechnol. 2003, 21, 1229–1232. [CrossRef] [PubMed] Feng, C.H.; Li, F.B.; Mai, H.J.; Li, X.Z. Bio-electro-Fenton process driven by microbial fuel cell for wastewater treatment. Environ. Sci. Technol. 2010, 44, 1875–1880. [CrossRef] [PubMed]

Materials 2017, 10, 169

4.

5. 6. 7. 8. 9.

10.

11. 12. 13.

14. 15.

16. 17. 18. 19.

20.

21.

22. 23. 24.

25.

14 of 15

Fernández de Dios, M.A.; Iglesias, O.; Bocos, E.; Pazos, M.; Sanromán, M.A. Application of benthonic microbial fuel cells and electro-Fenton process to dye decolourisation. J. Ind. Eng. Chem. 2014, 20, 3754–3760. [CrossRef] Wang, X.Q.; Liu, C.P.; Yuan, Y.; Li, F.B. Arsenite oxidation and removal driven by a bio-electro-Fenton process under neutral pH conditions. J. Hazard. Mater. 2014, 275, 200–209. [CrossRef] [PubMed] Zhang, L.; Yin, X.; Li, S.F.Y. Bio-electrochemical degradation of paracetamol in a microbial fuel cell-Fenton system. Chem. Eng. J. 2015, 276, 185–192. [CrossRef] Logan, B.E.; Hamelers, B.; Rozendal, R.; Schröder, U.; Keller, J.; Freguia, S.; Rabaey, K. Microbial fuel cells: Methodology and technology. Environ. Sci. Technol. 2006, 40, 5181–5192. [CrossRef] Ren, H.; Lee, H.S.; Chae, J. Miniaturizing microbial fuel cells for potential portable power sources: Promises and challenges. Microfluid. Nanofluid. 2012, 13, 353–381. [CrossRef] Ren, H.; Pyo, S.; Lee, J.I.; Park, T.J.; Gittleson, F.S.; Leung, F.C.C.; Kim, J.; Taylor, A.D.; Lee, H.-S.; Chae, J. A high power density miniaturized microbial fuel cell having carbon nanotube anodes. J. Power Sources 2015, 273, 823–830. [CrossRef] Ren, H.; Tina, H.; Gardner, C.L.; Ren, T.L.; Chae, J. A miniaturized microbial fuel cell with three-dimensional graphene macroporous scaffold anode demonstrating a record power density of over 10000 W/m3 . Nanoscale 2016, 8, 3539–3547. [CrossRef] [PubMed] Watanabe, K. Recent developments in microbial fuel cell technologies for sustainable bioenergy. J. Biosci. Bioeng. 2008, 106, 528–536. [CrossRef] [PubMed] Moon, H.; Chang, I.S.; Kim, B.H. Continuous electricity production from artificial wastewater using a mediator-less microbial fuel cell. Bioresour. Technol. 2006, 97, 621–627. [CrossRef] [PubMed] Cheng, S.; Liu, H.; Logam, B.E. Power densities using different cathode catalysts (Pt and CoTMPP) and polymer binders (Nafion and PTFE) in single chamber microbial fuel cells. Environ. Sci. Technol. 2006, 40, 364–369. [CrossRef] [PubMed] Yu, E.H.; Cheng, S.; Scott, K.; Logan, B.E. Microbial fuel cell performance with non-Pt cathode catalysts. J. Power Sources 2007, 171, 275–281. Zhao, F.; Harnisch, F.; Schröder, U.; Scholz, F.; Bogdanoff, P.; Herrmann, I. Application of pyrolysed iron (II) phthalocyanine and CoTMPP based oxygen reduction catalysts as cathode materials in microbial fuel cells. Electrochem. Commun. 2005, 7, 1405–1410. [CrossRef] Zagal, J.H.; Bedioui, F.; Dodelet, J.-P. N4-Macrocyclic Metal Complexes; Springer: New York, NY, USA, 2006. Zhao, F.; Harnisch, F.; Schröder, U.; Scholz, F.; Bogdanoff, P.; Herrmann, I. Challenges and constraints of using oxygen cathodes in microbial fuel cells. Environ. Sci. Technol. 2006, 40, 5193–5199. [CrossRef] [PubMed] Deng, L.; Zhou, M.; Liu, C.; Liu, L.; Liu, C.; Dong, S. Development of high performance of Co/Fe/N/CNT nanocatalyst for oxygen reduction in microbial fuel cells. Talanta 2010, 81, 444–448. [CrossRef] [PubMed] Wang, Y.; Li, B.; Cui, D.; Xiang, X.; Li, W. Nano-molybdenum carbide/carbon nanotubes composite as bifunctional anode catalyst for high-performance Escherichia coli-based microbial fuel cell. Biosens. Bioelectron. 2014, 51, 349–355. [CrossRef] [PubMed] Ghasemi, M.; Daud, W.R.W.; Hassan, S.H.A.; Jafary, T.; Rahimnejad, M.; Ahmad, A.; Yazdi, M.H. Carbon nanotube/polypyrrole nanocomposite as a novel cathode catalyst and proper alternative for Pt in microbial fuel cell. Int. J. Hydrog. Energy 2016, 41, 4872–4878. [CrossRef] Shaijumon, M.; Sundara, R.; Rajalakshmi, N. Platinum/multiwalled carbon nanotubes-platinum/carbon composites as electrocatalysts for oxygen reduction reaction in proton exchange membrane fuel cell. Appl. Phys. Lett. 2006, 88, 1–3. [CrossRef] Lee, S.W.; Sigmund, W.M. Formation of anatase TiO2 nanoparticles on carbon nanotubes. Chem. Commun. 2003, 21, 780–781. [CrossRef] Hang, S.T.; Yeh, C.Y.; Yao, P.C. Synthesis of carbon nanotubes over sol-gel deposited substrates from acetylene by thermal chemical vapor deposition. J. Sci. Eng. Technol. 2007, 3, 47–54. Pacios, M.; Martín-Fernández, I.; Villa, R.; Godignon, P.; Valle, M.D.; Bartrolí, J.; Esplandiu, M.J. Carbon Nanotubes as Suitable Electrochemical Platforms for Metalloprotein Sensors and Genosensors; InTech: Rijeka, Croatia, 2011. Li, Y.-H.; Wang, S.; Luan, Z.; Ding, J.; Xu, C.; Wu, D. Adsorption of cadmium(II) from aqueous solution by surface oxidized carbon nanotubes. Carbon 2012, 41, 1057–1062. [CrossRef]

Materials 2017, 10, 169

26. 27. 28. 29.

15 of 15

Lu, C.; Chiu, H. Adsorption of zinc (II) from water with purified carbon nanotubes. Chem. Eng. Sci. 2006, 61, 1138–1145. [CrossRef] Pocaznoi, D.; Calmet, A.; Etcheverry, L.; Bergel, A. Stainless steel is a promising electrode material for anodes of microbial fuel cells. Energy Environ. Sci. 2012, 5, 9645–9652. [CrossRef] Liu, Y.; Yue, X.; Li, K.; Qiao, J.; Wilkinson, D.P.; Zhang, J. PEM fuel cell electrocatalysts based on transition metal macrocyclic compounds. Coord. Chem. Rev. 2016, 315, 153–177. [CrossRef] Gong, K.; Du, F.; Xia, Z.; Durstock, M.; Dai, L. Nitrogen-doped carbon nanotube arrays with high electrocatalytic activity for oxygen reduction. Science 2009, 323, 760–764. [CrossRef] [PubMed] © 2017 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).