Abdellah Benzaouak*,Nour-Eddine TouachV.M.Ortiz-Martínez*,M.J.Salar-García*,F.Hernández-FernándezA.P.de los RíosMohammed El MahiEl Mostapha Lot fi*

1Laboratory of Mechanics and Industrial Processes,Chemical Sciences Research Team,ENSET,Mohammed V University in Rabat,Morocco

2Department of Chemical and Environmental Engineering,Technical University of Cartagena,Campus Muralla Del Mar Murcia,Spain

Keywords:Ferroelectric materials Tantalate Photocathode Microbial fuel cell Bioenergy Wastewater treatment

A B S T R A C T Microbial fuel cells (MFCs) are bio-electrochemical systems that can directly convert the chemical energy contained in an effluent into bioelectricity by the action of microorganisms.The performance of these devices is heavily impacted by the choice of the material that forms the cathode.This work focuses on the assessment of ferroelectric and photocatalytic materials as a new class of non-precious catalysts for MFC cathode construction. A series of cathodes based on mixed oxide solid solution of LiTaO3with WO3formulated as Li1-xTa1-xWxO3(x=0,0.10,0.20and 0.25),were prepared and investigated in MFCs.The catalyst phases were synthesized,identified and characterized by DRX,PSD,MET and UV–Vis absorption spectroscopy.The cathodes were tested as photoelectrocatalysts in the presence and in the absence of visible light in devices fed with industrial wastewater.The results revealed that the catalytic activity of the cathodes strongly depends on the ratio of substitution of W6+in the LiTaO3matrix.The maximum power densities generated by the MFC working with this series of cathodes increased from 60.45 mW·m-3for x=0.00(LiTaO3)to 107.2 mW·m-3for x=0.10,showing that insertion of W6+in the tantalate matrix can improve the photocatalytic activity of this material.Moreover,MFCs operating under optimal conditions were capable of reducing the load of chemical oxygen demand by 79%(CODinitial=1030 mg·L-1).

1.Introduction

Microbial fuel cells(MFCs)are devices that can convert the chemical energy of biodegradable substrates into electricity via bacteria[1–5].Therefore,this technology offers the possibility of treating wastewater and producing electricity at the same time. Along with other key factors，the efficiency of these systems is directly affected by cathode performance and thus by the materials employed for its construction.Many materials have been tested as cathode catalysts for the oxygen reduction reaction(ORR)in MFC systems[6–12],some of which are sensitive to light,increasing their activity when irradiated(photocatalysts)[8,13,14].In general,when an absorbent material is irradiated with UV light and/or visible light,with an energy(hν)greater than or equal to the energy of the band gap,electron transitions can take place between the different energy levels of the molecules,thereby generating pairs of electrons and holes.The photo-induced electrons(e-BC)can recombine inside the material or diffuse to the surface of the photocatalyst to react with the terminal electron acceptor at the interface of the adsorbate,and the resulting photo-induced holes(h+BV)can serve as acceptor of the electrons transferred from the anode.Titanium oxide(TiO2)is considered a good photocatalyst due to its semiconducting properties and band gap(near 3 eV)as well as to its low cost.However,the short lifetime of the electrons and holes in this material decreases its photocatalytic efficiency.To overcome this problem,TiO2is usually doped with other materials in order to contribute to charge separation and reduce the recombination rate of electrons and holes[15].

Materials with an internal electric field,especially ferroelectric materials,are a new generation of photocatalysts[16–21],displaying unique properties thanks to their permanent and stable polarization below the Curie temperature[22,23].This permanent polarization provides ferroelectric materials with photocatalytic properties due to the creation of apparent charges on the surface of the catalyst,separating the positive and negative faces[16,19,20,24].This property can(i)improve the transport of electrons at the interface of the catalyst and separation of charges,(ii)increase the lifetime of the photo induced electron-holes up to about 9 μs[25]and(iii)separate redox reactions and avoid reactions of intermediate species in reverse[20,26].Thus,the reduction reaction takes place on the positively charged face while the oxidation occurs on the negatively charged face,avoiding the reactions in the reverse direction.

Mixed oxides based on tantalates have been used in photocatalysis in several processes, some of these oxides, such as LiTaO3,being of ferroelectric nature[27–30].This work focuses on the study of ferroelectric tantalate as a new class of materials for cathode construction in MFCs.The ferroelectric properties can improve the photocatalytic activity by decreasing recombination speed and separate charges.On the other hand,tantalates are recognized by their corner-sharing structure in which octahedral positions are responsible for the high photocatalytic activity displayed by these materials,particularly in water splitting reaction[27,30–32].However,LiTaO3is characterized by its wide band gap[30].One strategy to reduce its energy gap consists of inserting other metal cations in the LiTaO3matrix through the combination with other oxide compounds.WO3presents a conduction band position below the position of the couple redox H+/H2.Thus,the combination of LiTaO3with WO3could result advantageous in enhancing its stability and the structure of the band gap compared to the redox couples O2/H2O and H+/H2.In this context,the present work offers a study of the performance of cathodes based on LiTaO3modified with W6+in single-chamber MFCs,investigating how the catalytic activity of the cathodes is affected by the substitution ratio in terms of power output.

2.Materials and Methods

2.1.Synthesis

Oxide solid solutions with chemical formula Li1-xTa1-xWxO3were prepared at several compositions(x=0,0.10,0.20 and 0.25).The oxides were synthesized by using Li2CO3(SOLVACHIM,99%),Ta2O5(SCHARLAU,TAO210,Extra-pur)and WO3(CERAC/pur,99.9%)as chemical precursors.The chemicals were mixed in stoichiometric proportion to obtain the desired oxide composition and then heated in a muffle furnace at 600°C for 48 h to eject CO2.Thereafter,the samples were subjected toa heat treatment at 800°Cfor 24 h and then to fast cooling.The synthesized products consisted of typically fine powders.

2.2.XRD,granulometry,TEM and EDS analyses

X-ray diffraction(XRD)technique was used to identify the synthesized phases and to obtain information on their crystalline structures and the extent of the solid solution.XRD analysis was performed using a Bruker D8 Advance diffractometer with Cu Kαradiation at 25 °C,in the 2θ range from 10°to 80°with a pitch length of 0.02°and a scanning speed of 0.05°·s-1.DRX at high temperature was used to determine the Curie temperature.

The particle sizes and morphology of the samples were determined with transmission electron microscopy(TEM,Tecnai G2Series,FEI Company)and energy-dispersive X-ray spectroscopy(EDS,Hitachi S-3500N).Their particle size distribution(PSD)and surface area were measured by a laser granulometer(MASTERSIZER2000).This technique consists of measuring the intensity as a function of the laser diffusion angle θ.The angular positions of the extremes are used to determine the particle diameters by comparison with the diffraction patterns calculated by the Mie theory.Indeed,the intensity depends on the volume of the diffusing particles and provides the cumulative volume of particles for each particle size class.

2.3.UV and visible absorption spectroscopy

The absorption and transmittance data of the samples were recorded with a Perkin Elmer Instrument lambda 900 UV/Vis/NIR spectrophotometer in the solid state at room temperature.The estimation of the band gap was determined from the absorbance spectrum using the Beer and Willardson equation:

where Egis the band gap energy of the material,νis the frequency of the incident radiation,h is Planck's constant,A is a constant related to the material,the coefficient n refers to the nature of transition(n=?for direct gap)and α is the absorption coefficient in cm-1,calculated by using the following equation:

where I0and I are the incident and transmitted intensities,respectively and t represents the length of the optical path.

The variation of the electrical conductivity was monitored as a function of the temperature at 100 Hz, 1 and 10 KHz using an LCR bridge type HP4192A.

2.4.MFC configuration

The prepared oxides were assessed in single-chamber MFCs consisting of glass bottle reactors of 250 ml anode capacity and equipped with an external jacket for temperature control(25°C).The cathode consisted of a mixture of the respective synthesized phases as catalysts and a solution of polytetrafluoroethylene as binder(PTFE)(60 wt%dispersion in water,Sigma-Aldrich)mechanically pressed onto a 4 cm diameter piece of carbon cloth.The total loading was of 60 mg·cm-2in a 1:9 mass ratio of PTFE:oxide.The anode comprises 100 g of graphite granules of 2–6 mm diameter(Graphite Store,USA)and a graphite rod of 3.18 mm diameter(Graphite Store,USA),which is externally connected to the cathode with a resistance load of 1 kΩ.MFC tests were run in batch mode.he anode chamber was loaded with 125 ml of industrial wastewater characterized by 1030 mg·L-1of COD and a pH of 7.3.All reactors were set up with a proton exchange membrane(PEM)of 4 cm diameter placed between the cathode and the anode chamber as explained in[33].MFCs were evaluated in the presence and in the absence of a light source.As light source,a visible lamp(500 W)was placed at a distance of 20 cm from the cathode.

Polarization and power curves were obtained by varying the external resistance load in the range 11 MΩ–1 Ω.The current(I)and power(P)densities were calculated according to the equations I=V(cell voltage)∕R(external resistor)and P=V2∕R,and then normalized to anode capacity.Internal resistance was calculated at the point of maximum power from power curves[5].Wastewater treatment was assessed in terms of chemical oxygen demand(COD)removal.COD was measured following the APHA protocol[34,35]employing a spectrophotometer Spectroquant Nova 30(Merck,Germany).The pH was measured using a digital pH meter(Crison Instruments,Spain)calibrated before each measurement.

3.Results and Discussions

3.1.XRD,granulometry and TEM

The formulas of the solid phases studied were expressed considering the anionic lattice of O3in order to facilitate the comparison of data with other works.These formulas were deducted from the system Li2CO3–Ta2O5–(WO3)2,which is known for the high solubility of WO3in LiTaO3[36,37].The XRD patterns of the samples obtained at several compositions are displayed in Fig.1.When the substitution ratio x takes a value below 0.25,the formation of a homogenous solid solution of LiTaO3and WO3is observed.For x=0.25,impurity peaks associated with TaO2phases reveal the presence of the non-isotype form of LiTaO3.This value is the same as that found by Elouadi et al.[38]and Wiegel et al.[37],who also observed the formation of a homogenous solid solution below x=0.25.The results are also in line with those reported by Blasse[36]and Ravez et al.[39],who fixed the limits for a homogenous solid solution at x=0.30.The small differences observed are possibly due to the synthesis conditions.Other works in the literature such as that of Kawakami et al. [40] reported homogenous structure compositions of up to x=0.40andx=0.5,which remains considerably higher compared with the values obtained in this work.

Fig.1.XRD patterns of LiTaO3modified with W6+.

For the studied powders,all XRD patterns are characteristic of torsional crystal with rhombohedral symmetry R3c.On the other hand,it can be observed that when the value of x increases in the system Li1-xTa1-xWxO3,the peaks corresponding to the crystallographic planes(104)and(110)decrease,suggesting that the W6+cations are inserted into the octahedral site by substituting the Ta5+and/or Li+cations.

3.2.Specific surface area

Table 1 shows the evolution of the specific surface area of the LiTaO3depending on the content of tungsten.As can be observed,the surface area decreases as the content of tungsten in the matrix increases up to x=0.20.For x=0.25,the value of the specific surface area increases again probably due to the formation of new species as was found by the XRD analysis(Fig.1).These values are relatively low due to the preparation method by solid route,which usually yields very small surface areas.However,it produces a high degree of crystallinity that promotes electron mobility[41].

Table 1Specific surface area of Li1-xTa1-xWxO3oxides

3.3.Particle size distribution

Fig.2 shows the plotted curves of the particle size distribution by laser diffractometry for all the synthesized phases.In the case of LiTaO3(x=0),only one type of main particle size is observed.By contrast,the insertion of tungsten in the tantalate matrix with values of x of 0.10 and 0.20,creates new particle classes with larger sizes compared with LiTaO3.In such cases,two main particle classes are noticeable.When x=0.25,up to three particle size classes are measured.In addition,all samples exhibit a high degree of crystallinity that increases as the tungsten insertion rate increases as confirmed by the TEM in Fig.3.The order of particle size for all cases(0.5–6 μm)is close to the size of a ferroelectric domain [42]. This parameter can play an important role in the photocatalytic activity displayed by the phases due to the existence of a permanent polarization,affecting the mechanism of recombination of photo-induced electrons as well as the involved redox reactions.

Fig.2.Particle size distribution by diffractometer laser of Li1-xTa1-xWxO3.

3.4.Optical properties

The understanding of the photocatalytic activity of the synthesized materials requires the study of their band structures and optical properties.Fig.4A includes the absorption spectra in the UV and visible ranges of the phases Li1-xTa1-xWxO3as a function of the wavelengthλ(nm).The results show that the lithium tantalate(x=0)only absorbs in the UV range and presents a maximum absorption at 240 nm.When tungsten ion is inserted into the LiTaO3matrix,the peak corresponding to the maximum absorption moves to the visible range,appearing at the wavelengths of 270 and 290 nm for x=0.10 and x=0.20,respectively.However,for x=0.25,the material presents two maximum absorption peaks at 245 and 295 nm,probably due to the presence of the Li0.75Ta0.75W0.25O3and TaO2phases,respectively.The presence of impurities of TaO2detected by X-ray diffraction(Fig.1)in this phase affects the absorption with a lower value of absorbed light as plotted in Fig.4A.On the other hand,the absorption spectra formed in the phases containing W6+with substitution ratios of 0.1and 0.2 are wider than in the case of LiTaO3.This can be explained by the presence of W--O bonds,which are more polarizable in comparison with the Ta--O bonds[37].

To evaluate the bandgap energy of the samples from absorption spectra,the linear portion of the plot(αhν)nas a function of photon energy(hν)was used(Fig.4B).The energy gap Egof the samples was determined by searching the y intercept value of the straight line corresponding to α=0.The energy values of direct band gap of these solid were estimated from absorption spectra at 4.81,3.85 and 3.67 eV for x=0,x=0.10 and x=0.20,respectively.The material corresponding to x=0.25 exhibits two gaps,Eg1=3.50 and Eg2=3.84.

Electrical measurements were performed to determine the electrical behavior of these products and to calculate the electrical activation energy(Table 1),which corresponds to the energy required to move an ion from its interstitial position through the defects,depending on the nature of the defect.The conductivity follows the Arrhenius law:

where k is the Boltzmann constant and T is the absolute temperature.

Fig.3.a)TEM images;b)representative EDS spectrum of Li1-xTa1-xWxO3(x=0.2).

Fig.4.A)-Absorption spectra of Li1-xTa1-xWxO3as a function of the wavelength λ.B)-(αhν)2as a function of the photon energy(hν)of Li1-xTa1-xWxO3.

The values of electric activation energy are displayed in Table 1.Among the phases synthesized,the oxide obtained with x=0.10 displays the highest conductivity and thus the minimum value of electric activation energy.The results obtained for 400°C≤T≤700°C are comparable to those published by S.Kawakami et al.[40].

3.5.MFC performance

Single chamber MFCs were used to evaluate the photocatalytic performance of the samples as cathode materials via oxygen reduction reaction(ORR)in the presence and in the absence of light.Figs.5 and 6 present the results obtained with the polarization tests performed on the MFCs working with the photocathodes based on Li1-xTa1-xWxO3and prepared by pressure method,when the systems offered a stable voltage response after 72 h.

Fig.5B shows that the maximum power density output under light conditions is reached by the MFC including the cathode based on Li0.90Ta0.90W0.10O3(x=0.10),with a total power output of 107.2 mW·m-3.The MFCs working with the cathodes based on the rest of oxide catalysts with a substitution ratio of x=0.25 and x=0.20 display respective maximum power outputs of 94.5 and 85.0.However,in the absence of the tungsten cation,the power output drops to 60.4 mW·m-3.Moreover,the value of current density for x=0.00 at the peak of power output is far below the values of current density obtained for the rest of substitution ratios(around 0.45 mA·dm-3).Fig.5A shows voltage response versus current density.Polarization curves typically include three types of regions that correspond to three types of potential losses for all studied compounds: activation loss(initial region),ohmic drop(intermediate region)and mass transfer limitation( final region)[43,44].The activation losses in the curves obtained with x=0.00,x=0.20 and x=0.25 are clearly more pronounced compared with x=0.10,indicating higher activation losses.

Fig.5.a)Voltage measured in polarization tests for the cathodes based on Li1-xTa1-xWxO3under light conditions.b)Generated power density of the studied phases under light conditions.

Fig.6.a)Voltage measured in polarization tests for the cathodes based on Li1-xTa1-xWxO3in the absence of light.b)Power curves of MFCs with Li1-xTa1-xWxO3cathodes in the absence of light.

Fig.6B illustrates the evolutions of the power densities of the MFCs in the dark for all the cathodes.As in the case of light conditions,it is noted that the production of electricity depends strongly on the tungsten content inserted in the LiTaO3matrix and that the maximum power densities generated by the Li1-xTa1-xWxO3cathodes vary as a function of the value of x.The highest observed power density is 72 mW·dm-3for x=0.10.Fig.6A shows the polarization curves of the MFC cathodes Li1-xTa1-xWxO3performed in the absence of light.The open circuit voltages(OCV)of all the cathodes are equal two by two;OCV for x=0 and x=0.25 are around 188 mV and OCV values for x=0.10 and x=0.20 are around 500 mV.In this case,a more pronounced decrease is observed in the potential from 500 to 202 mV for a current density of 0.09 mA·dm-3and from 500 to 210 mV for a current density of 0.26 mA·dm-3for x=0.10 and x=0.20,respectively.Beyond these values,the decrease is moderate to a potential around 150 mV for a current density of 0.6 mA·dm-3,and again the potential registered for a tungsten content of x=10 is higher for the medium interval of intensities.As expected,in the absence of tungsten,the polarization test offers the faster decay rate.The internal resistances(Rin)of MFCs working with Li1-xTa1-xWxO3and under light conditions are shown in Table 2.As observed,the addition of tungsten significantly reduces the internal resistance by more than 50%.

The performance of the catalysts with different tungsten contents under light and dark conditions is compared in Fig.7 in terms of maximum values of power density obtained from power curves,as one of the most relevant parameters in MFCs.As seen,the performance of the catalysts for all values of tungsten content is higher under light irradiation.Moreover,for both conditions,the maximum power is achieved when x=0.10,while the absence of this element in the LiTaO3matrix offers the poorest performance in both cases.

Table 2Results for UV–Visible absorption spectroscopy and MFC performance for Li1-xTa1-xWxO3

Fig.7.Maximum values of power density in MFCs with Li1-xTa1-xWxO3cathodes in the presence and in the absence of light.

According to these results,the insertion of W6+in the LiTaO3matrix clearly improves the photocatalytic activity at a concentration of x=0.10,enhancing the cathodic electron transfer process.The presence of W6+at such concentration increases the conductivity of LiTaO3and reduce its band gap(Table 2).As commented before,the highest value of maximum power was measured for the phase with x=0.10,which also displayed the higher conductivity and higher absorption light.The decrease of the power generated for x=0.20 and x=0.25 compared to x=0.10 cathode can be related to the fact that the mechanism of substitution in the phases changes when the tungsten concentration increases.The ions of W6+could be inserted into the matrix of LiTaO3according to several possibilities,either in a normal Ta5+or Li+site or in a Li+site already occupied by Ta5+,called the anti-site.Furthermore,following the principle of charge compensation, there may be a creation of vacant Li+sites.Thus,a W6+in a Ta5+site can leave a vacancy of Li+and W6+in a Li+site will create five vacancies of Li+[37].Comparing x=0.10 and x=0.20,the electric activation energy measured for the second concentration is higher.In the specific case of x=0.25,it was observed that the light absorption was hindered by the formation of different phases,with a higher electric activation energy required.The higher catalytic activity of the samples including W6+in comparison with the sample solely based on LiTaO3can be related to other factors such as the polar nature of these materials,which is associated with their ferroelectric depolarization field that increases the amount of molecules adsorbed per unit of surface[24].It should be noted that the redox reactions on ferroelectrics catalysts occur spatially separated.Therefore,the photo-generated electrons reduce the oxygen adsorbed,on the positive face,to form the superoxide anion radical that reacts with the H+crossing the membrane separator to be available in the double layer[16].

Fig.8.Power density and polarization curves with x=0.10 and Pt based catalysts.

Fig.8 compares the performance of the photocatalyst that displays the higher activity with x=0.10(Li0.9Ta0.9W0.1O3)and that obtained using a cathode based on platinum(Pt/C),which is recognized as the most efficient electrocatalysts for the ORR in MFCs.The maximum power density generated by the Li0.9Ta0.9W0.1O3photocatalytic phase(107.2 mW·m-3)is around 50%of the power output achieved with Pt/C(Fig.8).The performance is clearly higher when compared with other material configurations recently reported in previous work,e.g.MnO2/CNTs(69.54 mW·m-3)and MnO2supported on bare acetylene black(47.42 mW·m-3)configurations[44].

Finally,the degradation of organic matter in the MFCs using these materials was conducted by measuring the removal rate in the wastewater during 120 h as shown in Fig.9 for light conditions.The percentage of COD removal ranged between 64%for x=0.00 and 79%for x=0.10.For values of x=0.20and x=0.25 the final COD removal remains almost at the same level,around 67%.In all cases,it is observable that the highest percentage of COD removal in relative terms is reached after 48 h.For example,in the case of x=0.10 the 87%of the final COD rate is achieved after 48 h.The maximum rate of the organic load removal of 79%demonstrates a good efficacy of wastewater treatment for the MFCs working with the tungsten-doped synthesized phase.

Fig.9.COD removal over time for MFC working with cathodes based on Li1-xTa1-xWxO3.

4.Conclusions

In this work,several cathodes based on the solid solution Li1-xTa1-xWxO3(x=0,0.10,0.20and0.25)have been investigated in MFC systems.The catalyst phases were fully characterized by DRX,PSD,MET and UV–Vis absorption spectroscopy and assessed in the presence and in the absence of visible light.According to the results obtained,the insertion of tungsten with a ratio of x=0.10 in the LiTaO3matrix displays the highest photocatalytic activity,also offering the higher conductivity(minimum value of electric activation energy)and the highest level of power performance in MFCs when compared with the cathode materials analyzed.The level of power generated by the MFCs increases from 60.45 mW·m-3for x=0.00(LiTaO3)to 107.2 mW·m-3for the optimal value of tungsten content under light conditions,showing that insertion of W6+in the tantalate matrix can double the level of power output generated.Furthermore,MFCs operating under optimal conditions were capable of reducing the load of chemical oxygen demand by 72%.

Acknowledgments

This work was done as part of the ERASMUS-MUNDUS Mare Nostrum.We are grateful to the organizations responsible for managing the program.This work was partially supported by the Spanish Ministry of Science and Innovation(MICINN)and by the FEDER(Fondo Europeo de Desarrollo Regional),ref.CICYT ENE2011-25188 and by the Seneca Foundation 18975/JLI/2013 grants.