Professional Documents
Culture Documents
Please cite this article as: V. Becerril-Estrada, I. Robles, C. Martínez-Sánchez, L.A. Godínez, Study
of TiO2/Ti4O7 photo-anodes inserted in an activated carbon packed bed cathode: Towards the
development of 3D-type photo-electro-Fenton reactors for water treatment, Electrochimica Acta (2020),
doi: https://doi.org/10.1016/j.electacta.2020.135972.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
3 water treatment
5
a
6 Centro de Investigación y Desarrollo Tecnológico en Electroquímica S. C., Parque Tecnológico
9 México.
10
13
14
15
16 Abstract
18 cells were modified on the external walls with films of TiO2, Ti4O7 or TiO2/Ti4O7 mixtures.
19 Film characterization was carried out using SEM and UV-vis spectroscopy. The results of
21 on the decolorization of a methyl orange (MO) model dye solution showed that while
22 anatase provides better photocatalytic properties and the partially reduced Ti4O7 larger
24 substrate that combines the advantages of both materials (for PEC experiments for instance,
1
1 decolorization values for the model dye solution using TiO2, Ti4O7 and a TiO2/Ti4O7 mixed
3 effective photoanode material in a 3-D type reactor for water treatment processes, a
4 TiO2/Ti4O7 modified PMMA spectroscopy cell was inserted in an activated carbon (AC)
5 bed so that the semiconductor material could be illuminated using an external UV source
6 positioned inside the PMMA cell. The connected AC particles that were previously
7 saturated with MO dye were used as cathode sites for the oxygen reduction reaction so that
8 the photoelectrochemical reactions that take place in the anode could be complemented
10 resulted in an effective decolorization of the dye solution that results from a complex
12 pseudo-first order kinetic model so that a deeper understanding of the contribution of each
14
15 1. Introduction.
16 Advanced oxidation processes (AOP) are with no doubt some of the most promising
17 approaches for the treatment of wastewater containing recalcitrant pollutants due to their
18 speed and effectiveness. These processes are characterized by the generation and use of the
•
19 OH radical species which is a short-lived, non-selecting and powerful oxidant species that
20 reacts with almost any organic contaminant. There are several AOPs; i.e., numerous ways
21 in which the •OH radical can be produced and used [1,2]. Among these, electrochemical-
22 and photo-assisted AOPs are particularly interesting since electric and radiation stimuli can
2
1 photoelectrochemical (PEC), electro-Fenton (EF) and photo-electro-Fenton (PEF)
2 processes are all •OH radical generation techniques [4–7] in which electricity, light or an
4 are employed to produce a highly oxidant environment on the surface of either the anode,
8 material is titanium oxide (TiO2) [8–10]. This compound has several advantages such as
9 high photocatalytic activity [11,12], reasonable price, chemical stability and low toxicity
10 [13]. Commercially available TiO2, is usually composed of a mixture of anatase and rutile
11 (80-20%) [14] and it is well known that the former is characterized by a higher
13
14 On the other hand, Magneli phases are a sub-stoichiometric titanium oxides that are
15 described by the generic formula TinO2n-1, where n is an integer number between 4 and 10.
16 Among the different partially reduced degrees that can be found for magneli, Ti4O7 is
17 not only common and commercially available but also characterized by a high electric
18 conductivity (about 1500 S cm-1) that is similar to that of graphite [16]. There are some
19 other interesting properties of Ti4O7 such as its high chemical stability, hardness and low
20 toxicity [17] that have resulted in its use in batteries, fuel cell stands and photo-catalytic
21 substrates [18]. In addition, Ti4O7 has been reported to show a high potential for oxygen
22 evolution (2.5 V vs RHE) [16] and therefore, the potential use of magneli as an anode
23 material for EO processes, has called the attention of several research groups around the
24 world [19].
3
1
2 Depending on the nature of the excitation signal (light, electric polarization, or both), it
4 conducting [22–29] substrates. In the specific case of photo-anodes, optic fibers have been
5 chemically treated and surface modified with TiO2 so that the electrode can be illuminated
6 from within the optic fiber [30,31]. In spite of the promising results of these studies, the
7 SiO2 optic fibers that have been employed are fragile and therefore, a good alternative is
8 using plastic cuvettes of a transparent, low cost and easy handling material such as
10 photoanodes is that in principle, the modified cuvettes illuminated from the interior can be
11 inserted in 3D cathodes [35,36]. A good example of this type of electrodes are activated
13 adsorbent particles on which electrochemical reactions take place [37]. The use of 3D
16
17 In this context, we are hereby presenting the results of a study divided in two parts. In
18 the first one, commercially available PMMA spectroscopy cells were modified with films
19 of either TiO2, Ti4O7, or a mixture of both, and the photo- and photo-electrocatalytic
20 activity of the resulting electrodes was measured and compared. In the second part, the best
2 2. Experimental section.
3 2.1. Chemicals.
4 All chemicals employed in this work, were of analytical grade. methyl orange (MO),
5 KNO3, FeSO4•7H2O and acetone were purchased from Hycel, Fermont, JT Baker and
6 Sigma, respectively. While, TiO2 was obtained from Sigma Aldrich, Ti4O7 was purchased
7 from Magneli Materials-LLC. All solutions were prepared with ultra-pure deionized water
8 (18 MΩ cm) and when required, a HNO3 concentrated solution obtained from Merck was
9 employed to acidify the solution. pH measurements were carried out using a Thermo
10 Scientific Orion Star, A211, apparatus. Lignitic Activated carbon (AC) was obtained from
11 Clarimex (mesh size between 1.0 and 1.4 mm) and used as received.
12
13 2.2. PMMA cell modification with films of TiO2, Ti4O7 and TiO2/Ti4O7.
14 Commercial 4.5 mL PMMA cuvettes used for UV-vis spectroscopy analysis having flat
15 polymeric 1x4.5 cm2 sides and a 1x1 cm2 bottom, were modified with the semiconductor
16 material using a silicon-based commercial glue that was previously diluted with acetone
17 (3.5g of glue + 15 mL of acetone). In this way, 0.39 g of either TiO2, Ti4O7 or a 1:1 mixture
18 were evenly spread and fixed with the diluted glue on the 4 sides of the PMMA cuvette.
19 The modified cells were dried in an oven at 303 K for 24 h and while the mass of
20 semiconductor material was in all cases 0.39 g, the corresponding thickness was 200 µm.
21
23 Films anchored on the PMMA modified cells, were characterized using Scanning
24 Electron Microscopy (SEM) and UV-vis Spectroscopy. While SEM measurements were
5
1 carried out using a (JOEL JSM-5400L) microscope equipped with an X ray energy
2 dispersion accessory, the UV-vis spectroscopy experiments were performed with an UV-
3 Vis-NIR Agilent Cary 5000, apparatus. Using this spectrophotometer, band gap
8 As can be seen in Figure 1-A, the experimental cell to carry out PC, EO and PEC
9 experiments, consisted on a 100 mL glass flask partially filled with a MO dye containing
10 solution in which the semiconductor modified PMMA cuvette was immersed. For
11 polarization experiments, the semiconductor film supported on the polymeric substrate was
12 connected to a power source (Novak Technologies DCE 10/24-9) which in turn, was
13 coupled to a graphite rod GSP-250 (ø = 1.2 cm). The electrolytic medium consisted on 50
14 mL of a MO dye solution, 2.5x10-5 M and KNO3 0.1 M, which was maintained under
16 3 V between these two electrodes were applied for 60 min and samples for dye analysis
18
20 positioned inside the modified PMMA cuvette (see Figure 1-A). Electromagnetic
21 irradiation was continuously applied for 60 minutes and every 10 minutes a sample was
22 taken out from the solution in order to determine the concentration of the dye by means of a
24
6
1 For PEC experiments, the setup consisted on a system in which polarization and UV
6 works in re-circulation mode (110 mL min-1) and in which the semiconductor modified
7 PMMA cuvette (anode) is inserted in an AC packed bed (0.5 g cm-3, cathode) previously
8 saturated with the MO dye. Saturation conditions were achieved by exposing the AC
10 saturation adsorption capacity of AC with MO dye (29.3 mg g-1) was obtained from the
11 difference of absorbance measurements of the dye solution before and after saturation took
13 from PEC experiments in the cell shown in Figure 1-A), and EF/PEC’ experiments were
16
18 (5x10-5 M) also containing KNO3 0.1 M at pH 3. For EF/PEC’ experiments, the dye
19 contaminated solution contained 5.5 mg L-1 of FeSO4 and was maintained under oxygen
20 saturation conditions so that the Fenton mixture could be produced upon H2O2 electro-
22
7
1 As previously mentioned, the change in MO absorbance (λ = 505 nm) was measured in
2 sample solutions taken out from the reactor at 10 min time intervals using a Genesys 10S
5 3. Results
7 Figure 2 shows the morphology of the TiO2, Ti4O7 and the 1:1 TiO2/Ti4O7 films under
8 study. For the TiO2 film, a relatively dense structure that loses its consistency at the
9 outermost layers, can be observed. As has been previously discussed in the literature [40],
10 the cracks of ~20 to 25 µm in the surface structure have been suggested to be related to a
11 decrease in the TiO2 interparticle forces as the film becomes thick. This morphology is
12 quite different to that obtained for Ti4O7 (see Figure 2). Inspection of the relevant image
13 shows a disperse and porous structure with particle aggregates of about 50 µm; a feature
15 factor.
16
17 The TiO2/Ti4O7 film on the other hand, shows mixed morphology features. In this way,
18 the image shown in Figure 2 reveals not only a surface structure characterized by a
19 homogenous film with smaller surface cracks (about 10 µm), but also the absence of
22
23 Figure 2 also shows the EDS spectra of the three films under study. As expected, each
24 spectrum shows signals for titanium, silicon and oxygen. It is also interesting to note that in
8
1 addition to the peaks positioned at energies < 3 KeV, the presence of Ti(<IV) in Ti4O7 and
2 in TiO2/Ti4O7 can be readily identified by the relative intensity observed in the signal at 4.5
3 KeV.
5 Film characterization was also carried out in terms of the light absorption properties of
6 the surface confined materials. As can be seen in Figure 3, the data corresponding to diffuse
8 model [39], shows that the band gap (Eg) for TiO2 is 3.24 eV; a value that is close to that
9 reported in the literature (3.2 eV) [41]. In this context, a maximum absorption band could
10 not be identified for Ti4O7. This observation is consistent with reports that point out that
12 photocatalytic material when compared to TiO2 [42]. The TiO2/Ti4O7 mixed film on the
13 other hand shows intermediate properties, i.e., the response signal is located between the
15
17 The photocatalytic activity of the three films under study was tested as described in the
18 experimental section. In this way, the kinetics of dye decolorization was measured
21 performed in order to assess the adsorption capacity of the electrodes and as can be seen in
22 the dotted lines of Figure 4, the observed decolorization was in all cases negligible (< 2 %).
24 Inspection of the corresponding data shows that while the TiO2 and Ti4O7 films give the
9
1 best and worst photocatalytic degradation performances (19 and 3% after 60 min), the
5 These results support previous observations that point out that the photocatalytic
6 degradation of dye molecules using semiconductor materials, depends not only on the
8 these species to react with the pollutant molecules and with nearby oppositely charged
10 equations 1 to 3 [43,44] the photogenerated electrons and holes (see equation 1) can react
11 with either oxygen or water molecules to produce O2•- radical anions in the former and
12 surface adsorbed •OH radicals in the latter case (see equations 2 and 3, respectively).
13
17
19 Consistent with the previous set of experiments, the TiO2, Ti4O7 and TiO2/Ti4O7
20 mixed films were tested by following the kinetics of dye decolorization of a MO solution
21 under polarization conditions as described in the experimental section. In all cases, the
22 potential was applied in such a way that the semiconductor film was the anode and a
23 graphite rod worked as the cathode. Since surface adsorbed •OH radicals can be
10
1 electrochemically produced (see equation 4) [45], the experiments were performed
6 solution using the different semiconductor materials and the different polarization potential
7 values are presented. From the data in this Figure two important observations stand out. On
8 one hand, it is possible to see that within the experimental window the applied voltage is
9 proportional to the rate of decolorization (see Figures 5-A to -C) and therefore, to the
10 maximum amount of color disappearance that can be measured at a given time (see for
11 example decolorization percentages after 60 min for TiO2 in Figure 5-A, 19 %, 24 % and
13 extent of decolorization using the three different films under study suggests that as opposed
14 to what was observed for PC experiments, the best performance is given by Ti4O7, the
16 response that can be positioned between those associated to the other two materials (see
17 Figure 5-D).
18
19 The next step, consisted in testing the performance of the three semiconductor films
20 under simultaneous electric and light stimulation. In this way, PEC experiments were
21 carried out as described in the experimental section and as expected, the obtained responses
23
11
1 As can be seen from the relevant data shown in Figure 6, two main differences can be
2 readily identified. First, the decolorization percentages after 60 min are larger for PEC than
3 for the other two approaches (for example, 53% for PEC using TiO2/Ti4O7 vs 43 and 19%
4 for EO and PC using Ti4O7 and TiO2, respectively), and comparable to the data of PEC
5 experiments published by other authors [46–48]. The second difference is that, as opposed
6 to what was observed for PC and EO, the best performing semiconductor material is the
7 TiO2/Ti4O7 composite film. As can be seen from the experimental results in Figure 6 and
8 from equations 1-4, the combination of UV irradiation and electric polarization favors the
9 performance of the semiconductor film that not only can efficiently absorb photons and
10 generate (h+)/(e-) pairs, but that can also be effectively polarized so that charge carrier
12 [49,50]. In other words, the effective photo-anodic performance of the TiO2/Ti4O7 film
14 photocatalytic and electrochemical activities that were observed for TiO2 in the PC and for
15 Ti4O7 in EO experiments.
16
18 Considering that the mixed TiO2/Ti4O7 composite was the most effective film using a
19 PEC approach, the 3D-type reactor system shown in Figure 1-B was set up using the
21 and a UV lamp and a power source to provide electromagnetic radiation and electric
23 decolorization effects due to AC adsorption, the cathode material was previously saturated
24 with MO by exposing the material to a dye saturated solution for 72 h. Therefore, in the
12
1 experiments corresponding to the system in Figure 1-B, the initial concentration of the MO
2 dye was 5x10-5 M, the voltage applied by the power source was 3 V and the UV lamp
3 inserted in the PMMA cuvette provided continuous radiation at λ=365 nm and the reactor
5 It is also important to note that since the integral processes to be explored considered
6 not only the anodic dye-degradation reactions but also the cathodic effect in the 3D reactor,
7 the addition of 5.5 mg L-1 of FeSO4 to the electrolytic solution was carried out. As can be
8 seen by the reactions described in equation 5, the electrochemical generation of H2O2 (by
9 means of oxygen reduction at the polarized AC particle surface) in the presence of Fe(II)
10 cations gives rise to the well-known Fenton mixture which in turn, readily reacts as
11 described by equation 6 to produce the powerful •OH radical [51]. It is also important to
12 note that as equation 7 points out, the cathodic environment at the carbon-solution
13 interphase during the production of •OH radicals is also responsible for the regeneration of
14 Fe(II) from Fe(III); thus effectively sustaining the oxidant electro-generation conditions in
16
18 Fe 2+ + H 2 O2 → Fe 3+ + • OH + OH − (6)
19 Fe3+ + e − → Fe 2+ (7)
22 using this reactor, will only affect the performance of the anode. Therefore, in order to
23 avoid confusion with the experiments described in the previous section (see for instance the
13
1 discussion of PEC in figure 6), the PEC process in this part of the work will be labeled as
2 PEC’. In this way, while dye decolorization in the polarized and illuminated semiconductor
3 film immersed in the AC packed bed cathode will be marked as PEC’, the EF/PEC’ process
4 will stand for the combination of the photoelectrochemical contribution in the anode (PEC’)
5 and an electro-Fenton process (EF) taking place in the vicinity of the cathode surface (see
8 In this context, Figure 7 shows the color removal percentage (%CR) of a MO solution
9 obtained from experiments carried out in the absence (PEC’) and in the presence (EF/PEC’)
10 of Fe(II) species using a TiO2/Ti4O7 /PMMA anode and an AC packed cathode previously
11 saturated with the MO dye (see Figure 1-B). Both curves reveal a fast increasing CR within
12 the first 10 min. Afterwards, the rate of CR change is substantially smaller, resembling a
13 quasi-stationary type behavior that reaches CR percentages of 53 for PEC’ and 71% for
14 EF/PEC’. These two limiting values show that at long times, the decolorization processes
15 (PEC’, adsorption and EF when Fe(II) ions are present) are compensated with dye
16 desorption events from the previously saturated AC. Comparison of these two values on the
17 other hand, suggests that the EF contribution to the decolorization processes accounts for
18 about 30% of the total color removal observed at 60 min for EF/PEC’.
19
20 In the reactor shown in Figure 1-B, there is a semiconductor based anode and an AC
21 packed bed that not only works as a cathode for oxygen reduction but also as an adsorbent
22 material from which dye molecules are simultaneously being taken from and released to the
3 surface (PEC’, which would account for electrooxidation and photocatalytic contributions),
4 (2) Simultaneous adsorption and desorption processes of dye molecules at the surface and
5 pores of the particles of the AC packed bed (AD) and (3) Fenton induced decolorization of
8 Under this simplified scheme, equation 8 shows that the rate of mass change
9 (dMsol/dt) of the dye in solution (which is what can be experimentally observed trough
11 adsorption and desorption taking place on the surface of AC (dMAD/dt) and to PEC’
13
15 Assuming pseudo-first order kinetics for all the processes described in the right hand
16 side of equation 8 and considering that (dMAD/dt) describes two processes (related to
18
ௗெಲವ
19 = ܭௗ௦ ܯ௦௨ − ܭௗ௦ ܯ௦ (9)
ௗ௧
ௗெುಶ’
20 = −ܭா’ ܯ௦ (10)
ௗ௧
ௗெಶಷ
21 ௗ௧
= −ܭாி ܯ௦௨ (11)
22 This set of relationships shows that dMsol/dt depends on four relevant kinetic
23 constants (KPEC’, KEF, Kads and Kdes) and on the mass of the dye in solution (Msol) and on the
15
1 surface of the AC substrate (Msur). The signs of the kinetic constants on the other hand,
2 show that while there are three negative constants corresponding to dye consuming
3 processes (-KPEC’, -KEF, -Kads), there is also a positive kinetic constant (Kdes) that describes
6 The fraction of dye in solution on the other hand, corresponds to Msol / Msol,0 and
7 since the mass of dye on the adsorbent surface under saturation conditions at the beginning
9 approximately equal to Msur,0. Under these assumptions equation 8 can be written as,
10
13
ெೞೠೝ,బ
16 ݇ଷ = (15)
ெೞ,బ
17 The solution of equation 12 results in 16 (which strictly applies for EF/PEC’ but
18 can be simplified for PEC’ by removing the EF contribution, i.e., when k2=Kdes) in which
19 the fraction of mass of MO dye (which can be approximated to the decolorization fraction
21
ெೞ మ య మ య
22 = (1 + )݁భ ௧ − (16)
ெೞ,బ భ భ
16
1 A first inspection of equation 16 shows apparent consistency with the processes
2 described in Figure 8. In this way, while Msol =Msol,0 when t=0, (Msol / Msol,0) at t→∞
3 becomes -(k2k3/k1) which should be a value between 0 and 1. In this context, (k2k3/k1) can
4 be anticipated to be negative since k1 is negative (the sign of KPEC’ and Kads are both
5 negative) and k2 and k3 should be positive since k3=(Msur,0 / Msol,0)>0 and Kdes should be
6 larger than KEF (otherwise a quasi-stationary experimental response at long times would not
7 be observed).
8 In order to test these ideas and to find out how well the processes schematized in
9 Figure 8 describe the experimental data, the fraction of color removal values of MO (CR,
10 which corresponds to 1- (Msol / Msol,0)) in Figure 7 were fitted to the simplified kinetic
11 model described by equation 16. The resulting parameters k1 and k2 were computed in this
12 way (k3 was already known since it corresponds to the ratio of Msur,0 and Msol,0 values) and
13 used along with the previously determined experimental value of KPEC’ to calculate the
14 remaining kinetic constants Kads, Kdes and KEF. The complete set of values thus obtained is
15 shown in Table 1 where it can be seen that for the EF/PEC’ process, k2 is in fact positive
16 (Kdes-KEF = 0.023 min-1) and since k1 is negative (k1= -1.870 min-1), the ratio (Msol∞/ Msol,0)
17 = -(k2k3/k1)=0.296; a value that as expected, not only agrees well with the experimentally
18 observed %CR in Figure 7, but that also predicts the quasi-stationary behavior of the
19 system at long times (CR=0.70). For the PEC’ process on the other hand, the experiment
20 was carried out in the absence of Fe(II) species and therefore KEF=0. Under these
21 conditions, consistency of the fitted parameters is retained since k2 is still positive (in fact
22 larger than for the EF/PEC’ case) and k1 negative. The decolorization limit for PEC’ at long
23 times (Msol∞/ Msol,0)=-(k2k3/k1), corresponds in this case to 0.45 and the CR to 0.549; a value
17
1 that is in agreement with a process carried out in the absence of Fe(II) ions; thus reflecting
3 Simulated curves for both experiments using equation 16 and the constants in
4 reported in Table 1, are shown in Figure 7 as continuous lines. Inspection of the shape of
5 these curves in the short time region, reveals that both simulations predict a rapid increase
6 in the %CR that becomes a quasi-stationary response at rather short times (t<2 min). In the
7 2<t<60 min time window on the other hand, the data in Figure 7 not only suggests different
8 limiting CR values (0.549 and 0.7 for PEC’ and EF/PEC’, respectively) defined by a
9 process in which the rate of color change for the two experiments under study is balanced
10 by dye-consuming (adsorption, PEC’ and, when Fe(II) is present, EF) and dye-“generating”
11 (desorption from the previously saturated AC material) processes that are simultaneously
12 taking place in the reactor. By grouping the constants associated to these two types of
13 opposing effects in dye-concentration increasing and decreasing rate constants (Kincrease and
14 Kdecrease), it turns out that while Kincrease = 0.035 min-1 for both processes (as it should be
15 since dye desorption occurs in EF/PEC’ and PEC’ processes in exactly the same way),
16 Kdecrease= 1.858 and 1.882 min-1 for PEC’ and EF/PEC’, respectively. The difference
17 between the Kdecrease values for the two processes under study is obviously related to KEF
19 interesting to note that the small value of this difference is readily translated into the larger
20 effect for the limiting %CR levels observed in Figure 7, due to the different mass values of
21 the dye that must be used to compute the corresponding decolorization rates (in this case,
23
18
1 Conclusions
2 In this work, films of TiO2, Ti4O7 and a 1:1 TiO2/Ti4O7 mixture, were prepared on the
3 external surface of PMMA spectroscopy cuvettes and later, characterized in terms of their
5 dye aqueous solution. The results of photo-, electro- and photoelectrochemical experiments,
6 revealed that while TiO2 and Ti4O7 perform better in terms of photocatalytic and
8 the two materials, thus becoming the best semiconductor film for the photoelectrochemical
10
12 PMMA cuvettes allowed the illumination of the semiconductor film from within the
13 cuvette, making the immersion of this electrode in an AC packed bed possible. This
14 approach, which will be extended for the use of semiconductor modified optic fibers [31]
15 immersed in carbon particle based cathodes, was roughly tested in this work with a 3D-type
18 process.
19
20 The complex mixture of events that take place within the reactor gave rise to efficient
21 color removal performances that reached values close to a 70% in 60 min. Using a
22 simplified model, the performance of the reactor could be successfully fitted to the
23 associated rate constants of the corresponding processes, opening up the possibility for the
2 optimized.
3 Acknowledgments
4 The authors express their gratitude to the Mexican Council for Science and
5 Technology, CONACYT (2015-01-1574), and to the Bill and Melinda Gates Foundation
6 (OPP1156657), for their financial support of this work. V. B.-E. also thanks CONACYT
9 References
10 [1] J.A. Bañuelos, F.J. Rodríguez, J. Manríquez Rocha, E. Bustos, A. Rodríguez, J.C.
11 Cruz, L.G. Arriaga, L.A. Godínez, Novel electro-fenton approach for regeneration of
13 https://doi.org/10.1021/es401320e.
16 using a low cost carbon-felt air-diffusion cathode, J. Environ. Chem. Eng. 4 (2016)
17 2066–2075. https://doi.org/10.1016/j.jece.2016.03.012.
18 [3] K. Esquivel, M.G. García J., F.J. Rodríguez, M.V. González, L. Escobar-Alarcón, L.
20 Ti1−x O2, M: Ni, Co): synthesis and characterization for its potential application as
22 https://doi.org/10.1007/s11051-011-0245-y.
24 advanced oxidation processes: Today and tomorrow. A review, Environ. Sci. Pollut.
20
1 Res. 21 (2014) 8336–8367. https://doi.org/10.1007/s11356-014-2783-1.
5 [6] N. Oturan, C.T. Aravindakumar, H. Olvera-Vargas, M.M. Sunil Paul, M.A. Oturan,
9 [7] Z. Ye, E. Brillas, F. Centellas, P.L. Cabot, I. Sirés, Electro-Fenton process at mild
10 pH using Fe(III)-EDDS as soluble catalyst and carbon felt as cathode, Appl. Catal. B
14 nanowires versus TiO2 nanoparticles, J. Colloid Interface Sci. 498 (2017) 423–432.
15 https://doi.org/10.1016/j.jcis.2017.03.078.
16 [9] R. Miao, Z. Luo, W. Zhong, S.Y. Chen, T. Jiang, B. Dutta, Y. Nasr, Y. Zhang, S.L.
19 https://doi.org/10.1016/j.apcatb.2016.01.070.
20 [10] J. Zhang, X. Zhang, S. Dong, X. Zhou, S. Dong, N-doped carbon quantum dots/TiO
24 https://doi.org/10.1016/j.jphotochem.2016.04.012.
21
1 [11] S. Sakthivel, B. Neppolian, M. V Shankar, B. Arabindoo, M. Palanichamy, V.
7 https://doi.org/10.1016/j.mssp.2015.06.068.
8 [13] U.G. Akpan, B.H. Hameed, Parameters affecting the photocatalytic degradation of
9 dyes using TiO2-based photocatalysts: A review, J. Hazard. Mater. 170 (2009) 520–
10 529.
11 [14] Z. Luo, A.S. Poyraz, C.H. Kuo, R. Miao, Y. Meng, S.Y. Chen, T. Jiang, C. Wenos,
12 S.L. Suib, Crystalline mixed phase (anatase/rutile) mesoporous titanium dioxides for
14 https://doi.org/10.1021/cm5035112.
16 activities of various anatase and rutile forms of titania in pure liquid organic phases
18 https://doi.org/10.1021/jp9533584.
20 phase Ti 4 O 7 porous anode: Kinetics, products, and toxicity, Chem. Eng. J. 332
7 https://doi.org/10.1016/j.watres.2018.03.030.
8 [20] A. Ajmal, I. Majeed, R.N. Malik, M. Iqbal, M.A. Nadeem, I. Hussain, S. Yousuf,
11 2146. https://doi.org/10.1016/j.jece.2016.03.041.
12 [21] J. Singh, S.A. Khan, J. Shah, R.K. Kotnala, S. Mohapatra, Nanostructured TiO 2 thin
19 https://doi.org/10.1016/j.apcatb.2016.10.057.
20 [23] W. He, Y. Liu, J. Ye, G. Wang, Electrochemical degradation of azo dye methyl
2 https://doi.org/10.1016/j.chemosphere.2015.04.042.
6 https://doi.org/10.1016/j.jallcom.2016.09.119.
10 https://doi.org/10.1016/j.jelechem.2014.09.035.
11 [27] M.A.O. Soliu O. Ganiyu, Nihal Oturan, Clément Trellu, Stéphane Raffy, Marc
16 [28] C.F. Liu, C.P. Huang, C.C. Hu, C. Huang, A dual TiO 2 /Ti-stainless steel anode for
19 https://doi.org/10.1016/j.scitotenv.2018.12.224.
20 [29] E. Nurhayati, H. Yang, C. Chen, C. Liu, Y. Juang, C. Huang, C. chang Hu, Electro-
6 3603. https://doi.org/10.1016/j.watres.2009.05.035.
10 [33] W. Su, S. Wang, X. Wang, X. Fu, J. Weng, Plasma pre-treatment and TiO2 coating
13 [34] Y.H. Chen, Y.Y. Liu, R.H. Lin, F.S. Yen, Photocatalytic degradation of p-
16 https://doi.org/10.1016/j.jhazmat.2008.07.097.
23 using sludge derived carbon as catalytic particle electrodes and catalyst, J. Taiwan
4 [38] M. Zhou, L. Lei, The role of activated carbon on the removal of p-nitrophenol in an
6 https://doi.org/10.1016/j.chemosphere.2006.03.054.
10 [40] H.W. Chen, K.C. Huang, C.Y. Hsu, C.Y. Lin, J.G. Chen, C.P. Lee, L.Y. Lin, R.
11 Vittal, K.C. Ho, Electrophoretic deposition of TiO2 film on titanium foil for a
13 https://doi.org/10.1016/j.electacta.2010.10.099.
15 gap energy and flat band potential for application of modified TiO2 photoanodes in
17 https://doi.org/10.1016/j.jpowsour.2007.10.082.
18 [42] P. Geng, G. Chen, Magnéli Ti4O7 modified ceramic membrane for electrically-
19 assisted filtration with antifouling property, J. Memb. Sci. 498 (2016) 302–314.
20 https://doi.org/10.1016/j.memsci.2015.07.055.
23 degradation and inactivation of bacteria, Appl. Surf. Sci. 365 (2016) 171–179.
24 https://doi.org/10.1016/j.apsusc.2015.12.247.
26
1 [44] O.M. Ama, O.A. Arotiba, Exfoliated graphite/titanium dioxide for enhanced
4 https://doi.org/10.1016/j.jelechem.2017.09.015.
7 synthetic azo dye from wastewater by anodic oxidation process with an emphasis on
10 [46] M. José Martín de Vidales, L. Mais, C. Sáez, P. Cañizares, F.C. Walsh, M.A.
12 Methyl Orange on a TiO2 Nanotubular Anode Using a Flow Cell, Chem. Eng.
14 [47] J. Gong, W. Pu, C. Yang, J. Zhang, Tungsten and nitrogen co-doped TiO2 electrode
22 https://doi.org/10.1016/S0013-4686(01)00503-5.
3 https://doi.org/10.1016/j.jphotochem.2018.07.001.
10 https://doi.org/10.1016/j.electacta.2017.11.048.
11 [53] A. El-Ghenymy, F. Centellas, R.M. Rodríguez, P.L. Cabot, J.A. Garrido, I. Sirés, E.
15 https://doi.org/10.1016/j.electacta.2015.09.078.
16 [54] W. Zhou, L. Rajic, L. Chen, K. Kou, Y. Ding, X. Meng, Y. Wang, B. Mulaw, J. Gao,
20 https://doi.org/10.1016/j.electacta.2018.11.052.
21
28
Figure and Table Captions
Figure 1. (A) Electrochemical cell arrangement for PC, EO and PEC experiments and (B)
3D reactor for PEC’ and EF/PEC’ MO decolorization tests. (1) UV lamp, λ=365
nm, (2) Anode consisting on a PMMA cuvette modified with the semiconductor
films under study, (3) cathode consisting on -A, graphite rod and -B, AC packed
bed, (4) peristaltic pump and (5) dye solution containing tank.
Figure 2. SEM images and EDS spectra of TiO2, Ti4O7 and TiO2/Ti4O7 films prepared on
Figure 4. Decolorization kinetics (presented as the evolution of the fraction of the initial
presence (__) of UV light (λ=365 nm) at pH 3, using TiO2, Ti4O7 and TiO2/Ti4O7
semiconductor films.
(B) Ti4O7 and (C) TiO2/Ti4O7, PMMA modified cuvettes and applying different
Figure 7. Percentage of color removal (%CR) of MO for PEC’ and EF/PEC’ experiments
using the reactor shown in Figure 1-B. While the circles denote the experimental
data points, the solid lines correspond to simulations performed using equation
16. The experiments were carried out using 3V for electrical polarization, UV
Figure 8. Schematic representation of the MO color changing processes that take place in
processes on the surface and pores of the AC packed bed electrode, and (3) EF
Table 1. Experimental values for the mass of dye in the solution and at the AC surface at
the beginning of the experiment and kinetic constants, Ki, of equation 16 that fit
KEF 1.23x10-2
☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests: