Journal Pre-proof

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

V. Becerril-Estrada, I. Robles, C. Martínez-Sánchez, Luis A. Godínez

PII: S0013-4686(20)30364-9
DOI: https://doi.org/10.1016/j.electacta.2020.135972
Reference: EA 135972

To appear in: Electrochimica Acta

Received Date: 13 September 2019

Revised Date: 21 February 2020
Accepted Date: 24 February 2020

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.

© 2020 Published by Elsevier Ltd.

1 Study of TiO2/Ti4O7 photo-anodes inserted in an activated carbon packed bed

2 cathode: Towards the development of 3D-type photo-electro-Fenton reactors for

3 water treatment

4 V. Becerril-Estrada a, I. Robles a, C. Martínez-Sánchez b, Luis A. Godínez a, *

5
a
6 Centro de Investigación y Desarrollo Tecnológico en Electroquímica S. C., Parque Tecnológico

7 Querétaro, 76703 Sanfandila, Pedro Escobedo, Querétaro, México.

b
8 CONACYT – Centro de Investigación y Desarrollo Tecnológico en Electroquímica, Querétaro,

9 México.

10

11 * Corresponding author: lgodinez@cideteq.mx

12 Keywords: TiO2/Ti4O7; magneli; photoanode; advanced oxidation processes; electro-Fenton.

13

14

15

16 Abstract

17 In this work, commercially available Polymethyl-meta-acrylate (PMMA) spectroscopy

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

20 photocatalytic (PC), electro-oxidation (EO), and photoelectrochemical (PEC) experiments

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

23 electronic conductivity, the TiO2/Ti4O7 composite film behaves as a semiconductor

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

2 film, corresponded to 35 %, 46 % and 53 %, respectively). In order to test this film as an

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

9 with coupled electro-Fenton processes in the cathode. As expected, the combination

10 resulted in an effective decolorization of the dye solution that results from a complex

11 combination of processes. The experimental decolorization data was successfully fitted to a

12 pseudo-first order kinetic model so that a deeper understanding of the contribution of each

13 process in the reactor could be obtained.

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

23 be simultaneously applied to the system and combined on the surface of an electrode in a

24 synergistic fashion [3]. In this way, electro-oxidation (EO), photocatalytic (PC),

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

3 electrochemically or photo-electrochemically produced mixture of H2O2 and Fe(II) species,

4 are employed to produce a highly oxidant environment on the surface of either the anode,

5 the cathode, or both.

7 In photocatalytic as well as in photoelectrochemical studies, the most popular 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

12 photocatalytic activity than the later [15].

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

3 is often convenient to immobilize TiO2 or Ti4O7 in either non-conducting [20,21] or

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

9 polymethyl-meta-acrylate (PMMA) [32–34]. The advantage of preparing this type of

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

12 carbon (AC) packed beds which constitute a network of electrically interconnected

13 adsorbent particles on which electrochemical reactions take place [37]. The use of 3D

14 electrodes substantially increases the surface/volume ratio in electrochemical cells, thus

15 improving the efficiency of 3D reactors as compared to that of traditional 2D reactors [38].

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

21 performing films were introduced in a polarized AC packed bed so that a 3D

22 photoelectrochemical reactor could be tested and characterized as an alternative and

23 promising approach for the development of efficient and cost-competitive electrochemical

24 AOP based reactors for water treatment.

4
1

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

22 2.3. Semiconductor film characterization.

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

4 semiconductor assessment was made using the Kubelka-Munk approach [39].

6 2.4. Electrochemical and photocatalytic dye degradation experiments with

7 semiconductor modified PMMA substrates.

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

15 constant magnetic stirring at pH 3.0. In these experiments, potential differences of 1, 2 and

16 3 V between these two electrodes were applied for 60 min and samples for dye analysis

17 were extracted from the system every 10 minutes.

18

19 For PC experiments, a pencil-type UV Hg lamp (Spectroline 36-380, λ = 365 nm) was

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

23 standard absorbance measurement.

24
6
1 For PEC experiments, the setup consisted on a system in which polarization and UV

2 radiation were simultaneously applied to the semiconductor modified PMMA electrode.

4 2.5. Dye decolorization performance on a 3-D type AOP-PEC reactor

5 Figure 1-B shows the arrangement of a photo-assisted electro-Fenton reactor that

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

9 sample to a MO saturated aqueous solution for 72 h under stirring. The corresponding

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

12 place. Using the arrangement in Figure 1-B, photoelectrochemical (PEC’ to differentiate

13 from PEC experiments in the cell shown in Figure 1-A), and EF/PEC’ experiments were

14 performed applying a potential difference of 3V and introducing a pencil-type UV lamp (λ

15 = 365 nm) inside the PMMA modified cuvette.

16

17 In these experiments, the electrolyte to be treated consisted on 75 mL of a MO solution

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-

21 generation at the cathode surface.

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

3 Thermo Scientific spectrophotometer.

5 3. Results

6 3.1 Semiconductor film characterization.

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

14 that should be reflected by a more homogenous film, characterized by a large roughness

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

20 distinguishable particle aggregates. This is probably due to the better interparticle

21 interaction shown by Ti4O7 when compared to that displayed by TiO2.

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

7 reflectance spectroscopic measurements that are plotted following the Kubelka-Munk

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

11 although magneli is a reasonable electronic conductor, its performance is quite limited as a

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

14 spectra of TiO2 and of Ti4O7.

15

16 3.2 Photo-catalytic activity of the semiconductor films.

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

19 following the absorbance decrease (λ = 505nm) of a MO solution (initial concentration of

20 2.5x10-5 M) at pH 3. Control experiments in the absence of UV radiation were first

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 %).

23 Introducing UV radiation results in the continuous line responses shown in Figure 4.

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

2 mixed TiO2/Ti4O7 semiconductor material shows an intermediate behavior (14%), which is

3 in fact closer to that of TiO2.

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

7 photo-induced generation of hole/electron pairs (h+)/(e-) but also on the competition of

8 these species to react with the pollutant molecules and with nearby oppositely charged

9 carriers. In addition, different oxidant species can be formed as well. According to

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

14 TiO2 + hv → e-CB + h+VB (1)

15 e-CB + O2 → O2•- (2)

16 h+VB + H2O → •OH + H+ (3)

17

18 3.3 Electro- and photo-electrocatalytic tests.

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

2 applying 1, 2 and 3 V between the two electrodes.

4 M + H2O → MOx(•OH) + H+ + e- (4)

5 In Figure 5 the results of the evolution of the decolorization of the dye-contaminated

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

12 39 % for 1, 2 and 3 V, respectively). Relative to the second observation, comparison of the

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

15 worst is offered by TiO2 and the mixture TiO2/Ti4O7 is characterized by an experimental

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

22 reflected a combination of those previously observed for EO and PC experiments.

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

11 recombination is hindered and dye-consuming degradation reactions can be promoted

12 [49,50]. In other words, the effective photo-anodic performance of the TiO2/Ti4O7 film

13 observed in Figure 6 can be explained by the combination of the characteristic

14 photocatalytic and electrochemical activities that were observed for TiO2 in the PC and for

15 Ti4O7 in EO experiments.

16

17 3.4 Testing a TiO2/Ti4O7 mixed film in a photo-electro-Fenton 3D-type reactor.

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

20 TiO2/Ti4O7-modified PMMA cuvette as anode, an AC packed bed (0.5 g cm-2) as cathode

21 and a UV lamp and a power source to provide electromagnetic radiation and electric

22 potential stimuli, respectively. It is important to point out that in order to avoid

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

4 used previously saturated AC.

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

15 the so-called electro-Fenton process (EF) [52].

16

17 O2 + 2H + + 2e− → H2O2 (5)

18 Fe 2+ + H 2 O2 → Fe 3+ + • OH + OH − (6)

19 Fe3+ + e − → Fe 2+ (7)

20 Although photo-assisted electro-Fenton processes exists and are an important class of

21 AOPs [53], it is important to emphasize that incorporation of UV light in the experiments

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

6 equations 5 and 6).

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

23 electrolytic solution [54]. Therefore, as can be seen in Figure 8, there is a complex

24 combination of processes taking place in the photoelectrochemical reactor under study.

14
1 From the perspective of MO dye absorbance change, i.e., color removal, these processes

2 can be simplified as follows: (1) photoelectrochemical induced decolorization at the anode

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

6 the dye species at the cathode surface (EF).

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

10 absorbance measurements) can be expressed in terms of MO mass changes due to

11 adsorption and desorption taking place on the surface of AC (dMAD/dt) and to PEC’

12 (dMPEC’/dt) and EF (dMEF/dt) dye degradation processes.

13

ௗெೞ೚೗ ௗெಲವ ௗெುಶ಴’ ௗெಶಷ

14 = + + (8)
ௗ௧ ௗ௧ ௗ௧ ௗ௧

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

17 adsorption and desorption of MO), it is possible to write equations 9-11,

18

ௗெಲವ
19 = ‫ܭ‬ௗ௘௦ ‫ܯ‬௦௨௥ − ‫ܭ‬௔ௗ௦ ‫ܯ‬௦௢௟ (9)
ௗ௧

ௗெುಶ಴’
20 = −‫ܭ‬௉ா஼’ ‫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

4 dye desorption from the originally saturated AC packed bed.

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

8 of the experiment is rather high, the concentration of Msur can be assumed to be

9 approximately equal to Msur,0. Under these assumptions equation 8 can be written as,

10

ௗ(ெೞ೚೗ /ெೞ೚೗,బ ) ெೞ೚೗ ெೞೠೝ,బ

11 = (−‫ܭ‬௔ௗ௦ −‫ܭ‬௉ா஼’ ) + (‫ܭ‬ௗ௘௦ −‫ܭ‬ாி ) (12)
௧ ெೞ೚೗,బ ெೞ೚೗,బ

12 Where the constants k1, k2 and k3 can be defined as described by equations13-15,

13

14 ݇ଵ = −‫ܭ‬௔ௗ௦ − ‫ܭ‬௉ா஼’ (13)

15 ݇ଶ = ‫ܭ‬ௗ௘௦ − ‫ܭ‬ாி (14)

ெ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

20 of the solution) is observed to depend on time in an exponential way.

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

2 the EF contribution to the process.

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

18 (the contribution of the electro-Fenton decolorization process) and in this context, it is

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,

22 Msur and Msol are different in at least one order of magnitude).

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

4 structure, photocatalytic and electrochemical activity towards the decolorization of a MO

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

7 electrocatalytic activity, respectively, the TiO2/Ti4O7 mixture combines the properties of

8 the two materials, thus becoming the best semiconductor film for the photoelectrochemical

9 MO degradation tests surveyed.

10

11 The preparation of the TiO2/Ti4O7 composite photoanode on the external surface of

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

16 electrochemical reactor in which electro-Fenton events on the surface of the polarized AC

17 packed bed were successfully coupled to the photoelectrochemical anodic degradation

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

24 future design of a novel photo-electro-Fenton 3D reactor in which the adsorption, photo-

19
1 anodic and electro-Fenton cathodic processes, as well as the area/volume ratio, could be

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

7 for a graduate fellowship.

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

12 activated carbon, Environ. Sci. Technol. 47 (2013) 7927–7933.

13 https://doi.org/10.1021/es401320e.

14 [2] E. Brillas, F.J. Rodríguez-Valadez, A. El-Ghenymy, L.A. Godínez, O. García-

15 Rodríguez, J.A. Bañuelos, Advanced oxidation treatment of malachite green dye

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.

19 Ortiz-Frade, L. a. Godínez, Titanium dioxide doped with transition metals (M x

20 Ti1−x O2, M: Ni, Co): synthesis and characterization for its potential application as

21 photoanode, J. Nanoparticle Res. 13 (2011) 3313–3325.

22 https://doi.org/10.1007/s11051-011-0245-y.

23 [4] I. Sirés, E. Brillas, M.A. Oturan, M.A. Rodrigo, M. Panizza, Electrochemical

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.

2 [5] S. Garcia-Segura, E. Brillas, Applied photoelectrocatalysis on the degradation of

3 organic pollutants in wastewaters, J. Photochem. Photobiol. C Photochem. Rev. 31

4 (2017) 1–35. https://doi.org/10.1016/j.jphotochemrev.2017.01.005.

5 [6] N. Oturan, C.T. Aravindakumar, H. Olvera-Vargas, M.M. Sunil Paul, M.A. Oturan,

6 Electro-Fenton oxidation of para-aminosalicylic acid: degradation kinetics and

7 mineralization pathway using Pt/carbon-felt and BDD/carbon-felt cells, Environ. Sci.

8 Pollut. Res. 25 (2018) 20363–20373. https://doi.org/10.1007/s11356-017-9309-6.

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

11 Environ. 257 (2019) 117907. https://doi.org/10.1016/j.apcatb.2019.117907.

12 [8] H. Safajou, H. Khojasteh, M. Salavati-Niasari, S. Mortazavi-Derazkola, Enhanced

13 photocatalytic degradation of dyes over graphene/Pd/TiO2 nanocomposites: TiO2

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.

17 Suib, Mesoporous TiO2 modified with carbon quantum dots as a high-performance

18 visible light photocatalyst, Appl. Catal. B Environ. 189 (2016) 26–38.

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

21 2 hybrid composites with enhanced visible light driven photocatalytic activity

22 toward dye wastewater degradation and mechanism insight, J. Photochem.

23 Photobiol. A Chem. 325 (2016) 104–110.

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.

2 Murugesan, Solar photocatalytic degradation of azo dye: comparison of

3 photocatalytic efficiency of ZnO and TiO2, 77 (2003) 65–82.

4 [12] N. Shaham-Waldmann, Y. Paz, Away from TiO2: A critical minireview on the

5 developing of new photocatalysts for degradation of contaminants in water, Mater.

6 Sci. Semicond. Process. 42 (2016) 72–80.

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

13 visible light photocatalytic activity, Chem. Mater. 27 (2015) 6–17.

14 https://doi.org/10.1021/cm5035112.

15 [15] A. Sclafani, J.M. Herrmann, Comparison of the photoelectronic and photocatalytic

16 activities of various anatase and rutile forms of titania in pure liquid organic phases

17 and in aqueous solutions, J. Phys. Chem. 100 (1996) 13655–13661.

18 https://doi.org/10.1021/jp9533584.

19 [16] S. Liang, H. Lin, X. Yan, Q. Huang, Electro-oxidation of tetracycline by a Magnéli

20 phase Ti 4 O 7 porous anode: Kinetics, products, and toxicity, Chem. Eng. J. 332

21 (2018) 628–636. https://doi.org/10.1016/j.cej.2017.09.109.

22 [17] J. Maragatha, C. Rani, S. Rajendran, S. Karuppuchamy, Microwave synthesis of

23 nitrogen doped Ti4O7 for photocatalytic applications, Phys. E Low-Dimensional

24 Syst. Nanostructures. 93 (2017) 78–82. https://doi.org/10.1016/j.physe.2017.05.020.

22
1 [18] L. Guo, Y. Jing, B.P. Chaplin, Development and Characterization of Ultrafiltration

2 TiO2 Magnéli Phase Reactive Electrochemical Membranes, Environ. Sci. Technol.

3 50 (2016) 1428–1436. https://doi.org/10.1021/acs.est.5b04366.

4 [19] J. Wang, D. Zhi, H. Zhou, X. He, D. Zhang, Evaluating tetracycline degradation

5 pathway and intermediate toxicity during the electrochemical oxidation over a

6 Ti/Ti4O7 anode, Water Res. 137 (2018) 324–334.

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,

9 Zeshan, G. Mustafa, M.I. Zafar, M.A. Nadeem, Photocatalytic degradation of textile

10 dyes on Cu2O-CuO/TiO2 anatase powders, J. Environ. Chem. Eng. 4 (2016) 2138–

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

13 films prepared by RF magnetron sputtering for photocatalytic applications, Appl.

14 Surf. Sci. 422 (2017) 953–961. https://doi.org/10.1016/j.apsusc.2017.06.068.

15 [22] R. Xie, X. Meng, P. Sun, J. Niu, W. Jiang, L. Bottomley, D. Li, Y. Chen, J.

16 Crittenden, Electrochemical oxidation of ofloxacin using a TiO2-based SnO2-

17 Sb/polytetrafluoroethylene resin-PbO2 electrode: Reaction kinetics and mass transfer

18 impact, Appl. Catal. B Environ. 203 (2017) 515–525.

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

21 orange by anodic oxidation on Ti4O7 electrodes, J. Mater. Sci. Mater. Electron. 29

22 (2018) 14065–14072. https://doi.org/10.1007/s10854-018-9538-6.

23 [24] L.C. Almeida, B.F. Silva, M.V.B. Zanoni, Photoelectrocatalytic/photoelectro-Fenton

24 coupling system using a nanostructured photoanode for the oxidation of a textile

23
1 dye: Kinetics study and oxidation pathway, Chemosphere. 136 (2015) 63–71.

2 https://doi.org/10.1016/j.chemosphere.2015.04.042.

3 [25] C. Fu, M. Li, H. Li, C. Li, X. guo Wu, B. Yang, Fabrication of Au

4 nanoparticle/TiO2hybrid films for photoelectrocatalytic degradation of methyl

5 orange, J. Alloys Compd. 692 (2017) 727–733.

6 https://doi.org/10.1016/j.jallcom.2016.09.119.

7 [26] L.C. Almeida, B.F. Silva, M.V.B. Zanoni, Combined photoelectrocatalytic/electro-

8 Fenton process using a Pt/TiO2NTs photoanode for enhanced degradation of an azo

9 dye: A mechanistic study, J. Electroanal. Chem. 734 (2014) 43–52.

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

12 Cretin, Christel Causserand, Electrochemical Abatement of Analgesic Antipyretic

13 4‐Aminophenazone using Conductive Boron‐Doped Diamond and Sub‐

14 Stoichiometric Titanium Oxide Anodes: Kinetics, Mineralization and Toxicity

15 Assessment, ChemElectroChem. 6 (2019) 1808.

16 [28] C.F. Liu, C.P. Huang, C.C. Hu, C. Huang, A dual TiO 2 /Ti-stainless steel anode for

17 the degradation of orange G in a coupling photoelectrochemical and photo-electro-

18 Fenton system, Sci. Total Environ. 659 (2019) 221–229.

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-

21 photocatalytic fenton decolorization of orange g using mesoporous TiO2/stainless

22 steel mesh photo-electrode prepared by the sol-gel dip-coating method, Int. J.

23 Electrochem. Sci. 11 (2016) 3615–3632. https://doi.org/10.20964/110346.

24 [30] A. Danion, J. Disdier, C. Guillard, O. Païssé, N. Jaffrezic-Renault, Photocatalytic

24
1 degradation of imidazolinone fungicide in TiO 2-coated optical fiber reactor, Appl.

2 Catal. B Environ. 62 (2006) 274–281. https://doi.org/10.1016/j.apcatb.2005.08.008.

3 [31] K. Esquivel, L.G. Arriaga, F.J. Rodríguez, L. Martínez, L. a. Godínez, Development

4 of a TiO2 modified optical fiber electrode and its incorporation into a

5 photoelectrochemical reactor for wastewater treatment, Water Res. 43 (2009) 3593–

6 3603. https://doi.org/10.1016/j.watres.2009.05.035.

7 [32] L. Rizzo, J. Koch, V. Belgiorno, M.A. Anderson, Removal of methylene blue in a

8 photocatalytic reactor using polymethylmethacrylate supported TiO2 nanofilm,

9 Desalination. 211 (2007) 1–9. https://doi.org/10.1016/j.desal.2006.02.081.

10 [33] W. Su, S. Wang, X. Wang, X. Fu, J. Weng, Plasma pre-treatment and TiO2 coating

11 of PMMA for the improvement of antibacterial properties, Surf. Coatings Technol.

12 205 (2010) 465–469. https://doi.org/10.1016/j.surfcoat.2010.07.013.

13 [34] Y.H. Chen, Y.Y. Liu, R.H. Lin, F.S. Yen, Photocatalytic degradation of p-

14 phenylenediamine with TiO 2 -coated magnetic PMMA microspheres in an aqueous

15 solution, J. Hazard. Mater. 163 (2009) 973–981.

16 https://doi.org/10.1016/j.jhazmat.2008.07.097.

17 [35] B. Zhang, Y. Hou, Z. Yu, Y. Liu, J. Huang, L. Qian, J. Xiong, Three-dimensional

18 electro-Fenton degradation of Rhodamine B with efficient Fe-Cu/kaolin particle

19 electrodes: Electrodes optimization, kinetics, influencing factors and mechanism,

20 Sep. Purif. Technol. 210 (2019) 60–68. https://doi.org/10.1016/j.seppur.2018.07.084.

21 [36] B. Hou, H. Han, S. Jia, H. Zhuang, P. Xu, K. Li, Three-dimensional heterogeneous

22 electro-Fenton oxidation of biologically pretreated coal gasification wastewater

23 using sludge derived carbon as catalytic particle electrodes and catalyst, J. Taiwan

24 Inst. Chem. Eng. 60 (2016) 352–360. https://doi.org/10.1016/j.jtice.2015.10.032.

25
1 [37] C. Zhang, Y. Jiang, Y. Li, Z. Hu, L. Zhou, M. Zhou, Three-dimensional

2 electrochemical process for wastewater treatment: A general review, Chem. Eng. J.

3 228 (2013) 455–467. https://doi.org/10.1016/j.cej.2013.05.033 Review.

4 [38] M. Zhou, L. Lei, The role of activated carbon on the removal of p-nitrophenol in an

5 integrated three-phase electrochemical reactor, Chemosphere. 65 (2006) 1197–1203.

6 https://doi.org/10.1016/j.chemosphere.2006.03.054.

7 [39] R. López, R. Gómez, Band-gap energy estimation from diffuse reflectance

8 measurements on sol-gel and commercial TiO 2: A comparative study, J. Sol-Gel

9 Sci. Technol. 61 (2012) 1–7. https://doi.org/10.1007/s10971-011-2582-9.

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

12 flexible dye-sensitized solar cell, Electrochim. Acta. 56 (2011) 7991–7998.

13 https://doi.org/10.1016/j.electacta.2010.10.099.

14 [41] M. Radecka, M. Rekas, A. Trenczek-Zajac, K. Zakrzewska, Importance of the band

15 gap energy and flat band potential for application of modified TiO2 photoanodes in

16 water photolysis, J. Power Sources. 181 (2008) 46–55.

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.

21 [43] L. Korösi, M. Prato, A. Scarpellini, J. Kovács, D. Dömötör, T. Kovács, S. Papp, H 2

22 O 2 -assisted photocatalysis on flower-like rutile TiO 2 nanostructures: Rapid dye

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

2 photoelectrochemical degradation of methylene blue dye under simulated visible

3 light irradiation, J. Electroanal. Chem. 803 (2017) 157–164.

4 https://doi.org/10.1016/j.jelechem.2017.09.015.

5 [45] H. Hamad, D. Bassyouni, E.S. El-Ashtoukhy, N. Amin, M. Abd El-Latif,

6 Electrocatalytic degradation and minimization of specific energy consumption of

7 synthetic azo dye from wastewater by anodic oxidation process with an emphasis on

8 enhancing economic efficiency and reaction mechanism, Ecotoxicol. Environ. Saf.

9 148 (2018) 501–512. https://doi.org/10.1016/j.ecoenv.2017.10.061.

10 [46] M. José Martín de Vidales, L. Mais, C. Sáez, P. Cañizares, F.C. Walsh, M.A.

11 Rodrigo, C. de A. Rodrigues, C. Ponce de León, Photoelectrocatalytic Oxidation of

12 Methyl Orange on a TiO2 Nanotubular Anode Using a Flow Cell, Chem. Eng.

13 Technol. 39 (2016) 135–141. https://doi.org/10.1002/ceat.201500085.

14 [47] J. Gong, W. Pu, C. Yang, J. Zhang, Tungsten and nitrogen co-doped TiO2 electrode

15 sensitized with Fe-chlorophyllin for visible light photoelectrocatalysis, Chem. Eng.

16 J. 209 (2012) 94–101. https://doi.org/10.1016/j.cej.2012.07.137.

17 [48] S. V. Mohite, V. V. Ganbavle, K.Y. Rajpure, Photoelectrocatalytic activity of

18 immobilized Yb doped WO3 photocatalyst for degradation of methyl orange dye, J.

19 Energy Chem. 26 (2017) 440–447. https://doi.org/10.1016/j.jechem.2017.01.001.

20 [49] J. Luo, M. Hepel, Photoelectrochemical degradation of naphthol blue black diazo

21 dye on WO 3 film electrode, Electrochim. Acta. 46 (2001) 2913–2922.

22 https://doi.org/10.1016/S0013-4686(01)00503-5.

23 [50] Z. Shao, Y. Wang, Y. Zhang, G. Zhu, X. Yang, M. Zhong, Electrochemical

24 deposition synthesis of ZnO-NA/Cu 2 O-NPs type-II hierarchical heterojunction for

27
1 enhanced photoelectrochemical degradation of methyl orange (MO), J. Photochem.

2 Photobiol. A Chem. 364 (2018) 657–670.

3 https://doi.org/10.1016/j.jphotochem.2018.07.001.

4 [51] J. Ramírez, L.A. Godínez, M. Méndez, Y. Meas, F.J. Rodríguez, Heterogeneous

5 photo-electro-Fenton process using different iron supporting materials, J. Appl.

6 Electrochem. 40 (2010) 1729–1736. https://doi.org/10.1007/s10800-010-0157-z.

7 [52] Z. Zhang, H. Meng, Y. Wang, L. Shi, X. Wang, S. Chai, Fabrication of

8 graphene@graphite-based gas diffusion electrode for improving H2O2 generation in

9 Electro-Fenton process, Electrochim. Acta. 260 (2018) 112–120.

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.

12 Brillas, Comparative use of anodic oxidation, electro-Fenton and photoelectro-

13 Fenton with Pt or boron-doped diamond anode to decolorize and mineralize

14 Malachite Green oxalate dye, Electrochim. Acta. 182 (2015) 247–256.

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,

17 Y. Qin, A.N. Alshawabkeh, Activated carbon as effective cathode material in iron-

18 free Electro-Fenton process: Integrated H 2 O 2 electrogeneration, activation, and

19 pollutants adsorption, Electrochim. Acta. 296 (2019) 317–326.

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

the surface of PMMA substrates.

Figure 3. Kubelka-Munk representation of the Diffuse reflectance spectroscopy responses

for TiO2, Ti4O7 and the TiO2/Ti4O7 films.

Figure 4. Decolorization kinetics (presented as the evolution of the fraction of the initial

concentration) of a 2.5x10-5 M solution of MO in the absence (- - -) and in the

presence (__) of UV light (λ=365 nm) at pH 3, using TiO2, Ti4O7 and TiO2/Ti4O7

semiconductor films.

Figure 5. EO induced decolorization curves of 2.5x10-5 M, MO solutions, using (A) TiO2,

(B) Ti4O7 and (C) TiO2/Ti4O7, PMMA modified cuvettes and applying different

potentials. (D) Shows a comparison of the MO decolorization kinetics induced

by 2 V polarization for each one of the three films under study.

Figure 6. PEC induced decolorization curves of MO solutions using TiO2, Ti4O7 and

TiO2/Ti4O7 semiconductor films. Initial MO concentration 2.5x10-5 M, applied

potential 3 V and UV light with λ =365 nm.

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

light at λ =365 nm, a MO-saturated AC packed bed cathode and a cuvette

modified with a TiO2/Ti4O7 film as photo-anode.

Figure 8. Schematic representation of the MO color changing processes that take place in

the photoelectrochemical reactor under study. (1) photoelectrochemical

contribution in the anode (PEC’), (2) MO dye adsorption/desorption (AD)

processes on the surface and pores of the AC packed bed electrode, and (3) EF

in the vicinity of the cathode surface.

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

the experimental data in Figure 7.

Kinetic Constant, Ki Ki /min-1 R2

Kads 1.858 0.999

Kdes 3.5 x10-2 0.999

KEF 1.23x10-2

KPEC’ 1.21x10-2 ---

Msur,0 (mg) = 16.41

Msol,0 (mg) = 0.681
Declaration of interests

☒ 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: