Synthesis of TiO2-Carbon Nanotubes through ball-milling method for mineralization of oxamic acid (OMA) by photocatalytic ozonation

Accepted Manuscript Title: Synthesis of TiO2 -Carbon Nanotubes through ball-milling method for mineralization of oxamic acid (OMA) by photocatalytic ozonation Authors: Carla A. Orge, O. Salom´e G.P. Soares, Joaquim L. Faria, M. Fernando R. Pereira PII: DOI: Reference:

S2213-3437(17)30531-6 https://doi.org/10.1016/j.jece.2017.10.030 JECE 1940

To appear in: Received date: Revised date: Accepted date:

12-5-2017 9-10-2017 13-10-2017

Please cite this article as: Carla A.Orge, O.Salom´e G.P.Soares, Joaquim L.Faria, M.Fernando R.Pereira, Synthesis of TiO2-Carbon Nanotubes through ball-milling method for mineralization of oxamic acid (OMA) by photocatalytic ozonation, Journal of Environmental Chemical Engineering https://doi.org/10.1016/j.jece.2017.10.030 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

Synthesis of TiO2-Carbon Nanotubes through ball-milling method for mineralization of oxamic acid (OMA) by photocatalytic ozonation Carla A. Orge*, O. Salomé G.P. Soares, Joaquim L. Faria, M. Fernando R. Pereira Laboratory of Separation and Reaction Engineering - Laboratory of Catalysis and Materials (LSRE-LCM) Faculdade de Engenharia, Universidade do Porto, Rua Dr. Roberto Frias s/n, 4200-465 Porto, Portugal *[email protected], [email protected], [email protected], [email protected]

1

Abstract Several TiO2-carbon nanotube composites were prepared by ball milling procedure. Prepared materials were characterized by several techniques, including N 2 equilibrium adsorption isotherms, thermogravimetric analysis, elemental analysis, X-ray diffraction spectra and transmission emission microscopy, and tested in the mineralization of oxamic acid (OMA) by photocatalytic ozonation. The influence of milling conditions was evaluated and the performance was compared with samples synthesized by conventional procedures. Independently of the milling time, a high amount of OMA, approximately 70%, is removed in the first 10 min of reaction by composites of commercial TiO2 (P25) and carbon nanotubes (MWCNT). In the best conditions, the milled sample presents a reaction rate constant of 0.099 min -1, in contrast to 0.082 min1

obtained with conventional composite. The presence of N-groups produces a negative

effect, especially when N-precursor was added during P25 composite preparation, removing 30% of OMA after 10 min of reaction and decreasing the removal rate to ≈0.050 min-1. The OMA degradation is significantly more efficient in the presence of the milled composites containing synthesized TiO 2 (SG) when compared to the composite prepared by the conventional procedure increased OMA removal from 40% to 70% after 30 min of reaction. The reaction rate constant of milled sample with original MWCNT (0.032 min-1) is considerably higher than observed with unmilled (0.012 min -1). A remarkable conversion is observed for all SG milled composites during the first 30 min of reaction and the presence of Fe is advantageous, since Fe can promote the O3 decomposition into HO● radicals and surface reactions. Keywords: titanium dioxide; carbon nanotubes; ball milling; photocatalytic ozonation; oxamic acid

2

1.

Introduction

Advanced Oxidation Processes (AOPs) were widely used to degrade organic pollutants from water and wastewaters. Hydroxyl radical (HO●) is the major oxidizing agent and its potential is 2.80 eV [1]. The combination of various AOPs may result in an enhanced generation of oxidative species, and consequently an accelerated degradation and mineralization of refractory organic compounds [2]. In this sense, an increasing number of worldwide researchers have paid attention to a new combined oxidation method called photocatalytic ozonation, which couples photoactivated semiconductors with ozone (O 3) [3]. Combining these two processes may considerably reduce their individual limitations and enhance the efficiency of the pollutant removal [4]. This combination, resulting in synergistic effects, is thought to be a promising technique for wastewater decontamination [2]. On one hand, the synergetic effects between semiconductormediated photocatalysis and ozonation not only effectively improve the utilization of O 3, but also decrease the recombination rate of photo-generated electrons (e -) and holes (h+). On the other hand, the decomposition of dissolved O 3 is primarily responsible for the formation of non-selective HO● radicals, which react with almost all organic molecules at a rate of 10-6-10-9 M-1s-1 [5]. In the photocatalytic ozonation system, TiO 2 is one of the widely investigated materials due to its nontoxicity, good stability and excellent photocatalytic activity [2]. According to the literature, in addition to direct ozonation, in the presence of TiO2 under radiation, O3 can generate HO● radicals by the formation of an ozonide radical (O 3●-) in the adsorption layer [6]: TiO2 + hν → TiO2 + h+ + e−

(1) O3 + e− → O.− 3

(2)

h+ + H2 O → HO. + H +

(3)

The generated O3●- species quickly reacts with H+ in the solution to give HO3● radical, which evolves to give O2 and HO● as shown below:

3

. + O.− 3 + H ⇄ HO3

(4)

HO.3 ⟶ HO. + O2

(5)

In absence of O3, dissolved O2 itself can accept TiO2 conduction band electron and generate O2●. + O.− 2 + H ⇄ HO2

(6)

2HO.2 → H2 O2 + O2

(7)

. − H2 O2 + O.− 2 → HO + HO + O2

(8)

In contrast with HO3●, this species cannot give HO● radicals in a single step. This mechanism requires a total of three steps for the generation of a single HO ● species, which is a less favoured situation if compared with the one electron needed through the O3●- reaction pathway. Kopf et al. proposed a further possible reaction path [7]. In addition to direct O3 attack and direct electron transfer from TiO 2 to the O3 molecule, they suggested that O2 should have an influence on the photocatalytic ozonation: TiO2 + hν → TiO2 + h+ + e−

(1)

e− + O2 ⟺ O.− 2

(9)

.− O3 + O.− 2 ⟶ O3 + O2

(10)

. + O.− 3 + H ⇄ HO3

(4)

HO.3 ⟶ HO. + O2

(5)

The recombination of electrons and positive holes could be interfered by the reaction between O3 and electrons on the surface of TiO 2, equation 2. Consequently, a larger number of radicals is produced, thereby accelerating the photocatalytic reaction. Although excellent results were observed with suspensions of fine TiO2 particles, often using P25 (Evonik; Degussa), some limitations are verified. Since the bandgap energy of TiO2 is around 3.2 eV, it is necessary UV-light below 380 nm for the photocatalytic process involving TiO2, which limits its efficiency when sunlight is used [8]. The quantum yield of TiO2 is lower as a result of quick recombination of photogenerated e -/h+ pairs,

4

declining therefore the process efficiency. Thus, modifications of TiO 2 by changing the electrical properties and inducing a shift of the band gap were widely studied in order to improve the utilization efficiency of visible light and the photocatalytic quantum efficiency of TiO2. Doping of TiO2 with metal or non-metal ions was one of the strategies used to this end [9]. However, the drawback of doping is that the charge transfer properties would be destroyed, once defect states can be induced and consequently e -/h+ recombination is promoted [10]. Therefore, several efforts have been done to increase the interfacial electron-transfer rate and consequently enhance the quantum efficiencies for photocatalysis [11]. Amongst a different transitional metals, the loading with Fe is advantageous, since the catalyst can be recovery by an external magnetic field, in addition to the easily integration of Fe3+ into the crystal lattice of TiO2 that it is expected that reduced the recombination of photo-electrons [12-15] .More recently, the dispersion of a noble metal, as Ag, on to Fe-TiO2 was considered in photocatalytic reactions for pollutants elimination. Shojaie et al. verified that when UV light irradiated on the surface of Fe-TiO2-Ag catalyst, e-/h+ pairs were generated. Fe3+ ions can act as e-/h+ trap, hence formation of Fe2+ and Fe4+ ions which are less stable as compared to Fe 3+ ions. So, tends to generate Fe 3+, producing HO● radical and O2 anion. [16]. Preparation of composites with TiO 2 and a carbon phase is also an available alternative to form heterojunctions, which can provide a potential driving force for the separation of photo-generated charge carriers [17-19]. Combination of TiO2 with carbon materials in simple mixtures or as nanocomposites has received some attention [20-25]. Mixedphase TiO2-based composites revealed high photocatalytic activity due to the formation of solid-solid interfaces that facilitate charge transfer and spatial separation, reducing e /h+ recombination, and interfacial defect sites that act as “hot spots” [26]. Narrowing of band gap by non-metal such as nitrogen and sulphur is another approach to enhance the catalytic activity of TiO 2 under visible light [27, 28]. Nitrogen has widely considered as non-metal dopant due to its chemical stability, low cost and non-poison 5

character. Additionally, a decrease in the particle size in nanometer range and an increase of active surface area are promoted with some doping methods, resulting a reduction in the e-/h+ pair recombination. In contrast, some researchers found that Ndoped TiO2 presented less catalytic activity than pure TiO2 due to the strongly localized N 2p states at the top of valence band, which would tend to trap photogenerated electrons, and also reduce the oxidation power and the mobility of holes. The anion vacancies caused by the N-doping were also considered as a factor resulting in the decrease of the photocatalytic efficiency [29, 30]. In this research, composites of TiO2 and multi-walled carbon nanotubes (TiO2/MWCNT) prepared by ball milling were investigated as catalysts for the photocatalytic ozonation of oxamic acid (OMA). OMA was selected as model compound, since it is a simple molecule and one of the most common final products resulted from the degradation of a wide range of pollutants with high refractory character to oxidation [31]. The influence of milling time, vibration frequency, addition of solvent, doping with nitrogen and properties of TiO2 and MWCNT in the composites preparation and catalytic activity was investigated. The introduction of Fe, during composite synthesis or previously into the carbon phase, was also considered. Additionally, prepared catalysts were characterized by a vast range of techniques, such as N 2 equilibrium adsorption isotherms, thermogravimetric analysis, elemental analysis, X-ray diffraction spectra and transmission emission microscopy, in order to evaluate their textural, chemical and structural properties. The kinetic results were analysed taking into account their characterization. 2. 2.1

Experimental Preparation of materials

The ball milling used to prepare composites with TiO2 and MWCNT was a Retsch MM200 equipped with zirconium oxide balls and grinding jars [32]. The concentration of TiO2 in the composite was always the same (90% wt.). The selected composition was based on the results of previous research [33]. A commercial MWCNT, Nanocyl 3100, was 6

received from Nanocyl SA Belgium. Two kinds of TiO 2 were tested, one commercially obtained by Evonik Degussa Iberica SA, sample P25, and other prepared by the sol-gel procedure as described in [33, 34], sample SG. Times between 15 and 90 min at constant vibration frequency (10 vibration/s) and vibration frequency between 5 and 15 vibration/s during 30 min were used. The addition of solvent (ethanol) and N-precursor (melamine or urea) was evaluated in different steps of preparation and with different thermal treatments. The effect of oxygen-containing surface groups and presence of Fe on MWCNT surface was also studied. The following identification of the samples was used: TiO 2/MWCNT_X_Y_Z, where X is the time of milling in minutes, Y is the milling frequency in vibration/s and Z corresponds to the addition of ethanol (E) or nitrogen precursor (M for melamine and U for urea). The addition of ethanol before the milling process was also studied and the sample was named as (TiO2/MWCNT_X_Y)_E. When ethanol was added, before or after the milling, the final material was calcined at 500 ºC for 2 h and after the addition of N-precursor the catalyst was thermally treated under nitrogen flow at 600 ºC for 1 h. When MWCNT were previously functionalized with different surface chemical properties, the samples were labelled as TiO2/(MWCNT_A)_X_Y where A is: M or U, when MWCNT were previously N-doped in the ball mill with melamine or urea, respectively, followed by a thermal treatment at 600 ºC using an easy handle method [35]; it also can be Oxi or 2%Fe when MWCNT were oxidized with 7 M HNO 3 or impregnated with 2% of Fe by the incipient wetness method, respectively, prepared as described elsewhere [36]. The catalytic activity of ball milled materials was compared with composites synthesized by conventional procedures with the same composition. P25 based composites were compared with the composite synthesized by the hydration-dehydration technique (P25/MWCNT) and SG materials with the composite obtained by the sol-gel procedure (SG/MWCNT) [33].

7

Summarizing, in Table 1 is reported the list of samples selected in this work and the correspondent preparation method (HD: hydration-dehydration technique; SG: sol-gel method; BM: ball milling procedure). 2.2

Characterization of materials

The textural characterization of the materials was based on the corresponding N 2 equilibrium adsorption isotherms, determined at −196 ºC with a Quantachrome Instruments NOVA 4200e apparatus. BET surface area (S BET ) of the samples was calculated. The relative amount of metal oxide in the composite was determined by thermogravimetric analysis (TG) under air in a STA 409 PC/4/H Luxx Netzsch thermal analyzer. Elemental analysis (EA) of N-doped materials was carried out on a vario MICRO cube analyzer from Elemental GmbH in CHNS mode. Each element (CHNS) was determined by combustion of the sample at 1050 °C. XRD (X-ray diffraction) spectra were recorded on a Philips X’Pert MPD diffractometer (CuKα = 0.15406 nm). Transmission emission microscopy (TEM) images were obtained on a LEO 906E microscope operating with an accelerating voltage of 120 kV. 2.3

Photocatalytic ozonation experiments

Experiments of photocatalytic ozonation were carried out in a glass immersion photochemical reactor (diameter: 60 mm; height: 250 mm) loaded with 0.5 g L -1 of catalyst concentration and OMA 1 mmol L -1 at natural pH (~ 2.8). The reactor was equipped with a Heraeus TQ 150 medium-pressure mercury vapour lamp located axially and a DURAN 50® glass cooling jacket was placed around the lamp, resulting the main emission lines at λexc = 365, 405, 436, 546 and 578 nm, which corresponds a spectral energy range between 2.15 and 3.40 eV. The agitation was maintained constant at 400 rpm in order to keep the reactor content perfectly mixed. The gas was bubbled in the reactor by a diffuser with 1 cm of diameter. The experiments were performed at constant gas flow rate (150 cm3 min-1) and constant inlet ozone concentration (50 g m-3). Ozone was produced from pure oxygen in a BMT 802X ozone generator. The concentration of ozone in the gas phase was monitored with a BMT 964 ozone analyzer. Ozone in the 8

gas phase leaving the reactor was removed in a series of gas washing bottles filled with potassium iodide solution. The experimental conditions carried out in this work were selected according optimal conditions previously obtained. 2.4

Analytical techniques

Due to absence of intermediates during OMA degradation, since it is directly mineralized, the kinetic results were evaluated in terms of OMA concentration. The pollutant concentration was followed by HPLC with Hitachi Elite Lachrom system equipped with an ultraviolet detector. The stationary phase was an Altech AO-100 column (300 mm × 6.5 mm) working at room temperature under isocratic elution with H2SO4 5 mM at a flow rate of 0.5 cm3 min-1. The injection volume was 15 µL, the detector wavelength was 200 nm and the retention time was 9.3 min. 3. 3.1

Results Characterization of catalysts

The Ti and N content (determined by thermal and elemental analyses, respectively) and the BET surface area (from the N2 adsorption-desorption isotherms) of the prepared composites are collected in Table 1. No significant differences were observed between the expected (90%) and the determined content of TiO2, with the exception of the samples where ethanol was added, that present a slightly higher amount. These results indicated that the ball-milling procedure allow to prepare materials with the desired composition. Analysing the results of N-amount determined by EA, the N-doped samples prepared with melamine have a higher percentage of N than the urea treated materials, which is in accordance with the literature [35]. The preparation of composites with MWCNT previously doped with N-precursor allows incorporating more N-groups. The starting Ndoped MWCNT used for the preparation of the composites were prepared by an easy method that allows incorporate a N-content of 7.6% and 0.6% when melamine or urea,

9

respectively, are used as N-precursor [35]; consequently, the amount of N is higher in the composites prepared with melamine. In terms of BET surface area, P25 based materials prepared with fixed vibration (10 vibration/s) and different times of ball milling presented some differences. The sample milled during 15 min present the lowest specific surface area and the sample milled during 30 min the highest, which can be justified by the opening or also breaking of the carbon nanotubes with the increase of the milling time. A slightly decrease was verified when the ball milling time increases from 30 to 60 or 90 min, probably due to the formation of some agglomerates [37]. †

determined by TG; * determined by EA; DL: detection limit.

In the case of composites prepared at fixed milling time (30 min) and different vibration, the sample prepared at low frequency, P25/MWCNT_30_5, had the lowest S BET, suggesting that this vibration frequency was not enough to promote an efficient milling between P25 and MWCNT. Increasing the milling vibration frequency to 10 vibration/s an increase in the BET surface area was observed, slightly decreasing thereafter with the increase up to 15 vibration/s. Vibration frequency has a similar effect on the change of the textural properties than the milling time, and in both cases it goes through an optimum. This trend was also verified with neat MWCNT modified by ball-milling [32]. The addition of ethanol during the ball milling process, P25/MWCNT_30_10_E sample, or after the ball milling, (P25/MWCNT_30_10)_E, decreases the BET surface area. The presence of ethanol during the milling can prevent the breaking of the CNTs [35] and consequently the interaction between TiO 2 and MWCNT, avoiding the formation of homogeneous materials [38]. The addition of the N-precursors (urea or melamine) during the milling leads to a decrease of the BET surface area, which can be related to the introduction of the Ngroups which may block the access of N2.

10

In the case of SG based materials, the ball milling process decreased the BET surface area when compared with composite synthesized by the sol-gel method and this reduction was more pronounced for the composite prepared with original carbon material (SG/MWCNT_30_10). XRD spectra of some of the prepared composites are depicted in Figure 1. The percentage and the mean diameter of the different crystalline phases are presented in Table 2, as well as the amount of Fe. P25 based catalysts had characteristic peaks associated with anatase and rutile phases and the addition of solvent or N-precursor did not change the amount of each phase. Concerning the crystallite sizes, a pronounced reduction of TiO 2 (rutile) particle diameter was observed when melamine or urea was added, while the TiO 2 (anatase) particle size only slightly decreased. Modifications in crystallite size were not verified when ethanol is introduced during the milling process P25/MWCNT_30_10_E. In the case of SG containing composites, the milling leads to an increase in TiO 2 (anatase) diameter. The milling also promotes the formation of rutile phase, especially in the case of composite with SG and MWCNT previously impregnated with Fe. The TEM micrographs of some prepared samples are depicted in Figure 2. TEM results allowed to quickly conclude that the composites prepared with TiO2 obtained by the sol-gel method presented a smaller crystallite diameter than samples containing commercial TiO2, which is in agreement with the XRD results. In the case of P25 based materials, the TEM results revealed that non-milled composite presented agglomerates of TiO2 on the carbon material surface. On the other hand, in the sample prepared in the ball mill it is notorious that CNT are shortened by breaking up the tubes due to the milling and the occurrence of agglomerates is lower. The presence of Fe was not confirmed in P25/(MWCNT_2%Fe)_30_10 catalyst, under the resolution used, due to the low amount of iron in the composite, but similarly to that observed for the P25/MWCNT_30_10 sample, the breaking of the tubes was also notorious. 11

In SG/MWCNT_30_10 sample, carbon nanotubes are covered with TiO 2, which can justify the low BET surface area verified. On the other hand, in the sample SG/(MWCNT_2%Fe) TiO2 is not well distributed on the carbon material surface and there are distinct agglomerates of CNT and TiO 2. 3.2 Photocatalytic ozonation results 3.2.1

Influence of time

The performance of the samples was evaluated in the photocatalytic ozonation of OMA at natural pH (≈ 2.8). OMA has been identified among the most common oxidation final products from organic pollutants degradation with high refractory character to oxidation processes [33]. In Figure 3 is depicted OMA dimensionless concentration during photocatalytic ozonation in the presence of composites obtained at different ball milling times and fixed vibration frequency (10 vibration/s). The results of the non-catalytic run were also introduced. A first order kinetic model was considered to fit the experimental data. The photocatalytic degradation of OMA can be described by a Langmuir-Hinshelwood kinetic model [39] and assuming that due to combined effects of light and O3 the concentration of HO● with respect to OMA is quasi constant. According to this model, the evolution of OMA concentration during non-catalytic run is described by the following equation: −

dC = k hom C dt

(11)

where khom (min-1) represents the first-order apparent rate constant and C (mM) is the concentration of OMA in each instant. Integration of Eq. (11), considering C = C 0, when t = 0, leads to: ln

C0 = k hom t C

(12)

In the presence of prepared materials, both homogeneous and heterogeneous degradation occur. Therefore, OMA removal rate is the sum of the two contributions:

12

−

dC = (k hom + k het )C dt

(13)

where khet (min-1) represents the first-order apparent rate constant for the heterogeneous degradation. Integration of Eq. (13), considering kapp = khom + khet and C = C0, when t = 0, leads to: ln

C0 = k app t C

(14)

For the fitting procedure, the value of k hom previously calculated was fixed and the results of apparent first-order rate constant values (k) are listed in Table 1. Photo-ozonation by itself only removes 24% of OMA after 60 min of reaction. Independently of the milling time, all composites prepared in the ball mill present better catalytic activity than the non-milled sample, especially the catalysts prepared until 60 min. The results suggested that no significant differences were obtained among samples prepared with a milling time between 15 and 60 min (k ≈ 0.1 min -1) and the increase of the milling time up to 60 min is not advantageous, decreasing the removal rate to 0.065 min-1. Thus, the milling time selected for further tests was 30 min. 3.2.2

Effect of vibration frequency

The results of OMA removal during photocatalytic ozonation with composites prepared at different vibration frequency and fixed milling time (30 min) are presented in Figure 4. The composites prepared at 5 and 10 vibration/s present high catalytic activity, whereas sample synthesized at 15 vibration/s presents a slower removal rate (k = 0.048 min-1) than non-milled composite. Similarly to that observed for the ball milling time, there is an optimal value of vibration frequency, the catalytic performance decreasing thereafter. Nevertheless, the effect of vibration frequency is more evident, since the differences between the catalytic activity of the composite synthesized at 15 vibration/s and the samples obtained at 5 or 10 vibration/s are more pronounced. Thus, according to the results, the remaining synthesis were performed with 10 vibration/s. In order to better understand the kinetic results, experiments of adsorption and photocatalysis in the presence of P25/MWCNT_30_10 were also carried out (see Figure 13

4). The results indicate that adsorption (P25/MWCNT_30_10_O 2) scarcely contributes to the elimination of OMA compared to photocatalytic ozonation, and can therefore be neglected. However, photocatalysis (P25/MWCNT_30_10_O 2+Light) resulted in a removal of approximately 75% of OMA after 60 min of reaction. Until 15 min of reaction, the pollutant removal rate by photocatalysis is considerable, but after that time the amount of OMA in solution remains constant. With these additional experiments, it is possible to conclude that adsorption does not contribute to OMA elimination and the simultaneous use of ozone, radiation and catalyst yields an improvement in the OMA removal compared to photocatalysis. 3.2.3

Addition of solvent

The addition of a solvent was considered in order to evaluate the effect in the homogenization of the sample [38, 40]. Addition of ethanol before and after the ball milling treatment, followed by a calcination treatment, was carried out and the results are shown in Figure 5. The addition of ethanol after the milling ((P25/MWCNT_30_10)_E) slightly decreases the OMA removal rate, which must be related to the low surface area of this sample (see Table 1). On the other hand, the introduction of ethanol during the preparation (P25/MWCNT_30_10_E), almost does not change the OMA removal rate, decreasing from 0.098 min-1 to 0.093 min-1, not justifying the addition of solvent. 3.2.4

Addition of N-precursor

Composites doped with nitrogen were studied by the addition of N-precursors, melamine and urea, during the milling. Figure 6 presents the results of photocatalytic ozonation of OMA in the presence of P25/MWCNT composites N-doped with urea and melamine. The kinetics results revealed that the doping of the composites with nitrogen, independently of the N-precursor, is not advantageous, since the undoped sample, P25/MWCNT_30_10, has a better performance. In some cases, TiO 2 modified by metal/non-metal doping presents poor activity under UV irradiation [41], since the doped species could act as recombination centers. Therefore, doping is not always associated 14

with enhanced photocatalytic activity, since reactivity is a complex function of dopant concentration, distribution, energy levels in the TiO2 lattice, d-electron configuration and light intensity [41, 42]. Sun et al. [43] attributed the lower UV activity of sulphur-doped samples to S centers acting as recombination sites. On the other hand, Tachikawa et al. [44] concluded that S sites did not act as recombination centers based on flash photolysis experiments. Analysis of diffuse reflectance spectroscopy allowed to determine that the yield of charge carriers with 355 nm excitation is greater for undoped TiO 2 than for STiO2. However, the efficiency of hole transport to the surface is comparable for the two materials and the S-centers do not act as special recombination points. Summarizing, contradictory arguments are reported about the chemistry resultant from doped TiO 2 [45] and the lower performance verified with doped samples under UV radiation could not be shown under visible light. 3.2.5

Effect of MWCNT

In order to evaluate the effect of MWCNT properties, composites with P25 and MWCNT with different textural and surface chemical properties were prepared. For that purpose, MWCNT previous functionalized with HNO 3 (to introduce oxygen containing surface groups), doped with nitrogen (using melamine and urea as N-precursor) and impregnated with Fe were tested and the results of OMA removal by photocatalytic ozonation are presented in Figure 7. A slightly decrease in OMA conversion was verified when the composite was prepared with oxidized MWCNT (k = 0.092 min-1). Similar behavior was previously observed with a composite prepared by the hydration-dehydration technique that revealed inferior performance to the corresponding composite with original MWCNT [33]. In the case of photocatalysis the presence of large amount of oxygenated groups on the surface of carbon nanotubes promotes the dispersion of TiO 2 particles in the composite and formation of Ti-O-C bonds, as in esterification reactions between the carboxylic acid groups of carbon nanotubes and the hydroxyl groups of TiO 2. On the other hand, ozonation processes are favoured by carbon nanotubes with low acidic character [46, 15

47]. The ozone decomposition into radicals on the surface of carbon materials is preceded by an adsorption step where ozone is adsorbed on the carbon surface. Ozone molecules have a higher affinity for basic carbons, which are known to have a high density of delocalized 𝜋 electrons on the basal planes [46]. Consequently, carbon materials with less acidic character have a higher catalytic activity for ozone decomposition. The experimental results suggest that the negative effect caused by the presence of acid groups on the MWCNT surface overlays the positive effect, since photocatalytic ozonation with composites of oxidized MWCNT was less efficient than with original MWCNT. The previously functionalization of the MWCNT with N-groups did not increase the catalytic activity when compared with the undoped sample. However, the composites prepared with MWCNT previously N-doped have higher OMA removal rate than the composites where N-precursor is added simultaneously with P25 and MWCNT during the composite synthesis. The composite of MWCNT previously impregnated with 2% of Fe, sample P25/(MWCNT_2%Fe)_30_10, presents a high performance, but slightly lower than the sample containing original MWCNT. Concluding, the catalytic activity of P25 based composites during OMA degradation by photocatalytic ozonation is compromised by the presence of oxygen, nitrogen or Fe on the MWCNT surface. 3.2.6

Influence of TiO2 properties

With the aim of evaluating the influence of TiO 2 properties, TiO2 synthesized by the solgel procedure (SG) was considered for the preparation of composites in the ball milling and tested during photocatalytic ozonation of OMA. The optimal conditions obtained with commercial TiO2 (P25) were considered in these experiments (ball milling time and frequency of 30 min and 10 vibration/s, respectively) and the results are presented in Figure 8.

16

In contrast with the composite obtained by the traditional sol-gel method, all ball milled composites allow a fast OMA decay in the first 30 min of reaction, removing approximately 70% of the initial concentration. After this period, some differences are observed between samples. The previous N-doping of the MWCNT was not advantageous, especially in the case of urea, suggesting that the presence of N-groups decreases the number of active sites. On the other hand, the presence of Fe on MWCNT surface is useful, increasing the OMA removal rate from 0.032 min-1 to 0.042 min-1. The best performance after 60 min of reaction was obtained with the composite SG/(MWCNT_2%Fe)_30_10, leading to 92% of OMA removal. It is expected that the presence of Fe promotes the O3 decomposition into HO● radicals and surface reactions. TEM images confirm the availability of carbon nanotubes surface to react. The high catalytic activity verified with SG based composites prepared by ball milling can be attributed to the good contact between anatase TiO 2 particles and MWCNT promoted by the increase of TiO2 diameter [21]. Additionally, the presence of rutile phase increased in the ball milling preparation. It is important to emphasize that P25 and SG are morphologically different, and consequently the interactions with MWCNT in the composites are not similar. P25 is constituted by 80:20 ratio of anatase/rutile phases and its exclusive polymorphic characteristics that significantly improved photocatalytic reactions are very difficult to achieve during TiO2 synthesis. The high performance verified with prepared materials when these two oxidation processes, ozonation and photocatalysis, were carried out together may be explained by the high production of HO● resulted from the combination of several reactions [48]. Predominantly, photocatalytic reactions begin by photoexciting the surface of photocatalyst with UV–Vis radiation, which can provide the appropriate band gap energy to generate photoactivated e -/h+ pairs. In parallel, O3 molecules can adsorb on the surface of the photocatalyst through three different interactions: physical adsorption, formation of weak hydrogen bonds with surface hydroxyl groups, and molecular or 17

dissociative adsorption into Lewis acid sites [49], each interaction resulting in the production of active oxygen radicals (O ●). Huang and Li showed these active oxygen radicals react with water molecules to produce HO● [50]. The photogenerated electrons on the photocatalyst surface react with adsorbed O2 and O3 molecules as electron acceptors, and these reactions are important to decrease the recombination rate of electron–hole (e-/h+) pairs [51, 52]. The recombination of e-/h+ negatively affects reduction and oxidation reactions on the photocatalyst surface, quantitatively decreasing the effective interactions and consequently reducing the performance of this technology. The electron affinity of O 3 (≈ 2.1 eV) is higher than that of O2 (≈ 0.44 eV), which can promote photocatalytic reactions in the presence of O 3 more efficiently than in the presence of O 2 [53]. The H2O2 molecules that are generated can also react with photoexcited electrons on the photocatalyst surface, forming hydroxyl radicals [54]. Ball milling reveals to be an effective method to enhance the performance of TiO2/MWCNT composites and when compared to P25 based materials the differences are reduced. Additionally, the method preparation is easy and simple and the equipment to manufacture the catalysts is not expensive and does not require high expertise, which is an advantage for industrial application. 4. Conclusions Composites of TiO2 and carbon nanotubes were successfully prepared by ball milling and extensively characterized by different techniques. All composites of commercial TiO2 (P25) and carbon nanotubes (MWCNT) synthesized at different ball milling time and fixed vibration frequency presented better performances in the photocatalytic ozonation of OMA, than the catalyst prepared by the conventional method, especially the materials synthesized until 60 min. Analysing the effect of ball milling frequency, composites prepared at 5 and 10 vibration/s have higher removal rates than the non-milled sample. The increase of vibration frequency for 15 vibration/s has a negative effect. The addition of ethanol during composite synthesis almost does not 18

change the catalytic activity. The introduction of N-groups in carbon nanotubes or during composite synthesis is not advantageous. Ball milled composites containing TiO 2 prepared by the sol-gel procedure (SG) presented higher removal rates than the conventional composite and no significant differences were observed between samples during the first half-hour of reaction. After that, the prepared samples have distinct performances and the composite containing Fe presents the best catalytic activity, achieving a nearly total OMA removal after 1 h of reaction. In conclusion, ball milling is a promising way for preparing TiO 2/MWCNT composites.

Acknowledgements This work is a result of project “AIProcMat@N2020 - Advanced Industrial Processes and Materials for a Sustainable Northern Region of Portugal 2020”, with the reference NORTE-01-0145-FEDER-000006, supported by Norte Portugal Regional Operational Programme (NORTE 2020), under the Portugal 2020 Partnership Agreement, through the European Regional Development Fund (ERDF) and of Project POCI-01-0145FEDER-006984 – Associate Laboratory LSRE-LCM funded by ERDF through COMPETE2020 - Programa Operacional Competitividade e Internacionalização (POCI) – and by national funds through FCT - Fundação para a Ciência e a Tecnologia. C. A. Orge acknowledges the FCT grant (BPD/90309/2012.). References

[1] J. Huang, X. Wang, Z. Pan, X. Li, Y. Ling, L. Li, Efficient degradation of perfluorooctanoic acid (PFOA) by photocatalytic ozonation, Chem. Eng. J. 296 (2016) 329-334. [2] J. Yin, G. Liao, J. Zhou, C. Huang, Y. Ling, P. Lu, L. Li, High performance of magnetic BiFeO3 nanoparticle-mediated photocatalytic ozonation for wastewater decontamination, Sep. Purif. Technol. 168 (2016) 134-140.

19

[3] A.M. Chávez, A. Rey, F.J. Beltrán, P.M. Álvarez, Solar photo-ozonation: A novel treatment method for the degradation of water pollutants, J. Hazard. Mater. 317 (2016) 36-43. [4] A.C. Mecha, M.S. Onyango, A. Ochieng, C.J.S. Fourie, M.N.B. Momba, Synergistic effect of UV–vis and solar photocatalytic ozonation on the degradation of phenol in municipal wastewater: A comparative study, J. Catal. 341 (2016) 116-125. [5] M. Cheng, G. Zeng, D. Huang, C. Lai, P. Xu, C. Zhang, Y. Liu, Hydroxyl radicals based advanced oxidation processes (AOPs) for remediation of soils contaminated with organic compounds: A review, Chem. Eng. J. 284 (2016) 582598. [6] T.E. Agustina, H.M. Ang, V.K. Vareek, A review of synergistic effect of photocatalysis and ozonation on wastewater treatment, J. Photochem. Photobiol. C: Photochem. Rev. 6 (2005) 264-273. [7] P. Kopf, E. Gilbert, S.H. Eberle, TiO2 photocatalytic oxidation of monochloroacetic acid and pyridine: Influence of ozone, J Photochem Photobio A 136 (2000) 163-168. [8] D. Chen, Z. Jiang, J. Geng, Q. Wang, D. Yang, Carbon and Nitrogen Codoped TiO2 with Enhanced Visible-Light Photocatalytic Activity, Carbon and Nitrogen Co-doped TiO2 with Enhanced Visible-Light Photocatalytic Activity 46 (2007) 2741-2746. [9] R. Yuan, B. Zhou, D. Hua, C. Shi, Enhanced photocatalytic degradation of humic acids using Al and Fe co-doped TiO2 nanotubes under UV/ozonation for drinking water purification, J. Hazard. Mater. 262 (2013) 527-538.

20

[10] S. Liu, L. Yang, S. Xu, S. Luo, Q. Cai, Photocatalytic activities of C–N-doped TiO2 nanotube array/carbon nanorod composite, Electrochem. Commun. 11 (2009) 1748-1751. [11] M.R. Hoffmann, S.T. Martin, Wy Choi, D.W. Bahnemann, Environmental applications of semiconductor photocatalysis, Chem. Rev. 95 (1995) 69–96. [12] S. Sood, A. Umar, S.K. Mehta, S.K. Kansal, Highly effective Fe-doped TiO2 nanoparticles photocatalysts for visible-light driven photocatalytic degradation of toxic organic compounds, J. Coll. Inter. Sci. 450 (2015) 213-223. [13] V.M. Vasconcelos, F.L. Migliorini, J.R. Steter, M.R. Baldan, N.G. Ferreira, M.R. de Vasconcelos Lanza, Electrochemical oxidation of RB-19 dye using a highly BDD/Ti: Proposed pathway and toxicity, J. Environ. Chem. Eng. 4 (2016) 3900-3909. [14] Y. Zhao, C. Tao, G. Xiao, G. Wei, L. Li, C. Liu, H. Su, Controlled synthesis and photocatalysis of sea urchin-like Fe3O4@TiO2@Ag nanocomposites, Nanoscale 8 (2016) 5313-5326. [15] H.S. Mazloomi Tabaei, M. Kazemeini, M. Fattahi, Preparation and characterization of visible light sensitive nano titanium dioxide photocatalyst, Sci. Iran. 19 (2012) 1626-1631. [16] A. Shojaie, M. Fattahi, S. Jorfi, B. Ghasemi, Hydrothermal synthesis of FeTiO2-Ag nano-sphere for photocatalytic degradation of 4-Chlorophenol (4-CP): Investigating the effect of hydrothermal temperature and time as well as calcination temperature, J. Environ. Chem. Eng. 5 (2017) 4564-4572. [17] A. Kongkanand, R.M. Domínguez, P.V. Kamat, Single wall carbon nanotube scaffolds for photoelectrochemical solar cells. Capture and transport of photogenerated electrons, Nano Lett. 7 (2007) 676-680.

21

[18] D. He, L. Yang, S. Kuang, Q. Cai, Fabrication and catalytic properties of Pt and

Ru

decorated

TiO2\CNTs

catalyst

for

methanol

electrooxidation,

Electrochem. Commun. 9 (2007) 2467-2472. [19] L. Yang, Y. Xiao, G. Zeng, S. Luo, S. Kuang, Q. Cai, Fabrication and characterization of Pt/C - TiO2 nanotube arrays as anode materials for methanol electrocatalytic oxidation, Energy Fuels 23 (2009) 3134-3138. [20] C.G. Silva, J.L. Faria, Photocatalytic Oxidation of Phenolic Compounds by Using a Carbon Nanotube-Titanium Dioxide Composite Catalyst, ChemSusChem 3 (2010) 609-618. [21] R.R.N. Marques, M.J. Sampaio, P.M. Carrapiço, C.G. Silva, S. MoralesTorres, G. Dražić, J.L. Faria, A.M.T. Silva, Photocatalytic degradation of caffeine: Developing solutions for emerging pollutants, Catal. Today 209 (2013) 108-115. [22] M.J. Sampaio, R.R.N. Marques, P.B. Tavares, J.L. Faria, A.M.T. Silva, C.G. Silva, Tailoring the properties of immobilized titanium dioxide/carbon nanotube composites for photocatalytic water treatment, J. Environ. Chem. Eng. 1 (2013) 945-953. [23] L.M. Pastrana-Martínez, S. Morales-Torres, S.K. Papageorgiou, F.K. Katsaros, G.E. Romanos, J.L. Figueiredo, J.L. Faria, P. Falaras, A.M.T. Silva, Photocatalytic behaviour of nanocarbon–TiO2 composites and immobilization into hollow fibres, Appl. Catal. B: Environ. 142–143 (2013) 101-111. [24] S. Morales-Torres, L.M. Pastrana-Martínez, J.L. Figueiredo, J.L. Faria, A.M.T. Silva, Design of graphene-based TiO2 photocatalysts—a review, Environ. Sci. Pollut. R. 19 (2012) 3676-3687. [25] L. Lin, H. Wang, W. Jiang, A.R. Mkaouar, P. Xu, Comparison study on photocatalytic oxidation of pharmaceuticals by TiO2-Fe and TiO2-reduced

22

graphene oxide nanocomposites immobilized on optical fibers, J. Hazard. Mater. 333 (2017) 162-168. [26] Y. Yao, G. Li, S. Ciston, R.M. Lueptow, K.A. Gray, Photoreactive TiO2/Carbon Nanotube Composites: Synthesis and Reactivity, Environ. Sci. Technol. 42 (2008) 4952-4957. [27] W. Techitdheera, J. Rattanarak, W. Mekprasart, W. Pecharapa, Influence of Milling Time, NH3 Additive and Annealing Temperature on Physical Properties of Modified Commercial TiO2 Powders via Ball Milling Process, Energy Procedia 56 (2014) 667-672. [28] D. Avisar, I. Horovitz, L. Lozzi, F. Ruggieri, M. Baker, M.-L. Abel, H. Mamane, Impact of water quality on removal of carbamazepine in natural waters by Ndoped TiO2 photo-catalytic thin film surfaces, J. Hazard. Mater. 244–245 (2013) 463-471. [29] H. Li, S. Yin, Y. Wang, T. Sato, Efficient persistent photocatalytic decomposition of nitrogen monoxide over a fluorescence-assisted CaAl2O4:(Eu, Nd)/(Ta, N)-codoped TiO2/Fe2O3, Appl. Catal. B: Environ. 132-133 (2013) 487492. [30] H. Zangeneh, A.A.L. Zinatizadeh, M. Habibi, M. Akia, M. Hasnain Isa, Photocatalytic oxidation of organic dyes and pollutants in wastewater using different modified titanium dioxides: A comparative review, J. Ind. Eng. Chem. 26 (2015) 1-36. [31] C.A. Orge, J.L. Faria, M.F.R. Pereira, Removal of oxalic acid, oxamic acid and aniline by a combined photolysis and ozonation process, Environ. Technol. 36 (2014) 1075-1083.

23

[32] O.S.G.P. Soares, A.G. Gonçalves, J.J. Delgado, J.J.M. Órfão, M.F.R. Pereira, Modification of carbon nanotubes by ball-milling to be used as ozonation catalysts, Catal. Today 249 (2015) 199-203. [33] C.A. Orge, M.F.R. Pereira, J.L. Faria, Photocatalytic ozonation of model aqueous solutions of oxalic and oxamic acids, Appl. Catal. B: Environ. 174–175 (2015) 113-119. [34] C.G. Silva, J.L. Faria, Anatase vs. rutile efficiency on the photocatalytic degradation of clofibric acid under near UV to visible irradiation, Photochem. Photobiol. Sci. 8 (2009) 705-711. [35] O.S.G.P. Soares, R.P. Rocha, A.G. Gonçalves, J.L. Figueiredo, J.J.M. Órfão, M.F.R. Pereira, Easy method to prepare N-doped carbon nanotubes by ball milling, Carbon 91 (2015) 114-121. [36] O.S.G.P. Soares, J.J.M. Órfão, M.F.R. Pereira, Pd−Cu and Pt−Cu Catalysts Supported on Carbon Nanotubes for Nitrate Reduction in Water, Ind. Eng. Chem. Res. 49 (2010) 7183-7192. [37] A.S. Bolokang, F.R. Cummings, B.P. Dhonge, H.M.I. Abdallah, T. Moyo, H.C. Swart, C.J. Arendse, T.F.G. Muller, D.E. Motaung, Characteristics of the mechanical milling on the room temperature ferromagnetism and sensing properties of TiO2 nanoparticles, Appl. Surf. Sci. 331 (2015) 362-372. [38] S. H. R. Fatemi Nayeri, M. R. Aboutalebi, J.V. Khaki, The effect of milling conditions on the mechanical alloying and combustion synthesis of tio2-al-c powder mixture, Vol. 3, Numbers 1 and 2 ed., Iranian Journal of Materials Science and Engineering, Winter and Spring, 2006.

24

[39] D. Ollis, C.G. Silva, J. Faria, Simultaneous photochemical and photocatalyzed liquid phase reactions: Dye decolorization kinetics, Catal. Today 240, Part A (2015) 80-85. [40] J. Rattanarak, W. Mekprasart, W. Pecharapa, W. Techitdheera, Investigation of milling time on particle size reduction and photo-induced activity performance of commercial TiO2 for optical energy harvesting applications, in: International Conference on Photonics Solutions, 2013, pp. 88831-88835. [41] M. Lazar, S. Varghese, S. Nair, Photocatalytic Water Treatment by Titanium Dioxide: Recent Updates, Catalysts 2 (2012) 572-601. [42] W. Choi, A. Termin, M.R. Hoffmann, The Role of Metal Ion Dopants in Quantum-Sized TiO2: Correlation between Photoreactivity and Charge Carrier Recombination Dynamics, J. Phys. Chem. 98 (1994) 13669-13679. [43] H. Sun, Y. Bai, Y. Cheng, W. Jin, N. Xu, Preparation and Characterization of Visible-Light-Driven Carbon−Sulfur-Codoped TiO2 Photocatalysts, Ind. Eng. Chem. Res. 45 (2006) 4971-4976. [44] T. Tachikawa, S. Tojo, K. Kawai, M. Endo, M. Fujitsuka, T. Ohno, K. Nishijima, Z. Miyamoto, T. Majima, Photocatalytic Oxidation Reactivity of Holes in the Sulfur- and Carbon-Doped TiO2 Powders Studied by Time-Resolved Diffuse Reflectance Spectroscopy, J. Phys. Chem. B. 108 (2004) 19299-19306. [45] E.M. Rockafellow, L.K. Stewart, W.S. Jenks, Is sulfur-doped TiO2 an effective visible light photocatalyst for remediation?, Appl. Catal. B: Environ. 91 (2009) 554-562. [46] A.G. Gonçalves, J.L. Figueiredo, J.J.M. Órfão, M.F.R. Pereira, Influence of the surface chemistry of multi-walled carbon nanotubes on their activity as ozonation catalysts, Carbon 48 (2010) 4369-4381.

25

[47] C. Orge, J. Sousa, F. Gonçalves, C. Freire, J. Órfão, M. Pereira, Development of Novel Mesoporous Carbon Materials for the Catalytic Ozonation of Organic Pollutants, Catal. Lett. 132 (2009) 1-9. [48] M. Mehrjouei, S. Müller, D. Möller, A review on photocatalytic ozonation used for the treatment of water and wastewater, Chem. Eng. J. 263 (2015) 209-219. [49] K.M. Bulanin, J.C. Lavalley, A.A. Tsyganenko, IR spectra of adsorbed ozone, Colloid. Surface A 101 (1995) 153-158. [50] H. Huang, W. Li, Destruction of toluene by ozone-enhanced photocatalysis: Performance and mechanism, Appl. Catal. B: Environ. 102 (2011) 449-453. [51] V. Augugliaro, M. Litter, L. Palmisano, J. Soria, The combination of heterogeneous photocatalysis with chemical and physical operations: A tool for improving the photoprocess performance, J. Photochem. Photobio. C 7 (2006) 127-144. [52] F.J. Rivas, F.J. Beltrán, O. Gimeno, M. Carbajo, Fluorene oxidation by coupling of ozone, radiation, and semiconductors: A mathematical approach to the kinetics, Ind. Eng. Chem. Res. 45 (2006) 166-174. [53] P. Pichat, L. Cermenati, A. Albini, D. Mas, H. Delprat, C. Guillard, Degradation processes of organic compounds over uv-irradiated TiO2. effect of ozone, Res. Chem. Intermediat. 26 (2000) 161-170. [54] E. Mena, A. Rey, B. Acedo, F.J. Beltrán, S. Malato, On ozone-photocatalysis synergism in black-light induced reactions: Oxidizing species production in photocatalytic ozonation versus heterogeneous photocatalysis, Chem. Eng. J. 204–206 (2012) 131-140.

26

Figure Captions Figure 1: X-ray diffractograms of prepared composites. Figure 2: TEM images of prepared materials. Figure 3: Evolution of the OMA dimensionless concentration during photocatalytic ozonation in the presence of ball-milled composites prepared with different milling times. Figure 4: Evolution of the dimensionless concentration of OMA during photocatalytic ozonation in the presence of ball-milled composites prepared at different vibration frequencies. Adsorption (O2) and photocatalysis (O2+Light) results. Figure 5: Evolution of the dimensionless concentration of OMA during photocatalytic ozonation catalysed by composites prepared in the presence of ethanol (inset: magnification of the initial reaction time). Figure 6: Evolution of the dimensionless concentration of OMA during photocatalytic ozonation in the presence of ball-milled composites N-doped with urea and melamine. Figure 7: Evolution of the dimensionless concentration of OMA during photocatalytic ozonation in the presence of composites with P25 and different MWCNT (inset: magnification of the initial reaction time). Figure 8: Evolution of the dimensionless concentration of OMA during photocatalytic ozonation in the presence of ball-milled SG based composites.

27

P25/(MWCNT_2%Fe)_30_10

SG/(MWCNT_2%Fe)_30_10

Intensity (a.u.)

Intensity (a.u.)

P25/(MWCNT_U)_30_10

P25/(MWCNT_M)_30_10

SG/MWCNT_30_10

SG/MWCNT P25/MWCNT_30_10_E 20

30

40

50

2

60

70

80

90

P25/MWCNT_30_10

20

Figr-1

30

40

50

2

60

70

80

90

Figure 1

28

Figure 2

29

SG/(MWCNT_2%Fe)_30_10

SG/MWCNT_30_10

SG/MWCNT

P25/(MCNT_2%Fe)_30_10

P25/MWCNT_30_10

P25/MWCNT

1.0

0.8

no catalyst P25/MWCNT P25/MWCNT_15_10 P25/MWCNT_30_10 P25/MWCNT_60_10 P25/MWCNT_90_10

C/C0

0.6

0.4

0.2

0.0 0

Figr-3

10

20

30

40

50

60

t/min

Figure 3

30

1.0

no catalyst P25/MWCNT P25/MWCNT_30_5 P25/MWCNT_30_10 P25/MWCNT_30_15 P25/MWCNT_30_10_O2

0.8

C/C0

0.6

P25/MWCNT_30_10_O2+Light

0.4

0.2

0.0 0

Figr-4

10

20

30

40

50

60

t/min

Figure 4

31

1.0

no catalyst (P25/MWCNT_30_10)_E P25/MWCNT_30_10_E P25/MWCNT_30_10

0.8

1.0

0.8

0.6

C/C0

C/C0

0.6

0.4

0.4

0.2

0.2 0.0

0

5

20

30

t/min

10

15

0.0 0

Figr-5

10

40

50

60

t/min

Figure 5

32

1.0

0.8

no catalyst P25/MWCNT_30_10_U P25/MWCNT_30_10_M P25/MWCNT_30_10

C/C0

0.6

0.4

0.2

0.0 0

Figr-6

10

20

30

40

50

60

t/min

Figure 6

33

1.0

no catalyst P25/(MWCNT_Oxi)_30_10 P25/(MWCNT_M)_30_10 P25/(MWCNT_U)_30_10 P25/(MWCNT_2% Fe)_30_10 P25/MWCNT_30_10

0.8

0.6

C/C0

1.0

0.8

0.4 C/C0

0.6

0.2

0.4

0.2

0.0

0

5

t/min

10

15

0.0 0

Figr-7

10

20

30

40

50

60

t/min

Figure 7

34

Figr-8 1.0

0.8

C/C0

0.6

0.4

no catalyst SG/MWCNT SG/MWCNT_30_10 SG/(MWCNT_M)_30_10 SG/(MWCNT_U)_30_10 SG/(MWCNT_2% Fe)_30_10

0.2

0.0 0

10

20

30

40

50

60

t/min

Figure 8

35

Table 1: Preparation method, BET surface area, amount of TiO 2 and N of synthesized composites and apparent first-order rate constant values of OMA degradation.

Sample

Preparation SBET (m2 g- k (min† method %TiO2 %N* 1 1 ) )

P25/MWCNT

HD

86

0

74

0.082

P25/MWCNT_15_10

BM

92

0

60

0.093

P25/MWCNT_30_10

BM

88

0

77

0.098

P25/MWCNT_60_10

BM

90

0

68

0.099

P25/MWCNT_90_10

BM

90

0

68

0.065

P25/MWCNT_30_5

BM

89

0

48

0.095

P25/MWCNT_30_15

BM

90

0

69

0.048

P25/MWCNT_30_10_E

BM

95

0

56

0.093

(P25/MWCNT_30_10)_E

BM

95

0

32

0.055

P25/MWCNT_30_10_M

BM

92

0.02

59

0.047

P25/MWCNT_30_10_U

BM

91

0.01

53

0.050

P25/(MWCNT_M)_30_10

BM

90

0.14

59

0.092

P25/(MWCNT_U)_30_10

BM

90


74

0.080

P25/(MWCNT_Oxi)_30_10

BM

91

0

68

0.092

P25/(MWCNT_2%Fe)_30_10 BM

89

0

80

0.097

SG/MWCNT

SG

88

0

157

0.012

SG/MWCNT_30_10

BM

88

0

98

0.032

SG/(MWCNT_M)_30_10

BM

88

0.7

134

0.028

SG/(MWCNT_U)_30_10

BM

89

0.03

112

0.024

SG/(MWCNT_2%Fe)_30_10

BM

89

0

118

0.042

36

Table 2: Crystalline properties and amount of Fe of the prepared catalysts determined by XRD. % Crystalline Phase

dcrystallites

Anatase

Rutile

Anatase

Rutile

P25/MWCNT_30_10

85

15

29

45

-

P25/MWCNT_30_10_ E

85

15

28

41

-

P25/(MWCNT_M)_30_10

83

17

25

33

-

P25/(MWCNT_U)_30_10

84

16

26

37

-

P25/(MWCNT_2%Fe)_30_10 85

15

28

42

0.16

SG/MWCNT

99

1

9

-

-

SG/MWCNT_30_10

97

3

15

-

-

SG/(MWCNT_2%Fe)_30_10

90

10

11

24

0.13

Sample

%Fe

37