Fe-Modified TiO2 Nanotube Layer as a Photochemically Versatile Material for the Degradation of Organic Pollutants in Water "2279

TiO2 nanotube layers (TNT) are prepared by electrochemical anodization of Ti foil in an electrolyte composed of ethylene glycol, ammonium fluoride, and water. The surface of TNT is modified by iron using spin-coating of Fe(NO3)3/IPA (isopropyl alcohol) solution of different concentrations (10 μM−100 mM). The as-prepared materials are annealed at 450 ◦C for 2 h to form crystalline Fe-TNT. The phase identification and surface morphology of the materials are investigated by XRD and SEM/EDX, respectively. The novelty of this work is based on the investigation of different photochemical processes that could occur simultaneously, and it includes mainly photocatalysis and Fenton-based processes since iron is a Fenton-active element and TiO2 is a photocatalyst. To this end, the degradation of caffeine, an organic pollutant, is performed under solar-like radiation at pH = 3 using different systems that are: (i) Fe-TNT material alone; (ii) a radical precursor alone (H2O2 = 1 mM); and (iii) Fe-TNT combined with H2O2. It is worth noting the degradation mechanism of the organic pollutants occurs via advanced oxidation processes where hydroxyl radicals have been identified as the main reactive oxygen species. One of the main goals of this work is to determine the contribution of the different involved photochemical processes (photocatalysis, photo-Fenton, and photolysis) along with the potential synergy between all these processes. To resume, this work provides new insights into the concept of photochemical versatility (using Fe-TNT), which is scarcely described in the literature.


Introduction
Water plays a vital role for all living beings. Although one third of the Earth's surface is covered with water, only 3% is consumable, out of which <1% is accessible [1]. The anthropogenic activities, especially those from industries, increase the rate of pollutants in the natural environment by releasing treated wastewaters that are not efficiently cleaned. Therefore, persistent and emerging organic pollutants including toxic chemicals, pharmaceutical compounds, pesticides, etc., are accumulated in the environment, thus causing harmful diseases not only to mankind, but also to the ecosystem [2]. This is the case with caffeine, which is one of the most used psychoactive substances, and which is often detected in natural waters [3]. It is essential to look for efficient wastewater treatment. Advanced oxidation processes (AOPs) have become an alternative technology for efficiently removing persistent organic pollutants from the water due to the generation of highly reactive inorganic radicals, including • OH, that oxidize and mineralize the pollutants [4]. 2

of 6
Among the various AOPs, the photocatalysis and the Fenton-based processes are promising due to their low cost, ambient reaction conditions, and easy operation procedures. In photocatalysis, a semiconductor photocatalyst can generate electron (e − ) and hole (h + ) pairs (Equation (1)), which in turn forms reactive oxygen species (ROS) such as O 2 •− and • OH (Equations (2) and (3)), under light irradiation with energy higher than the photocatalyst energy bandgap (E g ) [5,6]. The Fenton and photo-Fenton process is based on the generation of • OH in the presence of Fe(III), H 2 O 2, and light irradiation (Equations (4)-(6)) [7]. Thus, the combination of Fenton-based processes with photocatalysis is an effective strategy to improve the efficiency of water pollutant removal [8].
The aim of our work is to understand the simultaneous coupling of photocatalytic and Fenton-based processes that are triggered using a photoactive nanomaterial. To this end, we report the use of a TiO 2 nanotube layer that is modified at its surface by Fe(III) (Fe-TNT) as a promising photochemically versatile material for efficient degradation of caffeine in water. The study of Fe-TNT layers provides new insight into the concept of photochemical versatility as an innovative water treatment technology.

Materials and Methods
TNT layers were prepared by electrochemical anodization at 60 V for 4 h of 1.5 cm × 1.5 cm Ti foil (99.7%, Sigma Aldrich, St. Louis, MI, USA) in an electrolyte composed of ethylene glycol (EG, CentralChem) and ammonium fluoride (NH 4 F, CentralChem). The as-deposited TNT layers were then annealed at 450 • C in the air for 2 h. Fe surfacemodified TNT layers were prepared by spin-coating technique of Fe(NO 3 ) 3 . 9H 2 O solution of different concentrations (10 µM, 100 µM, 1 mM, 10 mM, 100 mM) in IPA, followed by annealing at 450 • C in the air for 2 h to obtain Fe(III) surface-modified TNT (Fe-TNT). The morphology of Fe-TNT was analyzed by SEM/EDX (Tescan Lyra III). The crystalline phase composition was determined by XRD (PANalytical X-ray; Cu K α , λ = 1.5418 Å).
The photo-induced degradation of caffeine solution (CAF, ReagentPlus ® , Sigma Aldrich; 20 ppm) using Fe-TNT with a different concentration of Fe was studied under solar-like light (HQI TS-OSRAM 400 W, 1.3 mW cm -2 in the range 335-380 nm). The photocatalysis and Fenton-based processes were performed with and without 1 mM H 2 O 2 , respectively. All the photo-induced experiments were performed at pH = 3 in 30 mL solution with constant air bubbling for 4 h. Before irradiation, the mixtures were placed in the dark for 30 min to reach adsorption/desorption equilibrium. To investigate the degradation extents of CAF, 0.5 mL of the mixture was sampled out every 30 min and quenched into 100 µL MeOH. The absorbance of caffeine and its degradation byproducts was followed at 272 nm by a UV-vis spectrophotometer (Jasco V-530). In addition, the stability of the Fe-TNT nanomaterials was investigated by determining the amount of Fe leached in the solution during CAF degradation. To this end, 0.5 mL is sampled out every 30 min and mixed with 60 µL of 0.5M ascorbic acid (to reduce aqueous Fe(III) into Fe(II)), 50 µL of 20 mM ferrozine solution (to form Fe(II)-ferrozine complex), and 100 µL of phosphate buffer solution. The complex absorbance, along with its extinction coefficient (27,900 cm −1 M −1 at 562 nm), were used to estimate the concentration of leached iron by the Beer-Lamber equation.

Results and Discussion
The XRD patterns of TNT and the different Fe-TNT (Figure 1) confirm the presence of the TiO 2 anatase phase (ICDD 03-065-5714). The diffractions of metallic Ti (ICCD 00-044-1294) are from the foil, which supports the nanotube layers [9]. From the XRD patterns, the presence of iron oxide or oxohydroxide cannot be detected by XRD for two reasons: (i) the annealing temperature is not high enough to get crystalline compound; and (ii) the amount of deposited Fe(III) is low. It worth noting that higher annealing temperature than 450 • C was not used to avoid conversion of anatase into rutile TiO 2 phase. To confirm the presence of iron on the surface of TNT, EDX analyses are performed (Table 1) and the iron ratio is estimated in the range 0.1-1.5 at%. Such analysis also supports XRD measurements since the atomic ratio Ti:O is approximatively 1:2 ( Table 1). The morphology of TNT and Fe-TNT is investigated by SEM ( Figure 2). The images reveal the nanotubular structure of Fe-TNT with an average diameter of the nanotube of about 120 nm.
leached in the solution during CAF degradation. To this end, 0.5 mL is sampled out every 30 min and mixed with 60 µL of 0.5M ascorbic acid (to reduce aqueous Fe(III) into Fe(II)), 50 µL of 20 mM ferrozine solution (to form Fe(II)-ferrozine complex), and 100 µL of phosphate buffer solution. The complex absorbance, along with its extinction coefficient (27900 cm −1 M −1 at 562 nm), were used to estimate the concentration of leached iron by the Beer-Lamber equation.

Results and Discussion
The XRD patterns of TNT and the different Fe-TNT (Figure 1) confirm the presence of the TiO2 anatase phase (ICDD 03-065-5714). The diffractions of metallic Ti (ICCD 00-044-1294) are from the foil, which supports the nanotube layers [9]. From the XRD patterns, the presence of iron oxide or oxohydroxide cannot be detected by XRD for two reasons: (i) the annealing temperature is not high enough to get crystalline compound; and (ii) the amount of deposited Fe(III) is low. It worth noting that higher annealing temperature than 450 °C was not used to avoid conversion of anatase into rutile TiO2 phase. To confirm the presence of iron on the surface of TNT, EDX analyses are performed (Table 1) and the iron ratio is estimated in the range 0.1-1.5 at%. Such analysis also supports XRD measurements since the atomic ratio Ti:O is approximatively 1:2 (Table 1). The morphology of TNT and Fe-TNT is investigated by SEM (Figure 2). The images reveal the nanotubular structure of Fe-TNT with an average diameter of the nanotube of about 120 nm.

Results and Discussion
The XRD patterns of TNT and the different Fe-TNT (Figure 1) confirm the of the TiO2 anatase phase (ICDD 03-065-5714). The diffractions of metallic T 00-044-1294) are from the foil, which supports the nanotube layers [9]. From patterns, the presence of iron oxide or oxohydroxide cannot be detected by XRD reasons: (i) the annealing temperature is not high enough to get crystalline com and (ii) the amount of deposited Fe(III) is low. It worth noting that higher a temperature than 450 °C was not used to avoid conversion of anatase into ru phase. To confirm the presence of iron on the surface of TNT, EDX analyses formed (Table 1) and the iron ratio is estimated in the range 0.1-1.5 at%. Such also supports XRD measurements since the atomic ratio Ti:O is approximatively ble 1). The morphology of TNT and Fe-TNT is investigated by SEM (Figure 2). ages reveal the nanotubular structure of Fe-TNT with an average diameter of th tube of about 120 nm.  The degradation extents of CAF after 4 h reactions with Fe-TNT under solar-like radiation with and without H 2 O 2 are presented in Table 2. In the absence of H 2 O 2 , the degradation of CAF increases from 22% to 30% and 25% using 10 µM Fe-TNT and 100 µM Fe-TNT, respectively, compared to non-modified TNT. This observation is explained by the increase in •OH production (data not shown) probably due to a positive effect of Fe(III) on the separation of e -/h + pair. In addition, photo-Fenton reaction (Equation (6)) might occur since a small amount of iron is leached into the solution. However, at higher Fe(III) concentration, the degradation efficiency of CAF strongly decreases to 0%. It can be a consequence of iron aggregates that deactivate photocatalytic TiO 2 sites. On the other hand, in the presence of H 2 O 2 , the degradation of CAF increases from 27% to 100% for TNT modified by Fe(III) in the range 10 µM-100 mM, respectively. The observed increase is due to a strong contribution of Fenton-based processes (Equations (4)-(6)) along with H 2 O 2 photolysis ( Table 2). The contribution of the generation of •OH caused by the reaction of H 2 O 2 with photogenerated emight be lower since the photocatalysis decreases with increasing Fe(III) contents in the absence of H 2 O 2 . However, for non-modified TNT, this latter process is quite significant with an increase in degradation extent from 22% to 41% without and with H 2 O 2 , respectively (Table 2), so assuming 5% of H 2 O 2 activation by photogenerated e -(and 12% of photolysis). The use of Fe-TNT at pH = 3 as both photocatalyst and Fenton catalyst exhibit some limits: the iron leaching (Table 3). However, such an observation is crucial since it confirms that Fenton-based reactions probably occur in the solution (or close to the nanomaterial/ solution interface). Although the iron leaching can be seen as a drawback, it can lead to further development of photocatalysts as loading materials for iron to trigger Fenton-based processes. Indeed, at circumneutral pH, iron leaching does not occur (data not shown), but a selective iron leaching can be performed using non-toxic complexing agents in order to simultaneously couple heterogeneous photocatalysis and homogeneous Fenton-based processes [10,11]. To understand such photochemical versatility in Fe-TNT nanomaterials, insights into the contribution of photocatalysis and Fenton-based processes is assessed using the 1 mM iron-surface modified-TNT under solar-like radiation in the presence of H 2 O 2 at pH = 3. Indeed, this Fe-TNT is an optimal sample to determine the contribution of each process (due to their relative weak contributions). A tentative assessment of the process contribution (among the 48% degradation extent of CAF) is about 16% photocatalysis along with activation of H 2 O 2 by photogenerated e -, 12% H 2 O 2 photolysis, and 27% of Fenton-based processes. Therefore, the use of Fe-TNT is a promising photochemically versatile material that required further development for potential application in water treatments.

Conclusions
Fe(III)-surface modified TiO 2 nanotube layers (Fe-TNT) were successfully prepared using electrochemical anodization technique followed by spin-coating method. The CAF degradation at pH = 3 under solar-like radiation by Fe-TNT is the most efficient in the presence of 1 mM H 2 O 2 due to a strong contribution of Fenton-based processes. It is worth noting the Fenton-based reactions do not occur at the surface of the materials, but probably in the solution, since iron leaching is observed. Although it can be seen as a drawback due to the limited stability of Fe-TNT, it can open new research directions for designing photocatalysts as loading materials for iron to trigger Fenton-based processes. Indeed, at circumneutral pH, such Fe-TNT are stable and a selective iron leaching using non-toxic complexing agents can be performed, so heterogeneous photocatalysis is simultaneously coupled to homogeneous Fenton-based processes. Therefore, the use of Fe-TNT layers as photochemically versatile materials are promising for application in wastewater treatment.