WO 3 Fibers/g-C 3 N 4 Z-Scheme Heterostructure Photocatalysts for Simultaneous Oxidation/Reduction of Phenol/Cr (VI) in Aquatic Media

: A sequence of WO 3 /g-C 3 N 4 composites was synthesized at various % weight ratios (1, 5, 6.5, 8, 10, and 15%) of WO 3 into g-C 3 N 4 via electrospinning and wet-mixing method. The prepared photocatalytic materials were characterized by X-ray diffraction (XRD), Fourier transform-infrared (FT-IR) spectroscopy, UV–vis diffuse reflection spectroscopy (DRS), scanning electron microscopy (SEM), N 2 porosimetry and dynamic light scattering (DLS). Electrospun fibers of WO 3 with diameter 250–300 nm was prepared using polyvinylpyrrolidone (PVP) polymer and used for the synthesis of composite WO 3 /g-C 3 N 4 heterojunction structures. Results showed mesoporous materials with triclinic WO 3 crystal phase, surface areas up to 67.7 m 2 g −1 and band gaps lower than 2.5 eV confirming the absorption to visible light region. The photocatalytic performance of the prepared photocatalysts were assessed towards the oxidation of phenol and reduction of Cr (VI), in single and binary systems using simulated solar light illumination, that followed first-order kinetics. The WO 3 /g-C 3 N 4 composites were found to exhibit improved photocatalytic performances compared to the pure WO 3 and g-C 3 N 4 with 6.5 wt% WO 3 /g-C 3 N 4 and 5 wt% WO 3 /g-C 3 N 4 composites being the most efficient catalysts for the oxidation of phenolics and reduction of Cr (VI), respectively. The improved performance was explained by a Z-scheme photocatalytic mechanism which was proposed based on scavenging experiments and the determination of the corresponding energy levels of valence and conduction bands. The study demonstrated that such composites present interesting photocatalytic properties that can be further expanded to other environmental depollution applications as well as in energy applications.


Introduction
Semiconductor based solar light active photocatalysts have received a great attention in environmental remediation and energy production and conversion applications. Among other semiconductors, tungsten oxide (WO3) is considered as a promising photocatalyst because of its good chemical stability, small band gap (2.4-2.8 eV), non-toxicity and high oxidation ability to degrade persistent organic pollutants. However, the photocatalytic activity of WO3 is low because of the fast recombination of electron-holes pairs [1][2][3][4].
Another promising organic semiconductor, receiving the focus of recent studies on photocatalytic applications, is graphitic carbon nitride (g-C3N4). g-C3N4 with two-dimensional (2D) nanostructure has relatively narrow band gap (2.7 eV), chemical and thermal (up to 600 o C) stability due to s-triazine ring structure, high resistance in basic and acidic solutions and insolubility in common solvents as water and ethanol. Also, it can be fabricated easily with low-cost precursor compounds like melamine, urea etc. and is non-toxic. However, its photocatalytic efficiency is also limited because of the fast recombination of photogenerated electron-holes pairs and low ability to generate • OH radicals [5][6][7].
In modern strategies for photocatalyst-engineered materials, heterostructuring with a second or a third semiconductor represents one the most innovative approaches. Heterostructuring can overdraw usual limitations of single semiconductors such as low exploitation of solar light, fast electron-hole recombination, inappropriate redox potentials for oxidation or reduction reactions. Direct z-scheme photocatalytic systems, mimicking natural photosynthesis, represent the most attractive photocatalytic systems because of their advantages such as fast charge separation, suitable band potentials and usually increase light-response [8,9]. There are few works exploring the WO3/g-C3N4 heterojunction structure via various methods such as hydrothermal treatment [10][11][12][13][14][15], wet mixing method [16][17][18][19][20][21][22], thermal treatment [23,24] for different purposes such as removal of dyes, degradation of pharmaceuticals and hydrogen production etc.
A simple and low-cost method for the preparation of nanofibers is offered by the electrospinning technique using various polymers (over 200 till now) produced in industrial scale. The (1D) nanofibers presented large surface area to volume ratio and they can weaken the recombination of electron-hole (h + -e − ) better than spherical particles, thus, the photocatalytic activity of a material could improve. The electrospun fibers/materials are being used in different applications such as catalysis, tissue engineering, drug delivery etc [25,26]. According to the current bibliography, a study dealing with the fabrication of WO3 fibers/g-C3N4 heterojunction structure combining the electrospinning and wet mixing techniques is lacking. In addition, previously prepared WO3/g-C3N4 composites by other methods have not been applied in simultaneous oxido-reduction processes such as the oxidation of organic pollutants and the reduction of toxic heavy metal cations. As a result, the present work deals with the preparation of g-C3N4/WO3 composite materials with a range of WO3 electrospun fibers loadings (up to 15% wt%), their extended physicochemical characterization by a battery of techniques and the evaluation of their photocatalytic performance towards the simultaneous oxidation-reduction processes of phenol and chromium (VI), respectively. Finally, the photocatalytic mechanism taking place in the process was assessed by scavenging and • OH radical determination, proving a direct Z-scheme system with enhanced performance.
For g-C3N4 two characteristic intense peaks at 13.2° and 27.5° were observed. The peak at 2θ = 13.2° corresponds to the (100) plane of g-C3N4 and it is attributed to the inplane repetitive and continuous heptazine network. The peak located at 2θ = 27.5° corresponds to the (002) plane of g-C3N4 and it is assigned to the stacking of the conjugated aromatic system. [5][6][7]. An interplanar distance of aromatic units is calculated to be 3.24 Å using Bragg's law.
The prepared composite catalysts present closely the same patterns to pristine WO3. The diffraction peak intensity of g-C3N4 becomes increasingly lower with increasing the loading of WO3, because of the relative lowering content of g-C3N4 as well as the poor crystallinity of g-C3N4 compared to that of WO3. In contrast, the diffraction peak intensity of WO3 becomes more evident with increasing the loading of WO3 to g-C3N4. The diffraction patterns of g-C3N4 are closely disappeared at 15%wt WO3 loading, indicating that g-C3N4 has been almost covered by WO3 fibers. The determined crystal sizes of the composite materials (12.4-15.3 nm) according to the Scherrer equation presented low variation in respect to pristine WO3 (12.5 nm) and g-C3N4 (10.8 nm) materials (Table 1). The FT-IR spectra of the synthesized materials are presented in Figure 2. The peaks of g-C3N4 can be observed at 1539, 1455, 1394, 1316, 1232 cm −1 confirming the stretching vibration of C-N(-C)-C or C-NH-C heterocycles. The peak at 804 cm −1 is the most characteristic of heptazine rings. The peaks between 3400 and 3000 cm −1 are related to the stretching vibration of N-H and stretching vibration mode O-H bond. Also, the peak at 1739 cm −1 corresponds to C=N bending vibration and the bending vibration of W-O-W was verified in 500-900 cm −1 . The characteristic WO3 broad peak was more obvious in the materials with the higher WO3 loadings. Finally, the observed FT-IR spectra of prepared catalysts confirm the formation of composite heterostructures. Figure 2. FT-IR spectra of WO3 fibers, g-C3N4 and composite materials of WO3/g-C3N4.

Morphology and UV-Vis Diffuse Reflectance
The structural features of all catalysts were also studied by SEM. The WO3 fibers and the composite materials appeared in Figure 3 and Figure S1. The diameter of fibers varies between 250-300 nm (Figure 3a and Figure S1e). In figure S1(a) it is shown the structure of bulk g-C3N4 consisting of stacking sheets. SEM images of composite materials ( Figure  1b-d) demonstrated aggregated and disordered WO3 fibers on g-C3N4 flakes.  Figure S2. According to the IUPAC classification, the WO3 fibers, g-C3N4 and the composite materials are typical mesoporous materials with type IVa adsorption isotherms and H3, H2 (b) (for WO3 fibers) hysteresis loop type [28]. BET specific surface area for the WO3/g-C3N4 composites ranged between 49.0 m 2 g −1 and 67.7 m 2 g −1 . The calculated specific surface area, pore diameter and total pore volume of the catalysts are shown in Table 1. Pore volume and surface area of the composites were about 5-7-fold higher than those of WO3 and pristine g-C3N4. The reason might be the unordered intercalation of WO3 fibers between aggregated g-C3N4 flakes.  The optical properties of prepared composite catalysts were investigated by UV-Vis diffuse reflectance spectroscopy. Τhe absorption edge of g-C3N4 sample is determined at 494 nm, and the introduction of WO3 fibers in g-C3N4 contributed to the increased absorbance of visible-light and red shift. Figure 5 displays the UV-vis absorption spectra of the materials and the energy band gap (Eg) of each photocatalyst determined using the Kübelka-Munk function ( Figures 5(b) and S3). The Eg's and the absorption edge of the materials are shown in Table 1. The absorption edges of materials varied between 494 -558 nm and the energy band gap 2.22-2.51 eV. Calculated Eg values for g-C3N4 and WO3 agree well with previous reports [14,19,22].
The band edge position of CB and VB of prepared materials was calculated by the following equations: where X is the electronegativity of g-C3N4 (4.67 eV) and WO3 (6.59 eV); Ee (4.5 eV), which is the energy of free electrons on the hydrogen scale; Eg is the energy band gap of the semiconductors and the EVB and ECB are the potentials for the valence and the conduction band, respectively. According of the above formulas, the EVB and ECB of WO3 was calculated at 3.2 eV and 0.89 eV while for g-C3N4 the corresponded values were 1.425 eV and −1.085 eV.

Photocatalytic Activity
The photocatalytic activity of the prepared catalysts was studied against phenol in single solute systems and against phenol -chromium in binary solute systems. Since it is well-known that phenol photocatalytic degradation proceeds through the formation of various phenolics intermediates, phenolics were also determined in single and binary systems. The results for all photocatalytic experiments are shown in Figure 6, Table 2 and  Table 3.

Photocatalytic Mechanism for the WO3/g-C3N4 composite catalysts
In order to investigate the photocatalytic mechanism, a series of photocatalytic experiments in the presence scavengers have been performed. Isopropanol (IPA), triethanolamine (TEOA), sodium azide (NaN3), and superoxide dismutase (SOD) can act as scavengers of • OH, h + , singlet oxygen ( 1 O2) and • OH; O2 •− , respectively. Figure 7 shows the effects of scavengers on the photocatalytic kinetics for 10%WCN catalyst. After adding, IPA, NaN3 and SOD the degradation rate constant for phenol was decreased by 68%, 71% and 57%, respectively. When TEOA was added, the photocatalytic efficiency was significantly reduced by 92%. The degradation kinetics of phenol in the presence of IPA and NaN3 are almost similar, indicating that 1 O2 doesn't display a significant role on the photocatalytic process. In addition, the high inhibition in the presence of SOD denoted also the formation of O2 •− and the effective charge separation with the electrons being captured by the molecular oxygen.

10%WCN
decreased degradation efficiency indicating that the major active species involved the photocatalytic reaction are • OH, h + and O2 •− . If a type II heterojunction (Figure 8) is considered for the composite catalysts, the photogenerated electrons would transport from the CB of g-C3N4 to the CB of WO3 and the holes would form from the VB of WO3 to VB of g-C3N4. However, in this case scenario, the accumulated electrons in the CB of WO3 couldn't reduce O2 to generate • O2 − radicals (WO3 ECB = +0.89 eV; O2/O2 •− (−0.33 eV vs NHE)) and the collected holes couldn't oxidize OHor H2O to form • OH (g-C3N4, EVB = + 1.425 eV; OH − / • OH (+1.99 eV vs NHE); H2O/ • OH (+2.4 eV vs NHE)). As a result, type II photocatalytic mechanism is excluded. On the contrary, the formation and the participation of • OH and O2 •− in the photocatalytic degradation based on the scavenging experiments and fluorescence measurements as well as the determined band edge positions for WO3 and g-C3N4 demonstrate a Z-scheme mechanism (Figure 9).  The gap between the CB g-C3N4 and the CB WO3 is about 1.98 eV, which is four times larger than the gap between the VB of g-C3N4 and the CB of WO3 (0.53 eV), leading the electrons from CB of WO3 to transfer to the VB of g-C3N4. Consequently, the holes and the electrons will gather in the VB of WO3 and the CB of g-C3N4, respectively. The corresponding band potentials matched well with the production of the reactive species determined as well as on the oxidation of phenol and reduction of chromium species. In this way, the heterojunction will exhibit strong oxidize and reduction capability, representing an interesting system for both oxido-reductive applications.

Recyclability of the Composite Catalyst
In order to investigate the recyclability of the synthesized photocatalysts, the photocatalytic activity against phenolics in binary systems using one of the most efficient photocatalyst 6.5%WCN was studied for three consecutive cycles. According to the experimental results ( Figure S4) the apparent reaction constants determined for the first, second and third catalytic cycle were 0.0469 min −1 , 0.0460 min −1 and 0.0433 min −1 for the degradation of phenolics, respectively. A loss of 8% of its catalytic performance was recorded after the third cycle. In combination with SEM ( Figure S5) and FT-IR ( Figure S6) analysis, a high stability of the catalyst can be suggested taking into account also potential losses during the catalyst recovery process.

Materials and Chemicals
Poly

Preparation of electrospun WO3 fibers, g-C3N4and composite materials WO3/g-C3N4
The WO3 fibers were synthesized in a typical horizontal set up of electrospinning apparatus using a grounded collecting metal plate. Firstly, a PVP (0.2 g) solution in methanol (2.5 mL) and a second solution of AMH (0.125 g) in bi-distilled water (1.25 mL) were prepared and stirred by vortex to ensure complete dissolution. Then, AMH solution was transferred to PVP solution and stirred again by vortex until to achieve a homogeneous solution. Afterwards, the final mixture was added into a plastic syringe and placed in the syringe-pump (Holliston, Ma, USA). Electrospinning process was carried out at room temperature, kept between 27-30 °C, with a relative humidity of 32-35% according to previously selected conditions [29]. Moreover, the applied voltage was set at 20kV, the distance between the tip and the collector was 15 cm and the feed rate of solution was 1mLh −1 . The fabricated fibers were calcinated at 500 °C for 3 hours with a heating rate of 2 °C min −1 . The g-C3N4 powder was synthesized using melamine as precursor compound. The melamine was calcinated in air at 550 °C for 4h with a heating rate of 10 °C min −1 in a covered quartz crucible, which was whole wrapped with aluminum foil, then allowed to cool down naturally. The obtained yellow color solid was ground well into a fine powder in an agate mortar. The WO3/g-C3N4 composite catalysts were synthesized by the wet mixing method. Each time, the as-prepared catalyst powders were added in 50 mL of bi-distilled water and the resulting suspension was stirred for 2 hours and heated until dryness. The collected catalysts were calcinated at 520 °C for 2 hours with a heating rate of 5 °C min −1 . The as-prepared catalysts were named as 1%WCN, 5%WCN, 6 .5%WCN, 8%WCN, 10%WCN, 15%WCN, respectively to the %weight of WO3. For comparison purposes, a composite catalyst with WO3 particles and 6.5% weight of WO3 was also synthesized according to the previous methodology and named as 6.5%WCNp.

Characterization
The X-ray diffraction (XRD) patterns of the fabricated catalysts were recorded using a Bruker Advance D8 instrument (Billerica, MA, USA) working with Cu-Ka (λ=1.5406 Å) radiation. Diffractograms were scanned from 2θ 10 o to 70 o in steps of 0.02 o and a rate 0.01 o θ sec −1 . The patterns were assigned to crystal phases with the use of the International Center for Diffraction Data (ICCD). The morphology and the size of fibers were observed by scanning electron microscopy (SEM) using a JEOL JSM 5600 instrument (Tokyo, Japan) working at 20kV.
Nitrogen adsorption-desorption isotherms were measured using a Quantachrome Autosorb-1 instrument (Bounton Beach, FL, USA) at 77K. The prepared catalysts (≈ 80 mg) were degassed at 423 K for 3 h. Brunauer-Emmet-Teller (BET) method was used at relative pressure between 0.05-0.3, in order to calculate the specific surface area (SSA) of each material. Adsorbed amount of nitrogen at relative pressure P/P0 = 0.95 was used in order to calculate the total pore volume (VTOT). The BJH (Barrett, Joyner and Halenda) method was used to determine the pore size distribution (PSD) of the photocatalysts.
A Shimadzu SALD-2300 laser diffraction particle size analyzer (Kyoto, Japan) working with dynamic light scattering (DLS) mode was used for catalysts' hydrodynamic particle size measurements. Suspensions of the catalysts were prepared by stirring for 2.5 h.

Spectroscopy Measurements
Attenuated Total Reflectance-Fourier Transform Infrared spectra (ATR-FT-IR) were obtained by a Shimadzu IR Spirit QATR-S (Kyoto, Japan). The fabricated photocatalysts were scanned in the range 4000-400 min −1 .
Diffuse reflectance spectra (DRS) of the fabricated catalysts were carried out on a Shimadzu 2600 spectrophotometer bearing an IRS-2600 integrating sphere (Kyoto, Japan) in the wavelength range of 200-800 nm at room temperature using BaSO4 (Nacalai Tesque, extra pure reagent, Kyoto, Japan), as a reference sample.

Determination of . OH Radicals by Fluorescence Measurements
The formation of hydroxyl radicals was studied using terephthalic acid (TA) method. An aqueous solution (100 mL) of NaOH (2 × 10 −3 M, 99% Riedel-de Haën, Seelze, Germany) and TA (5 × 10 −4 M) was prepared and then 10 mg of the photocatalyst powder was added and the suspension was placed in the photocatalytic reactor [30,31] following the same irradiation conditions followed in the photocatalytic experiments.

Evaluation of Photocatalytic performance
The photocatalytic activity was evaluated towards the oxidation of phenol and the reduction of chromium. The solar simulator apparatus Suntest XLS+ (Atlas, Germany) disposing a xenon lamp 2.2kW jacked with special 290 nm cut-off glass filter was used for the photocatalytic experiments. The irradiation intensity was maintained at 500 Wm −2 during the experiments. The set-up of the photocatlytic reactor was described in previous studies [29,30]. For the photocatalytic experiments, 100 mL of phenol (10 mgL −1 ) solution were loaded in Pyrex glass reactor thermostated at ambient conditions (≈20 °C), by water circuit flowing in the double-skin of the reactor and air-flow, under continuous stirring. The molar ratio of phenol: Cr (VI) was set to 1:5 while the pH of solutions was adjusted by H2SO4 at pH=2. Before illumination the suspension was magnetically stirred for 30 minutes to ensure the establishment of adsorption-desorption equilibrium onto the catalyst surface. The samples taken at different time intervals were filtered through 0.22 μm PTFE syringe filters. The concentrations of Cr (VI) and phenolics were determined by the diphenyl-carbazide and Folin-Ciocalteu reagent, respectively. The samples were analyzed by UV-Vis-spectroscopy (Jasco-V630, Tokyo, Japan) measuring the absorbance at the characteristic wavelength of 540 nm for Cr (VI) and 765 nm for phenolics. Determination of phenol concentration was carried out by high-performance liquid chromatography (HPLC) (Schimadzu, LC 10AD, Diode Array Detector SPD-M10A, Degasser DGU-14A). The mobile phase was a mixture of HPLC grade water (50%) and MeOH (50%). The column oven (Schimadzu, CTO-10A) was set at 40 °C.

Conclusions
In summary, g-C3N4/WO3 composite catalysts (up to 15% wt of WO3) have been successfully prepared combining the electrospinning process and the wet mixing method. The composite materials presented mainly triclinic crystal phase of WO3, mesoporosity and increased response to visible light irradiation. The materials presented increased efficiency for the both oxidation and reduction processes as determined towards the oxidation of phenol and reduction of Cr (VI) as well as high stability. Photocatalytic efficiencies of materials are better in binary systems than the single systems. A Z-scheme photocatalytic mechanism was proposed for the composites make them promising systems for versatile photocatalytic application in environmental remediation and energy conversion. The ease preparation in large quantities constitute another advantage for their applications.