- Email: [email protected]
Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today journal homepage: www.elsevier.com/locate/cattod
Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes Guilherme Garcia Bessegato, Juliano Carvalho Cardoso, Maria Valnice Boldrin Zanoni ∗ Departamento de Química Analítica, Instituto de Química, Universidade Estadual Paulista - UNESP, Av. Prof. Francisco Degni, 55, C. P. 355, 14801-970 Araraquara, SP, Brazil
a r t i c l e
i n f o
Article history: Received 18 January 2014 Received in revised form 23 March 2014 Accepted 28 March 2014 Available online xxx Keywords: TiO2 nanotube arrays B-doped TiO2 Bandgap engineering Photoelectrochemical degradation Photoelectrochemistry
a b s t r a c t The present work evaluates the performance of boron-doped TiO2 nanotubes (B–TiO2 NTs) prepared by electrochemical anodization in the presence of 70, 140, 280 and 560 ppm of boron when activated by UV/visible irradiation. The presence of boron was characterized by FEG-MEV, XRD, DRS, XPS and photocurrent curves, where the maximum values were obtained for B–TiO2 NTs. Assessment of the electrodes were conducted in the hair dye degradation (Acid Yellow 1 dye) and the degradation rate was 2 times higher at the doped electrode containing 280 ppm of boron, where the maximum electrode activation was obtained under UV/vis irradiation. The best results indicate 100% of discoloration and up to 95% of TOC removal when 100 ppm of Acid Yellow 1 dye was treated at B–TiO2 NTs electrode during 120 min at Eapp = +1.2 V in 0.01 mol L−1 of Na2 SO4 at pH 2. These findings showed that the photoelectrocatalysis is an efficient method to remove hazardous organic compounds from water and B-doped TiO2 electrodes are an important step in the search for efficient and stable catalysts for photo(electro)catalysis. © 2014 Published by Elsevier B.V.
1. Introduction The combination of photocatalysis and electrochemical techniques has been successfully explored in the degradation of hazardous organic compounds since the beginning of the 1990s [1,2], although Fujishima and Honda [3] had already used the photoelectrocatalysis technique for water splitting in 1972. The great goal is that under external positive bias potential there is better charges (e− /h+ ) separation and minimization due to competitive recombination process. Among several semiconductor materials TiO2 has promoted high efficiency when activated by UV irradiation ( ≤ 387 nm). In this condition the photogenerated holes (h+ ) at the photoanode surface can oxidize adsorbed water molecules and/or hydroxyl ions producing hydroxyl radicals (• OH) [4], which are the base to promote efficient degradation of organic compound as a contaminant [4–6]. Nevertheless, the use of sunlight for photoelectrodes activation is limited (because it provides up to 5% of UV light), and in order to extend the absorption onset of TiO2 with light at longer wavelengths (redshift) several authors have adopted the
∗ Corresponding author. Tel.: +55 1633019619. E-mail addresses: [email protected] (G.G. Bessegato), [email protected] (M.V.B. Zanoni).
insertion of cation or anion into the TiO2 lattice [7], such as N-doped TiO2 [8,9], C [10], S [11], B [12], Cr [13], Fe [14] and W [15]. Among them, boron-doped TiO2 has gained attention since boron atoms can be introduced in the TiO2 lattice and the p orbital of B is mixed with O 2p orbitals, leading to a shift in the optical response to the visible range [12,16]. Therefore, it is a promising path towards photocatalysis under visible light. B-doped TiO2 NTs were obtained by chemical vapor deposition treatment [12,17], electrodeposition method [18] and in a one-step process during the electrochemical anodization [19] with addition of boric acid in the electrolyte of growth of nanotubes. In addition, the doping process coupled to a nanostructured material has improved the interest of this material in photoelectrocatalysis, due to improvement in the active surface area (reaction/interaction can be facilitated between the catalyst and the interacting media) and excellent electric properties, once the charges carriers transfer is mainly governed by the quantum confinement phenomenon [20]. For this purpose TiO2 nanotube arrays (TiO2 NTs) have shown high structural organization and excellent electron percolation based on vectorial charge transfer between interfaces [6,21,22], which has been applied in several wastewater treatment. The present work investigates the fabrication and characterization of boron-doped TiO2 nanotubes (B–TiO2 NTs) by electrochemical anodization (one-step doping) using NaBF4 as a
http://dx.doi.org/10.1016/j.cattod.2014.03.073 0920-5861/© 2014 Published by Elsevier B.V.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
2
boron source and the application in the wastewater treatment containing acid dyes. For this, our main objective was to investigate the effect of boron doping in the electronic and optical properties when TiO2 is activated by using a simple commercial lamp (125 W Hg high pressure lamp) as a source of irradiation of UV and visible region. The electrodes were tested in the photoelectrochemical degradation of hazardous dye AY1 (also called Naphthol Yellow S) [23] extensively used in oxidative and semipermanent hair colorants. The dye degradation was optimized and performed by direct photolysis, electrocatalysis, photocatalysis and photoelectrocatalysis. Then, the degradation was evaluated by UV/vis spectrophotometry, total organic carbon (TOC) removal and liquid chromatography coupled to mass spectrometry (LC-MS/MS) with the aim to assess low levels of dye remaining. 2. Material and methods 2.1. Preparation of TiO2 NTs and boron-doped TiO2 NTs Titanium samples (Realum, Brazil) with 5 cm × 5 cm were polished using silicon carbide sandpaper of successively finer roughness (220, 320, 400, 800, 1200, and 1500 grit). Then, the sheets were degreased by successively sonication for 15 min in isopropanol, acetone and ultrapure water and dried in a N2 stream. TiO2 NTs were produced by electrochemical anodization using as electrolyte 1.0 mol L−1 (NH4 )SO4 + 0.5 wt.% NH4 F (Sigma–Aldrich). The applied potential was 20 V for 2 h (room temperature) using a DC power supply (Minipa MPL-1303) in a two-electrode electrochemical cell using a Ti/Ru sheet (25 cm2 ) as a counter electrode. The boron-doped TiO2 NTs electrodes were grown using the same conditions, adding NaBF4 in the electrolyte as a boron source, in the concentrations of 70, 140, 280 and 560 mg L−1 . Then, the samples were rinsed with DI water, dried in a nitrogen stream and annealed at 450 ◦ C for 2 h in air, with heating rates of 2 ◦ C min−1 [24]. 2.2. Characterization of materials The prepared TiO2 NTs and B-doped TiO2 NTs samples were characterized by X-ray diffraction (XRD) on a SIEMENS D5000, DIFFRAC PLUS XRD Commander X-ray diffractometer with Cu K␣ radiation. The morphological characterization was carried out by Field Emission Gun-Scanning Electron Microscopy (FEG-SEM) on a JEOL 7500F Microscope. The chemical composition was examined using X-ray photoelectron spectroscopy (XPS, UNI-SPECS UHV). The optical properties were evaluated by diffuse reflectance spectroscopy (DRS), using an UV/vis/NIR spectrometer (PerkinElmer Lambda 1050) with an Integrating Sphere-150 mm UV/vis/NIR (InGaAs) Module. The equipment was calibrated with a Spectralon standard (Labsphere USRS-99-020, 99% reflectance) and the reflectance was measured in the 250–800 nm range. The electronic properties were evaluated by photocurrent curves registered by linear sweep voltammetry technique at = 10 mV s−1 in 0.1 mol L−1 Na2 SO4 (Sigma–Aldrich) using a potentiostat/galvanostat Autolab PGSTAT302N. 2.3. Degradation of the dye Acid Yellow 1 The degradations were performed in a 500 mL cylindrical glass reactor at a controlled temperature of 25 ◦ C by a thermostatic bath (Quimis, Brazil), as shown in Fig. 1. The anode was irradiated by a commercial Hg high pressure lamp of 125 W (Osram) (without the bulb) vertically inserted in a central quartz tube. The photoelectrochemical reactor consisted of a conventional threeelectrode electrochemical cell where a Ti/Ru plate and a saturated
Fig. 1. Scheme of the photoelectrochemical reactor.
Ag/AgCl were used as counter and reference electrodes, respectively. The TiO2 NTs or the B–TiO2 NTs electrode were used as working electrodes. The electrolyte was Na2 SO4 and all the photoelectrochemical experiments were carried out in a potentiostat/galvanostat Autolab PGSTAT302N. 2.4. Evaluation of degradation In the course of the Acid Yellow 1 degradation, aliquots were taken at controlled times and discoloration was monitored by UV/vis spectrophotometry (Hewlett-Packard 8453 spectrophotometer) and the mineralization of organic matter with a total organic carbon analyzer (TOC-VCPN , Shimadzu). 3. Results and discussion 3.1. Characterization of doped and bare TiO2 NT electrodes Images of FEG-MEV (Fig. 2) show the formation of self-aligned and self-organized nanotubes perpendicularly to the metallic substrate when Ti foil is submitted to 20 V for 2 h in 1.0 mol L−1 (NH4 )SO4 + 0.5 wt.% NH4 F, with an average diameter of 110 nm and a tube wall of 15 nm. Fig. 2B shows a cross sectional image of TiO2 nanotubes, presenting 800 nm in length. The introduction of boron atoms in the TiO2 lattice did not cause changes in the morphology of the nanotubes, as can be seen in Fig. 2C–F, for all the boron-doped samples. However, the XPS analysis confirmed the presence of boron in the TiO2 NTs lattice, as can be observed in the B 1s spectra in Fig. 3A. For B–TiO2 samples, the B 1s peak appeared at a binding energy of 191.6 eV. According to earlier reports of Zaleska et al. [25] and Gopal et al. [26] this position is consistent with substitutional boron at Ti sites (Ti–O–B–O–Ti). In the Ti 2p spectrum (not shown) the symmetrical 2p3/2 peak is located at 459.0 eV for the doped and undoped samples. As this binding energy is characteristic for pristine TiO2 phase, no influence of B doping on the peak position was detected, excluding a direct Ti–B interaction. In the case of boron atoms inserted in interstitial positions (also compatible with a binding energy of 191.6 eV) the presence of the Ti3+ phase at about 456 eV is expected [27], which was not detected in the Ti 2p spectra. Fig. 3A shows that the intensity of the B 1s peak increases when the concentration of NaBF4 in the anodization electrolyte increases. The quantitative XPS analysis showed an increase of the boron content from 0.4 to 0.7 at.% when changing the
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
3
Fig. 2. FEG-SEM images of (A) top view of TiO2 NTs; (B) cross section of TiO2 NTs; and top view of boron-doped TiO2 NTs with addition of: (C) 70 ppm, (D) 140 ppm, (E) 280 ppm and (F) 560 ppm of NaBF4 in the electrolyte of nanotubes growth.
Fig. 3. (A) B 1s XPS spectrum of boron-doped TiO2 NTs with (1) 280 ppm and (2) 560 ppm of NaBF4 added in the electrolyte; and (B) X-ray diffraction patterns of (1) not annealed TiO2 NT sample, (2) TiO2 NTs and boron-doped TiO2 NTs with addition of (3) 70 ppm, (D) 140 ppm, (E) 280 ppm and (F) 560 ppm of NaBF4 in the electrolyte of nanotubes growth.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7 4
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 4. (A) DRS spectrum to: (1) bare TiO2 NTs; and B-doped TiO2 NTs sample with addition of (2) 70 ppm, (3) 560 ppm, (4) 140 ppm and (5) 280 ppm of NaBF4 in the electrolyte of nanotubes growth. (B) Photocurrent versus potential curves to (1) dark current; (2) TiO2 NTs before annealing, (3) bare TiO2 NTs; and doped samples with addition of (4) 70 ppm, (5) 140 ppm, (6) 560 ppm and (7) 280 ppm of NaBF4 in the electrolyte of nanotubes growth. Linear scan voltammetry, = 10 mV s−1 , in 0.1 mol L−1 Na2 SO4 using UV/vis irradiation of a 125 W Hg mercury lamp. The inset in (A) shows the Tauc’s Plot and band gap of (1) B280ppm –TiO2 NTs and (2) undoped TiO2 NTs.
concentration of NaBF4 in the electrolyte form 280 and 560 ppm. For lower NaBF4 concentrations the boron content was below the detection limit. By X-ray diffraction (XRD) the peaks indicate the formation of anatase phase after annealing the TiO2 NTs (Fig. 3B), and no difference between doped and undoped samples can be observed. The optical properties of the TiO2 NTs and B–TiO2 samples were studied by diffuse reflectance spectroscopy (DRS), which the spectra are shown in Fig. 4A. All samples obtained for TiO2 nanotube grown in different concentrations of NaBF4 added in the electrolyte showed identical absorption spectra. Doped materials presented a new optical absorption shoulder at wavelengths greater than 400 nm, showing that the substitutional boron atoms in the TiO2 lattice may change the optical properties of the material, extending the optical absorption spectrum into the visible region. The inset in Fig. 4A shows the Tauc’s graph [28,29] used to estimate the band gap energy of the materials. This method consists of extrapolating 1/ the linear portion of a plot of ˛(h) as a function of h (eV), where the intercept at ˛ = 0 is the optical band gap (Ebg in eV). In these expressions ˛ is the absorption coefficient, h is Planck’s constant (J s), is the frequency (s−1 ) and ␥ is the power coefficient that can take the values 1/2, 3/2, 2 or 3 depending on the type of electronic transition: direct allowed, direct forbidden, indirect allowed and indirect forbidden, respectively. More accurate results for TiO2 are obtained with the use of = 2 [29]. The diffuse reflectance measurements can be converted to an equivalent absorption coefficient with Kubelka–Munk equation (K–M ≈ ˛ = (1 − R)2 /2R). Based on the described method, the band gap for TiO2 NTs was 3.34 eV and for the boron-doped TiO2 NTs the obtained values were 3.27 and 2.22 eV. There are two intercepts for B-doped TiO2 samples probably due to the formation of intermediary energy levels (created by substituted B atoms) between the valence and conduction band of the TiO2 [30], in agreement with results of Li et al. [18]. There is a debate in the literature about the likely mechanism of decreasing the band gap of titanium dioxide in second-generation photocatalysts (cation or anion doping) [7]. Some papers have reported that anion doping (as B-doping) can introduce localized dopant states in the band gap [7,18,31], while some studies reported that intrinsic defects, as oxygen vacancies, could enhance visible light photoactivity of doped TiO2 through the creation of color centers, including Ti3+ centers [7]. Another situation is the true narrowing of the original TiO2 band gap, which needs heavy anion doping that may compromise its original integrity [7]. Thus, the new absorption shoulder at the visible region in our material can be assigned to excitation of electrons from localized dopant energy levels above the valence band of TiO2 (created by substitutional boron doping) to the conduction band of TiO2 [18].
Fig. 5 shows a schematic representation of a probable mechanism of charges generation and electron flow in a photoelectrochemical system, to B-doped samples. The photoactivity of the doped (Fig. 4B, Curves 4–7) and bare materials (Fig. 4B, Curve 3) were evaluated by comparing photocurrent-potential curves under dark conditions (Fig. 4B, Curve 1) and under UV/vis irradiation both in 0.1 mol L−1 Na2 SO4 pH 6. By illuminating the semiconductor with light energy greater than that of the band gap, electron-hole pairs are generated at the electrode surface. Under UV light, by applying a bias potential higher than the flat band potential of TiO2 , there is a bending of the CB and VB bands causing a more effective charges separation that increases the photocurrent [32]. The B-doped samples (Curves 4–7) showed larger photocurrent than the undoped one (Curve 3). The photocurrent taken at +1.5 V indicates an increase of photoresponse when the boron concentration is increased from 70 to 280 ppm. This is due to the fact that the B-doped samples present a higher utilization of the radiation emitted by the lamp in the UV and visible regions, as already evidenced by Fig. 4A. But, at higher boron concentrations (560 ppm, Curve 6), there was a decrease in the photocurrent, where the boron is not acting as a dopant anymore, but it is trapping the electron generated in the system. Thus, as a better activity is obtained for B–TiO2 NTs when doped with 280 ppm this condition was chosen for further measurements. Curve 2 of Fig. 4B
Fig. 5. Proposal of a photoelectrocatalytic mechanism in the boron-doped TiO2 NTs samples.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
5
Fig. 6. Optimization of different conditions for treatment of 50 ppm of AY1 in 0.1 M Na2 SO4 , where: (A) effect of the amount of boron added to the electrolyte of nanotube growth (pH 6 under Eapp = +1.2 V); (B) effect of annealing (B280ppm –TiO2 NTs, pH 6 under Eapp = +1.2 V); (C) effect of pH (B280ppm –TiO2 NTs, Eapp = +1.2 V); (D) effect of applied potential (B280ppm –TiO2 NTs, pH 2).
shows the photocurrent of B–TiO2 before annealing, proving that the amorphous phase is not photoactive, presenting a photocurrent negligible as the dark current. Fig. 5 illustrates a possible mechanism of charge generation in a photoelectrochemical system using B–TiO2 NTs. The application of a higher potential than the flat band potential drives the electrons to the counter electrode and holes are more available to promote the oxidation reactions. The innovation is, the use of B–TiO2 NTs anode promotes a better use of light, since the material is activated by both UV light and visible light radiated by the commercial lamp. Therefore, more holes (h+ ) are concentrated on the electrode surface, producing a higher oxidant agent (hydroxyl radicals), which could improve the degradation of organic compounds in solution. In order to evaluate the performance of the B–TiO2 NTs electrodes in photoelectrocatalytic oxidation, the Acid Yellow 1 dye, strongly used in hair dye formulations, was chosen as a target compound for degradation.
3.2. Degradation of AY1 cosmetic dye 3.2.1. Optimization of photoelectrocatalysis conditions The influence of boron doping in the TiO2 NTs electrode was evaluated comparing the performance of B–TiO2 NTs doped with 70, 140, 280 and 560 ppm of NaBF4 . For this, the photoelectrocatalytic degradation of the 50 ppm of AY1 was monitored in 1.0 M Na2 SO4 , pH 6 under bias potential of +1.2 V and UV/vis irradiation for comparison the doped and the bare photoanodes. There was 100% of discoloration in all tested electrodes, but the performance is different for total organic carbon (TOC) removal as shown Fig. 6A. Although a maximum of 70% of TOC removal is obtained for bare TiO2 NTs electrode, the results are lower than the borondoped electrodes that show a maximum of 85% of TOC removal. Taking into account the performance during all the treatment, the TiO2 NT electrode prepared with 280 ppm of boron (0.4 at.%) in the anodization was chosen for further experiments.
Fig. 7. (A) Percentage of AY1 degradation monitored at 393 nm; and (B) Percentage of TOC removal of AY1 in the degradation by the different techniques. All used conditions were 100 ppm of AY1 in 0.01 M Na2 SO4 , pH 2, Hg 125 W lamp and Eapp = +1.2 V.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model
ARTICLE IN PRESS
CATTOD-9023; No. of Pages 7
G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
6
Table 1 Discoloration rate constant (k) obtained for ln[A/A0 ] versus time plot during degradation of 100 mg L−1 AY1 dye in 0.01 mol L−1 Na2 SO4 at pH 2; using UV/vis irradiation from Hg 125 W lamp and Eapp = +1.2 V in PEC experiments. Treatment
Electrode
k (min−1 )
Direct photolysis (DP)
–
−0.0083
Photocatalysis (PC)
TiO2 NTs B–TiO2 NTs
−0.0102 −0.0109
Photoelectrocatalysis (PEC)
TiO2 NTs B–TiO2 NTs
−0.0199 −0.0235
Fig. 6B shows the efficiency of TOC removal with not annealed TiO2 and B–TiO2 electrodes in comparison with the annealed ones. There was about 35% of TOC removal at the end of 120 min, due to degradation by direct photolysis. These data confirm what has already been observed by photocurrent curves (Fig. 4B, Curve 2) where the photocurrent generated was negligible, next to the current obtained in the dark. The effect of original pH (2, 4.5 and 6) in the TOC removal efficiency for B–TiO2 NTs electrode was tested in the photoelectrocatalytic degradation of 50 ppm of AY1 in 0.1 M Na2 SO4 , under bias potential of +1.2 V and UV/vis irradiation (Fig. 6C). Higher degradation efficiency (85%) was reached after 60 min of treatment in pH 2. This illustrates that the acid dye can be strongly adsorbed on the electrode surface at very acid conditions, when the pHpzc for TiO2 is around 5.3 [33] and the dye is in anionic form (sulfonate dye) [34]. Thus, pH 2 was chosen for further experiments. In order to check the influence of applied potential, further experiments were carried out comparing the performance at +0.50 V, +1.2 V and +1.5 V (Fig. 6D). The results indicate that the TOC removal is improved at Eapp = +1.2 V reaching 93% of mineralization. All the applied potential is higher than the flat band potential so the electrons are depleted and holes enriched at the surface of B–TiO2 NTs. At a higher potential than +1.2 V, such as +1.5 V it could happen the formation of a competitive reaction such as oxygen evolution that decreases the photoelectrocatalytic performance [5,35]. Thus, considering the TOC removal, a potential of +1.2 V was used. Taking into account that in a photoelectrochemical system the electrolytes interfere drastically in the band bending of the electrode/solution interface, was investigated the effect of electrolyte composition, testing Na2 SO4 concentration of 0.1 mol L−1 , 0.05 mol L−1 and 0.01 mol L−1 . The results do not show any important difference in the TOC removal showing that the ideal conductivity can be reached at sulfate concentration of 0.01 mol L−1 and this condition was adopted in further experiments [5]. Comparison of B-doped TiO2 NT and TiO2 NT electrodes on AY1 degradation under optimized conditions. Using the best experimental conditions defined previously, it was evaluated the photoelectrocatalytic (PEC) performance for the treatment of 100 ppm of AY1 dye in solution of 0.01 M Na2 SO4 at pH 2 under application of +1.2 V when compared to electrocatalysis (EC; Eapp = +1.2 V + catalyst); direct photolysis (DP; UV/vis commercial irradiation); and photocatalysis (PC; UV/vis irradiation + catalyst). The results are shown in Fig. 7 comparing discoloration (Fig. 7A) and TOC removal (Fig. 7B) during 120 min of treatment. Fig. 7A shows the percentage of AY1 degradation monitored at 393 nm by UV/vis spectrophotometry. Except for the electrocatalysis when the catalyst is not active since there is no irradiation, all the other processes promoted 100% discoloration. The degradation kinetics follows a pseudo-first-order reaction (Langmuir–Hinshelwood kinetics) and the obtained constant rates of discoloration at B-doped and undoped electrodes are shown in Table 1. The results indicate that PEC degradation rate is about 2.1
Fig. 8. (A) Photocurrent density curves to: (1) dark current, (2) TiO2 NTs in glass tube, (3) TiO2 NTs in quartz tube, (4) B–TiO2 NTs in glass tube and (5) B–TiO2 NTs in quartz tube; (B) percentage of AY1degradation monitored at 393 nm; and (C) percentage of TOC removal of AY1 by the different techniques using glass tube (g) and quartz tube (q). All used conditions were 100 ppm of AY1 in 0.01 M Na2 SO4 , pH 2 and Eapp = +1.2 V.
times faster than the PC degradation showing the important effect of the bias potential. It is clear in Fig. 7B that the PEC degradation is much more effective in the AY1 mineralization than EC, PC and DP indicating that the separation of charges promoted by PEC treatment improves the degradation efficiency. In addition, the doped electrodes increase the degradation performance (75% of TOC removal in 90 min). This is due to the fact that the doped electrode can absorb at higher wavelengths range emitted by the radiation source. Thus, the higher amount of photogenerated holes can produce more hydroxyl radicals, promoting an increased mineralization rate. Regarding the comparison between the PC and the PEC processes, it is clear that the mineralization percentage of PEC is higher than the PC, around 40% superior at 90 min of degradation. These results prove that photoelectrocatalysis allows more effective
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
separation of photogenerated charges thus increasing the lifetime of e− /h+ pairs [4,36]. The DP can remove only 34% of total organic carbon, and in the EC the TOC removal is negligible. In order to evaluate the visible activity of B-doped TiO2 NTs produced, we change the quartz tube which the lamp is inserted in the photoelectrochemical reactor, for a glass tube that can filter large part of the UV radiation. Fig. 8A shows the photocurrent curves taken with the lamp inserted in a glass tube (g) and in a quartz tube (q) for doped and undoped electrodes. It can be seen that there was a greater percentage reduction of the photocurrent to the undoped electrode than for B-doped one, when the tube was changed. Fig. 8B and C shows the graphs of AY1 degradation and TOC removal for a comparison between degradation efficiency using a glass or quartz tube. In Fig. 8B, the PEC AY1 degradation using quartz tube was very similar. But when using a glass tube, the PEC degradation rate using B-doped electrode was much higher than the PEC using bare TiO2 NTs. In Fig. 8C the B-doped electrode was almost 2 times more efficient than the undoped electrode in TOC removal at the end of 120 min, while using the quartz tube the difference between the two electrodes was smaller. 4. Conclusions Our findings show that boron-doped TiO2 nanotubes can be easily obtained by a one step electrochemical process, where boron atoms are introduced into TiO2 lattice during growth of the nanotubes by electrochemical anodization. The obtained material showed absorption at the visible region of spectrum (presenting a shoulder with band gap of 2.22 eV) and higher photoactivity than the bare TiO2 nanotubes, due to the insertion of boron atoms into the TiO2 lattice. The photoelectrocatalytic treatment applied in the degradation of AY1 by using B–TiO2 NTs was more efficient than the one on the bare electrode due to a better use of the irradiation emitted in the UV and visible region by using a commercial lamp. The effect of the applied potential improved the degradation rate around 2 times when the PEC is compared with a PC technique. Thus, the material consists of an important step in the search for efficient and stable catalysts for application in photo(electro)catalysis, even being activated by solar radiation. Acknowledgments The authors acknowledge the financial support provided by the Brazilian funding agency FAPESP (process 2012/183570; 2011/23128-7 and 2008/10449-7). FEG-SEM facilities were
7
provided by LMA-IQ and X-Ray Diffraction measurements by GFQM-IQ. We are grateful for the contributions of Prof. Dr. Peter Hammer on acquisitions and discussion of XPS data. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]
[31] [32] [33] [34]
[35] [36]
K. Vinodgopal, S. Hotchandani, P.V. Kamat, J. Phys. Chem. 97 (1993) 9040–9044. D.H. Kim, M.A. Anderson, Environ. Sci. Technol. 28 (1994) 479–483. A. Fujishima, K. Honda, Nature 238 (1972) 37–38. J. Georgieva, E. Valova, S. Armyanov, N. Philippidis, I. Poulios, S. Sotiropoulos, J. Hazard. Mater. 211 (2012) 30–46. R. Daghrir, P. Drogui, D. Robert, J. Photochem. Photobiol. A 238 (2012) 41–52. J.C. Cardoso, T.M. Lizier, M.V. Boldrin Zanoni, Appl. Catal. B 99 (2010) 96–102. N. Serpone, J. Phys. Chem. B 110 (2006) 24287–24293. R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293 (2001) 269–271. A. Ghicov, J.M. Macak, H. Tsuchiya, J. Kunze, V. Haeublein, L. Frey, P. Schmuki, Nano Lett. 6 (2006) 1080–1082. J.H. Park, S. Kim, A.J. Bard, Nano Lett. 6 (2006) 24–28. G. Yan, M. Zhang, J. Hou, J. Yang, Mater. Chem. Phys. 129 (2011) 553–557. Y. Su, S. Han, X. Zhang, X. Chen, L. Lei, Mater. Chem. Phys. 110 (2008) 239–246. A. Ghicov, B. Schmidt, J. Kunze, P. Schmuki, Chem. Phys. Lett. 433 (2007) 323–326. Z. Xu, J. Yu, Nanoscale 3 (2011) 3138–3144. J. Gong, W. Pu, C. Yang, J. Zhang, Catal. Commun. 36 (2013) 89–93. C. Di Valentin, G. Pacchioni, Catal. Today 206 (2013) 12–18. N. Lu, X. Quan, J. Li, S. Chen, H. Yu, G. Chen, J. Phys. Chem. C 111 (2007) 11836–11842. J. Li, N. Lu, X. Quan, S. Chen, H. Zhao, Ind. Eng. Chem. Res. 47 (2008) 3804–3808. A. Subramanian, H.-W. Wang, Appl. Surf. Sci. 258 (2012) 6479–6484. A.E.R. Mohamed, S. Rohani, Energy Environ. Sci. 4 (2011) 1065–1086. K. Shankar, J.I. Basham, N.K. Allam, O.K. Varghese, G.K. Mor, X. Feng, M. Paulose, J.A. Seabold, K.-S. Choi, C.A. Grimes, J. Phys. Chem. C 113 (2009) 6327–6359. C.A. Grimes, G.K. Mor, TiO2 Nanotube Arrays: Synthesis, Properties, and Applications, Springer, New York, 2009. R. Jain, V.K. Gupta, S. Sikarwar, J. Hazard. Mater. 182 (2010) 749–756. G.G. Bessegato, J.C. Cardoso, B.F. Silva, M.V.B. Zanoni, J. Photochem. Photobiol. A 276 (2013) 96–103. A. Zaleska, J.W. Sobczak, E. Grabowska, J. Hupka, Appl. Catal. B 78 (2008) 92–100. N.O. Gopal, H.-H. Lo, S.-C. Ke, J. Am. Chem. Soc. 130 (2008) 2760–2761. G. Liu, H.G. Yang, X. Wang, L. Cheng, H. Lu, L. Wang, G.Q. Lu, H.-M. Cheng, J. Phys. Chem. C 113 (2009) 21784–21788. J. Tauc, R. Grigorovici, A. Vancu, Phys. Status Solidi 15 (1966) 627–637. R. Lopez, R. Gomez, J. Sol-Gel Sci. Technol. 61 (2012) 1–7. M. Pelaez, N.T. Nolan, S.C. Pillai, M.K. Seery, P. Falaras, A.G. Kontos, P.S.M. Dunlop, J.W.J. Hamilton, J.A. Byrne, K. O’Shea, M.H. Entezari, D.D. Dionysiou, Appl. Catal. B 125 (2012) 331–349. L. Deng, Y. Chen, M. Yao, S. Wang, B. Zhu, W. Huang, S. Zhang, J. Sol-Gel Sci. Technol. 53 (2010) 535–541. H.O. Finklea, Semiconductor Electrodes, Elsevier, New York, 1988. S. Malato, P. Fernandez-Ibanez, M.I. Maldonado, J. Blanco, W. Gernjak, Catal. Today 147 (2009) 1–59. K. Rajeshwar, M.E. Osugi, W. Chanmanee, C.R. Chenthamarakshan, M.V.B. Zanoni, P. Kajitvichyanukul, R. Krishnan-Ayer, J. Photochem. Photobiol. C 9 (2008) 171–192. Y. Zhang, X. Xiong, Y. Han, X. Zhang, F. Shen, S. Deng, H. Xiao, X. Yang, G. Yang, H. Peng, Chemosphere 88 (2012) 145–154. M.V.B. Zanoni, J.J. Sene, M.A. Anderson, J. Photochem. Photobiol. A 157 (2003) 55–63.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
ARTICLE IN PRESS Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today journal homepage: www.elsevier.com/locate/cattod
Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes Guilherme Garcia Bessegato, Juliano Carvalho Cardoso, Maria Valnice Boldrin Zanoni ∗ Departamento de Química Analítica, Instituto de Química, Universidade Estadual Paulista - UNESP, Av. Prof. Francisco Degni, 55, C. P. 355, 14801-970 Araraquara, SP, Brazil
a r t i c l e
i n f o
Article history: Received 18 January 2014 Received in revised form 23 March 2014 Accepted 28 March 2014 Available online xxx Keywords: TiO2 nanotube arrays B-doped TiO2 Bandgap engineering Photoelectrochemical degradation Photoelectrochemistry
a b s t r a c t The present work evaluates the performance of boron-doped TiO2 nanotubes (B–TiO2 NTs) prepared by electrochemical anodization in the presence of 70, 140, 280 and 560 ppm of boron when activated by UV/visible irradiation. The presence of boron was characterized by FEG-MEV, XRD, DRS, XPS and photocurrent curves, where the maximum values were obtained for B–TiO2 NTs. Assessment of the electrodes were conducted in the hair dye degradation (Acid Yellow 1 dye) and the degradation rate was 2 times higher at the doped electrode containing 280 ppm of boron, where the maximum electrode activation was obtained under UV/vis irradiation. The best results indicate 100% of discoloration and up to 95% of TOC removal when 100 ppm of Acid Yellow 1 dye was treated at B–TiO2 NTs electrode during 120 min at Eapp = +1.2 V in 0.01 mol L−1 of Na2 SO4 at pH 2. These findings showed that the photoelectrocatalysis is an efficient method to remove hazardous organic compounds from water and B-doped TiO2 electrodes are an important step in the search for efficient and stable catalysts for photo(electro)catalysis. © 2014 Published by Elsevier B.V.
1. Introduction The combination of photocatalysis and electrochemical techniques has been successfully explored in the degradation of hazardous organic compounds since the beginning of the 1990s [1,2], although Fujishima and Honda [3] had already used the photoelectrocatalysis technique for water splitting in 1972. The great goal is that under external positive bias potential there is better charges (e− /h+ ) separation and minimization due to competitive recombination process. Among several semiconductor materials TiO2 has promoted high efficiency when activated by UV irradiation ( ≤ 387 nm). In this condition the photogenerated holes (h+ ) at the photoanode surface can oxidize adsorbed water molecules and/or hydroxyl ions producing hydroxyl radicals (• OH) [4], which are the base to promote efficient degradation of organic compound as a contaminant [4–6]. Nevertheless, the use of sunlight for photoelectrodes activation is limited (because it provides up to 5% of UV light), and in order to extend the absorption onset of TiO2 with light at longer wavelengths (redshift) several authors have adopted the
∗ Corresponding author. Tel.: +55 1633019619. E-mail addresses: [email protected] (G.G. Bessegato), [email protected] (M.V.B. Zanoni).
insertion of cation or anion into the TiO2 lattice [7], such as N-doped TiO2 [8,9], C [10], S [11], B [12], Cr [13], Fe [14] and W [15]. Among them, boron-doped TiO2 has gained attention since boron atoms can be introduced in the TiO2 lattice and the p orbital of B is mixed with O 2p orbitals, leading to a shift in the optical response to the visible range [12,16]. Therefore, it is a promising path towards photocatalysis under visible light. B-doped TiO2 NTs were obtained by chemical vapor deposition treatment [12,17], electrodeposition method [18] and in a one-step process during the electrochemical anodization [19] with addition of boric acid in the electrolyte of growth of nanotubes. In addition, the doping process coupled to a nanostructured material has improved the interest of this material in photoelectrocatalysis, due to improvement in the active surface area (reaction/interaction can be facilitated between the catalyst and the interacting media) and excellent electric properties, once the charges carriers transfer is mainly governed by the quantum confinement phenomenon [20]. For this purpose TiO2 nanotube arrays (TiO2 NTs) have shown high structural organization and excellent electron percolation based on vectorial charge transfer between interfaces [6,21,22], which has been applied in several wastewater treatment. The present work investigates the fabrication and characterization of boron-doped TiO2 nanotubes (B–TiO2 NTs) by electrochemical anodization (one-step doping) using NaBF4 as a
http://dx.doi.org/10.1016/j.cattod.2014.03.073 0920-5861/© 2014 Published by Elsevier B.V.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
2
boron source and the application in the wastewater treatment containing acid dyes. For this, our main objective was to investigate the effect of boron doping in the electronic and optical properties when TiO2 is activated by using a simple commercial lamp (125 W Hg high pressure lamp) as a source of irradiation of UV and visible region. The electrodes were tested in the photoelectrochemical degradation of hazardous dye AY1 (also called Naphthol Yellow S) [23] extensively used in oxidative and semipermanent hair colorants. The dye degradation was optimized and performed by direct photolysis, electrocatalysis, photocatalysis and photoelectrocatalysis. Then, the degradation was evaluated by UV/vis spectrophotometry, total organic carbon (TOC) removal and liquid chromatography coupled to mass spectrometry (LC-MS/MS) with the aim to assess low levels of dye remaining. 2. Material and methods 2.1. Preparation of TiO2 NTs and boron-doped TiO2 NTs Titanium samples (Realum, Brazil) with 5 cm × 5 cm were polished using silicon carbide sandpaper of successively finer roughness (220, 320, 400, 800, 1200, and 1500 grit). Then, the sheets were degreased by successively sonication for 15 min in isopropanol, acetone and ultrapure water and dried in a N2 stream. TiO2 NTs were produced by electrochemical anodization using as electrolyte 1.0 mol L−1 (NH4 )SO4 + 0.5 wt.% NH4 F (Sigma–Aldrich). The applied potential was 20 V for 2 h (room temperature) using a DC power supply (Minipa MPL-1303) in a two-electrode electrochemical cell using a Ti/Ru sheet (25 cm2 ) as a counter electrode. The boron-doped TiO2 NTs electrodes were grown using the same conditions, adding NaBF4 in the electrolyte as a boron source, in the concentrations of 70, 140, 280 and 560 mg L−1 . Then, the samples were rinsed with DI water, dried in a nitrogen stream and annealed at 450 ◦ C for 2 h in air, with heating rates of 2 ◦ C min−1 [24]. 2.2. Characterization of materials The prepared TiO2 NTs and B-doped TiO2 NTs samples were characterized by X-ray diffraction (XRD) on a SIEMENS D5000, DIFFRAC PLUS XRD Commander X-ray diffractometer with Cu K␣ radiation. The morphological characterization was carried out by Field Emission Gun-Scanning Electron Microscopy (FEG-SEM) on a JEOL 7500F Microscope. The chemical composition was examined using X-ray photoelectron spectroscopy (XPS, UNI-SPECS UHV). The optical properties were evaluated by diffuse reflectance spectroscopy (DRS), using an UV/vis/NIR spectrometer (PerkinElmer Lambda 1050) with an Integrating Sphere-150 mm UV/vis/NIR (InGaAs) Module. The equipment was calibrated with a Spectralon standard (Labsphere USRS-99-020, 99% reflectance) and the reflectance was measured in the 250–800 nm range. The electronic properties were evaluated by photocurrent curves registered by linear sweep voltammetry technique at = 10 mV s−1 in 0.1 mol L−1 Na2 SO4 (Sigma–Aldrich) using a potentiostat/galvanostat Autolab PGSTAT302N. 2.3. Degradation of the dye Acid Yellow 1 The degradations were performed in a 500 mL cylindrical glass reactor at a controlled temperature of 25 ◦ C by a thermostatic bath (Quimis, Brazil), as shown in Fig. 1. The anode was irradiated by a commercial Hg high pressure lamp of 125 W (Osram) (without the bulb) vertically inserted in a central quartz tube. The photoelectrochemical reactor consisted of a conventional threeelectrode electrochemical cell where a Ti/Ru plate and a saturated
Fig. 1. Scheme of the photoelectrochemical reactor.
Ag/AgCl were used as counter and reference electrodes, respectively. The TiO2 NTs or the B–TiO2 NTs electrode were used as working electrodes. The electrolyte was Na2 SO4 and all the photoelectrochemical experiments were carried out in a potentiostat/galvanostat Autolab PGSTAT302N. 2.4. Evaluation of degradation In the course of the Acid Yellow 1 degradation, aliquots were taken at controlled times and discoloration was monitored by UV/vis spectrophotometry (Hewlett-Packard 8453 spectrophotometer) and the mineralization of organic matter with a total organic carbon analyzer (TOC-VCPN , Shimadzu). 3. Results and discussion 3.1. Characterization of doped and bare TiO2 NT electrodes Images of FEG-MEV (Fig. 2) show the formation of self-aligned and self-organized nanotubes perpendicularly to the metallic substrate when Ti foil is submitted to 20 V for 2 h in 1.0 mol L−1 (NH4 )SO4 + 0.5 wt.% NH4 F, with an average diameter of 110 nm and a tube wall of 15 nm. Fig. 2B shows a cross sectional image of TiO2 nanotubes, presenting 800 nm in length. The introduction of boron atoms in the TiO2 lattice did not cause changes in the morphology of the nanotubes, as can be seen in Fig. 2C–F, for all the boron-doped samples. However, the XPS analysis confirmed the presence of boron in the TiO2 NTs lattice, as can be observed in the B 1s spectra in Fig. 3A. For B–TiO2 samples, the B 1s peak appeared at a binding energy of 191.6 eV. According to earlier reports of Zaleska et al. [25] and Gopal et al. [26] this position is consistent with substitutional boron at Ti sites (Ti–O–B–O–Ti). In the Ti 2p spectrum (not shown) the symmetrical 2p3/2 peak is located at 459.0 eV for the doped and undoped samples. As this binding energy is characteristic for pristine TiO2 phase, no influence of B doping on the peak position was detected, excluding a direct Ti–B interaction. In the case of boron atoms inserted in interstitial positions (also compatible with a binding energy of 191.6 eV) the presence of the Ti3+ phase at about 456 eV is expected [27], which was not detected in the Ti 2p spectra. Fig. 3A shows that the intensity of the B 1s peak increases when the concentration of NaBF4 in the anodization electrolyte increases. The quantitative XPS analysis showed an increase of the boron content from 0.4 to 0.7 at.% when changing the
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
3
Fig. 2. FEG-SEM images of (A) top view of TiO2 NTs; (B) cross section of TiO2 NTs; and top view of boron-doped TiO2 NTs with addition of: (C) 70 ppm, (D) 140 ppm, (E) 280 ppm and (F) 560 ppm of NaBF4 in the electrolyte of nanotubes growth.
Fig. 3. (A) B 1s XPS spectrum of boron-doped TiO2 NTs with (1) 280 ppm and (2) 560 ppm of NaBF4 added in the electrolyte; and (B) X-ray diffraction patterns of (1) not annealed TiO2 NT sample, (2) TiO2 NTs and boron-doped TiO2 NTs with addition of (3) 70 ppm, (D) 140 ppm, (E) 280 ppm and (F) 560 ppm of NaBF4 in the electrolyte of nanotubes growth.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7 4
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 4. (A) DRS spectrum to: (1) bare TiO2 NTs; and B-doped TiO2 NTs sample with addition of (2) 70 ppm, (3) 560 ppm, (4) 140 ppm and (5) 280 ppm of NaBF4 in the electrolyte of nanotubes growth. (B) Photocurrent versus potential curves to (1) dark current; (2) TiO2 NTs before annealing, (3) bare TiO2 NTs; and doped samples with addition of (4) 70 ppm, (5) 140 ppm, (6) 560 ppm and (7) 280 ppm of NaBF4 in the electrolyte of nanotubes growth. Linear scan voltammetry, = 10 mV s−1 , in 0.1 mol L−1 Na2 SO4 using UV/vis irradiation of a 125 W Hg mercury lamp. The inset in (A) shows the Tauc’s Plot and band gap of (1) B280ppm –TiO2 NTs and (2) undoped TiO2 NTs.
concentration of NaBF4 in the electrolyte form 280 and 560 ppm. For lower NaBF4 concentrations the boron content was below the detection limit. By X-ray diffraction (XRD) the peaks indicate the formation of anatase phase after annealing the TiO2 NTs (Fig. 3B), and no difference between doped and undoped samples can be observed. The optical properties of the TiO2 NTs and B–TiO2 samples were studied by diffuse reflectance spectroscopy (DRS), which the spectra are shown in Fig. 4A. All samples obtained for TiO2 nanotube grown in different concentrations of NaBF4 added in the electrolyte showed identical absorption spectra. Doped materials presented a new optical absorption shoulder at wavelengths greater than 400 nm, showing that the substitutional boron atoms in the TiO2 lattice may change the optical properties of the material, extending the optical absorption spectrum into the visible region. The inset in Fig. 4A shows the Tauc’s graph [28,29] used to estimate the band gap energy of the materials. This method consists of extrapolating 1/ the linear portion of a plot of ˛(h) as a function of h (eV), where the intercept at ˛ = 0 is the optical band gap (Ebg in eV). In these expressions ˛ is the absorption coefficient, h is Planck’s constant (J s), is the frequency (s−1 ) and ␥ is the power coefficient that can take the values 1/2, 3/2, 2 or 3 depending on the type of electronic transition: direct allowed, direct forbidden, indirect allowed and indirect forbidden, respectively. More accurate results for TiO2 are obtained with the use of = 2 [29]. The diffuse reflectance measurements can be converted to an equivalent absorption coefficient with Kubelka–Munk equation (K–M ≈ ˛ = (1 − R)2 /2R). Based on the described method, the band gap for TiO2 NTs was 3.34 eV and for the boron-doped TiO2 NTs the obtained values were 3.27 and 2.22 eV. There are two intercepts for B-doped TiO2 samples probably due to the formation of intermediary energy levels (created by substituted B atoms) between the valence and conduction band of the TiO2 [30], in agreement with results of Li et al. [18]. There is a debate in the literature about the likely mechanism of decreasing the band gap of titanium dioxide in second-generation photocatalysts (cation or anion doping) [7]. Some papers have reported that anion doping (as B-doping) can introduce localized dopant states in the band gap [7,18,31], while some studies reported that intrinsic defects, as oxygen vacancies, could enhance visible light photoactivity of doped TiO2 through the creation of color centers, including Ti3+ centers [7]. Another situation is the true narrowing of the original TiO2 band gap, which needs heavy anion doping that may compromise its original integrity [7]. Thus, the new absorption shoulder at the visible region in our material can be assigned to excitation of electrons from localized dopant energy levels above the valence band of TiO2 (created by substitutional boron doping) to the conduction band of TiO2 [18].
Fig. 5 shows a schematic representation of a probable mechanism of charges generation and electron flow in a photoelectrochemical system, to B-doped samples. The photoactivity of the doped (Fig. 4B, Curves 4–7) and bare materials (Fig. 4B, Curve 3) were evaluated by comparing photocurrent-potential curves under dark conditions (Fig. 4B, Curve 1) and under UV/vis irradiation both in 0.1 mol L−1 Na2 SO4 pH 6. By illuminating the semiconductor with light energy greater than that of the band gap, electron-hole pairs are generated at the electrode surface. Under UV light, by applying a bias potential higher than the flat band potential of TiO2 , there is a bending of the CB and VB bands causing a more effective charges separation that increases the photocurrent [32]. The B-doped samples (Curves 4–7) showed larger photocurrent than the undoped one (Curve 3). The photocurrent taken at +1.5 V indicates an increase of photoresponse when the boron concentration is increased from 70 to 280 ppm. This is due to the fact that the B-doped samples present a higher utilization of the radiation emitted by the lamp in the UV and visible regions, as already evidenced by Fig. 4A. But, at higher boron concentrations (560 ppm, Curve 6), there was a decrease in the photocurrent, where the boron is not acting as a dopant anymore, but it is trapping the electron generated in the system. Thus, as a better activity is obtained for B–TiO2 NTs when doped with 280 ppm this condition was chosen for further measurements. Curve 2 of Fig. 4B
Fig. 5. Proposal of a photoelectrocatalytic mechanism in the boron-doped TiO2 NTs samples.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
5
Fig. 6. Optimization of different conditions for treatment of 50 ppm of AY1 in 0.1 M Na2 SO4 , where: (A) effect of the amount of boron added to the electrolyte of nanotube growth (pH 6 under Eapp = +1.2 V); (B) effect of annealing (B280ppm –TiO2 NTs, pH 6 under Eapp = +1.2 V); (C) effect of pH (B280ppm –TiO2 NTs, Eapp = +1.2 V); (D) effect of applied potential (B280ppm –TiO2 NTs, pH 2).
shows the photocurrent of B–TiO2 before annealing, proving that the amorphous phase is not photoactive, presenting a photocurrent negligible as the dark current. Fig. 5 illustrates a possible mechanism of charge generation in a photoelectrochemical system using B–TiO2 NTs. The application of a higher potential than the flat band potential drives the electrons to the counter electrode and holes are more available to promote the oxidation reactions. The innovation is, the use of B–TiO2 NTs anode promotes a better use of light, since the material is activated by both UV light and visible light radiated by the commercial lamp. Therefore, more holes (h+ ) are concentrated on the electrode surface, producing a higher oxidant agent (hydroxyl radicals), which could improve the degradation of organic compounds in solution. In order to evaluate the performance of the B–TiO2 NTs electrodes in photoelectrocatalytic oxidation, the Acid Yellow 1 dye, strongly used in hair dye formulations, was chosen as a target compound for degradation.
3.2. Degradation of AY1 cosmetic dye 3.2.1. Optimization of photoelectrocatalysis conditions The influence of boron doping in the TiO2 NTs electrode was evaluated comparing the performance of B–TiO2 NTs doped with 70, 140, 280 and 560 ppm of NaBF4 . For this, the photoelectrocatalytic degradation of the 50 ppm of AY1 was monitored in 1.0 M Na2 SO4 , pH 6 under bias potential of +1.2 V and UV/vis irradiation for comparison the doped and the bare photoanodes. There was 100% of discoloration in all tested electrodes, but the performance is different for total organic carbon (TOC) removal as shown Fig. 6A. Although a maximum of 70% of TOC removal is obtained for bare TiO2 NTs electrode, the results are lower than the borondoped electrodes that show a maximum of 85% of TOC removal. Taking into account the performance during all the treatment, the TiO2 NT electrode prepared with 280 ppm of boron (0.4 at.%) in the anodization was chosen for further experiments.
Fig. 7. (A) Percentage of AY1 degradation monitored at 393 nm; and (B) Percentage of TOC removal of AY1 in the degradation by the different techniques. All used conditions were 100 ppm of AY1 in 0.01 M Na2 SO4 , pH 2, Hg 125 W lamp and Eapp = +1.2 V.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model
ARTICLE IN PRESS
CATTOD-9023; No. of Pages 7
G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
6
Table 1 Discoloration rate constant (k) obtained for ln[A/A0 ] versus time plot during degradation of 100 mg L−1 AY1 dye in 0.01 mol L−1 Na2 SO4 at pH 2; using UV/vis irradiation from Hg 125 W lamp and Eapp = +1.2 V in PEC experiments. Treatment
Electrode
k (min−1 )
Direct photolysis (DP)
–
−0.0083
Photocatalysis (PC)
TiO2 NTs B–TiO2 NTs
−0.0102 −0.0109
Photoelectrocatalysis (PEC)
TiO2 NTs B–TiO2 NTs
−0.0199 −0.0235
Fig. 6B shows the efficiency of TOC removal with not annealed TiO2 and B–TiO2 electrodes in comparison with the annealed ones. There was about 35% of TOC removal at the end of 120 min, due to degradation by direct photolysis. These data confirm what has already been observed by photocurrent curves (Fig. 4B, Curve 2) where the photocurrent generated was negligible, next to the current obtained in the dark. The effect of original pH (2, 4.5 and 6) in the TOC removal efficiency for B–TiO2 NTs electrode was tested in the photoelectrocatalytic degradation of 50 ppm of AY1 in 0.1 M Na2 SO4 , under bias potential of +1.2 V and UV/vis irradiation (Fig. 6C). Higher degradation efficiency (85%) was reached after 60 min of treatment in pH 2. This illustrates that the acid dye can be strongly adsorbed on the electrode surface at very acid conditions, when the pHpzc for TiO2 is around 5.3 [33] and the dye is in anionic form (sulfonate dye) [34]. Thus, pH 2 was chosen for further experiments. In order to check the influence of applied potential, further experiments were carried out comparing the performance at +0.50 V, +1.2 V and +1.5 V (Fig. 6D). The results indicate that the TOC removal is improved at Eapp = +1.2 V reaching 93% of mineralization. All the applied potential is higher than the flat band potential so the electrons are depleted and holes enriched at the surface of B–TiO2 NTs. At a higher potential than +1.2 V, such as +1.5 V it could happen the formation of a competitive reaction such as oxygen evolution that decreases the photoelectrocatalytic performance [5,35]. Thus, considering the TOC removal, a potential of +1.2 V was used. Taking into account that in a photoelectrochemical system the electrolytes interfere drastically in the band bending of the electrode/solution interface, was investigated the effect of electrolyte composition, testing Na2 SO4 concentration of 0.1 mol L−1 , 0.05 mol L−1 and 0.01 mol L−1 . The results do not show any important difference in the TOC removal showing that the ideal conductivity can be reached at sulfate concentration of 0.01 mol L−1 and this condition was adopted in further experiments [5]. Comparison of B-doped TiO2 NT and TiO2 NT electrodes on AY1 degradation under optimized conditions. Using the best experimental conditions defined previously, it was evaluated the photoelectrocatalytic (PEC) performance for the treatment of 100 ppm of AY1 dye in solution of 0.01 M Na2 SO4 at pH 2 under application of +1.2 V when compared to electrocatalysis (EC; Eapp = +1.2 V + catalyst); direct photolysis (DP; UV/vis commercial irradiation); and photocatalysis (PC; UV/vis irradiation + catalyst). The results are shown in Fig. 7 comparing discoloration (Fig. 7A) and TOC removal (Fig. 7B) during 120 min of treatment. Fig. 7A shows the percentage of AY1 degradation monitored at 393 nm by UV/vis spectrophotometry. Except for the electrocatalysis when the catalyst is not active since there is no irradiation, all the other processes promoted 100% discoloration. The degradation kinetics follows a pseudo-first-order reaction (Langmuir–Hinshelwood kinetics) and the obtained constant rates of discoloration at B-doped and undoped electrodes are shown in Table 1. The results indicate that PEC degradation rate is about 2.1
Fig. 8. (A) Photocurrent density curves to: (1) dark current, (2) TiO2 NTs in glass tube, (3) TiO2 NTs in quartz tube, (4) B–TiO2 NTs in glass tube and (5) B–TiO2 NTs in quartz tube; (B) percentage of AY1degradation monitored at 393 nm; and (C) percentage of TOC removal of AY1 by the different techniques using glass tube (g) and quartz tube (q). All used conditions were 100 ppm of AY1 in 0.01 M Na2 SO4 , pH 2 and Eapp = +1.2 V.
times faster than the PC degradation showing the important effect of the bias potential. It is clear in Fig. 7B that the PEC degradation is much more effective in the AY1 mineralization than EC, PC and DP indicating that the separation of charges promoted by PEC treatment improves the degradation efficiency. In addition, the doped electrodes increase the degradation performance (75% of TOC removal in 90 min). This is due to the fact that the doped electrode can absorb at higher wavelengths range emitted by the radiation source. Thus, the higher amount of photogenerated holes can produce more hydroxyl radicals, promoting an increased mineralization rate. Regarding the comparison between the PC and the PEC processes, it is clear that the mineralization percentage of PEC is higher than the PC, around 40% superior at 90 min of degradation. These results prove that photoelectrocatalysis allows more effective
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073
G Model CATTOD-9023; No. of Pages 7
ARTICLE IN PRESS G.G. Bessegato et al. / Catalysis Today xxx (2014) xxx–xxx
separation of photogenerated charges thus increasing the lifetime of e− /h+ pairs [4,36]. The DP can remove only 34% of total organic carbon, and in the EC the TOC removal is negligible. In order to evaluate the visible activity of B-doped TiO2 NTs produced, we change the quartz tube which the lamp is inserted in the photoelectrochemical reactor, for a glass tube that can filter large part of the UV radiation. Fig. 8A shows the photocurrent curves taken with the lamp inserted in a glass tube (g) and in a quartz tube (q) for doped and undoped electrodes. It can be seen that there was a greater percentage reduction of the photocurrent to the undoped electrode than for B-doped one, when the tube was changed. Fig. 8B and C shows the graphs of AY1 degradation and TOC removal for a comparison between degradation efficiency using a glass or quartz tube. In Fig. 8B, the PEC AY1 degradation using quartz tube was very similar. But when using a glass tube, the PEC degradation rate using B-doped electrode was much higher than the PEC using bare TiO2 NTs. In Fig. 8C the B-doped electrode was almost 2 times more efficient than the undoped electrode in TOC removal at the end of 120 min, while using the quartz tube the difference between the two electrodes was smaller. 4. Conclusions Our findings show that boron-doped TiO2 nanotubes can be easily obtained by a one step electrochemical process, where boron atoms are introduced into TiO2 lattice during growth of the nanotubes by electrochemical anodization. The obtained material showed absorption at the visible region of spectrum (presenting a shoulder with band gap of 2.22 eV) and higher photoactivity than the bare TiO2 nanotubes, due to the insertion of boron atoms into the TiO2 lattice. The photoelectrocatalytic treatment applied in the degradation of AY1 by using B–TiO2 NTs was more efficient than the one on the bare electrode due to a better use of the irradiation emitted in the UV and visible region by using a commercial lamp. The effect of the applied potential improved the degradation rate around 2 times when the PEC is compared with a PC technique. Thus, the material consists of an important step in the search for efficient and stable catalysts for application in photo(electro)catalysis, even being activated by solar radiation. Acknowledgments The authors acknowledge the financial support provided by the Brazilian funding agency FAPESP (process 2012/183570; 2011/23128-7 and 2008/10449-7). FEG-SEM facilities were
7
provided by LMA-IQ and X-Ray Diffraction measurements by GFQM-IQ. We are grateful for the contributions of Prof. Dr. Peter Hammer on acquisitions and discussion of XPS data. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]
[31] [32] [33] [34]
[35] [36]
K. Vinodgopal, S. Hotchandani, P.V. Kamat, J. Phys. Chem. 97 (1993) 9040–9044. D.H. Kim, M.A. Anderson, Environ. Sci. Technol. 28 (1994) 479–483. A. Fujishima, K. Honda, Nature 238 (1972) 37–38. J. Georgieva, E. Valova, S. Armyanov, N. Philippidis, I. Poulios, S. Sotiropoulos, J. Hazard. Mater. 211 (2012) 30–46. R. Daghrir, P. Drogui, D. Robert, J. Photochem. Photobiol. A 238 (2012) 41–52. J.C. Cardoso, T.M. Lizier, M.V. Boldrin Zanoni, Appl. Catal. B 99 (2010) 96–102. N. Serpone, J. Phys. Chem. B 110 (2006) 24287–24293. R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293 (2001) 269–271. A. Ghicov, J.M. Macak, H. Tsuchiya, J. Kunze, V. Haeublein, L. Frey, P. Schmuki, Nano Lett. 6 (2006) 1080–1082. J.H. Park, S. Kim, A.J. Bard, Nano Lett. 6 (2006) 24–28. G. Yan, M. Zhang, J. Hou, J. Yang, Mater. Chem. Phys. 129 (2011) 553–557. Y. Su, S. Han, X. Zhang, X. Chen, L. Lei, Mater. Chem. Phys. 110 (2008) 239–246. A. Ghicov, B. Schmidt, J. Kunze, P. Schmuki, Chem. Phys. Lett. 433 (2007) 323–326. Z. Xu, J. Yu, Nanoscale 3 (2011) 3138–3144. J. Gong, W. Pu, C. Yang, J. Zhang, Catal. Commun. 36 (2013) 89–93. C. Di Valentin, G. Pacchioni, Catal. Today 206 (2013) 12–18. N. Lu, X. Quan, J. Li, S. Chen, H. Yu, G. Chen, J. Phys. Chem. C 111 (2007) 11836–11842. J. Li, N. Lu, X. Quan, S. Chen, H. Zhao, Ind. Eng. Chem. Res. 47 (2008) 3804–3808. A. Subramanian, H.-W. Wang, Appl. Surf. Sci. 258 (2012) 6479–6484. A.E.R. Mohamed, S. Rohani, Energy Environ. Sci. 4 (2011) 1065–1086. K. Shankar, J.I. Basham, N.K. Allam, O.K. Varghese, G.K. Mor, X. Feng, M. Paulose, J.A. Seabold, K.-S. Choi, C.A. Grimes, J. Phys. Chem. C 113 (2009) 6327–6359. C.A. Grimes, G.K. Mor, TiO2 Nanotube Arrays: Synthesis, Properties, and Applications, Springer, New York, 2009. R. Jain, V.K. Gupta, S. Sikarwar, J. Hazard. Mater. 182 (2010) 749–756. G.G. Bessegato, J.C. Cardoso, B.F. Silva, M.V.B. Zanoni, J. Photochem. Photobiol. A 276 (2013) 96–103. A. Zaleska, J.W. Sobczak, E. Grabowska, J. Hupka, Appl. Catal. B 78 (2008) 92–100. N.O. Gopal, H.-H. Lo, S.-C. Ke, J. Am. Chem. Soc. 130 (2008) 2760–2761. G. Liu, H.G. Yang, X. Wang, L. Cheng, H. Lu, L. Wang, G.Q. Lu, H.-M. Cheng, J. Phys. Chem. C 113 (2009) 21784–21788. J. Tauc, R. Grigorovici, A. Vancu, Phys. Status Solidi 15 (1966) 627–637. R. Lopez, R. Gomez, J. Sol-Gel Sci. Technol. 61 (2012) 1–7. M. Pelaez, N.T. Nolan, S.C. Pillai, M.K. Seery, P. Falaras, A.G. Kontos, P.S.M. Dunlop, J.W.J. Hamilton, J.A. Byrne, K. O’Shea, M.H. Entezari, D.D. Dionysiou, Appl. Catal. B 125 (2012) 331–349. L. Deng, Y. Chen, M. Yao, S. Wang, B. Zhu, W. Huang, S. Zhang, J. Sol-Gel Sci. Technol. 53 (2010) 535–541. H.O. Finklea, Semiconductor Electrodes, Elsevier, New York, 1988. S. Malato, P. Fernandez-Ibanez, M.I. Maldonado, J. Blanco, W. Gernjak, Catal. Today 147 (2009) 1–59. K. Rajeshwar, M.E. Osugi, W. Chanmanee, C.R. Chenthamarakshan, M.V.B. Zanoni, P. Kajitvichyanukul, R. Krishnan-Ayer, J. Photochem. Photobiol. C 9 (2008) 171–192. Y. Zhang, X. Xiong, Y. Han, X. Zhang, F. Shen, S. Deng, H. Xiao, X. Yang, G. Yang, H. Peng, Chemosphere 88 (2012) 145–154. M.V.B. Zanoni, J.J. Sene, M.A. Anderson, J. Photochem. Photobiol. A 157 (2003) 55–63.
Please cite this article in press as: G.G. Bessegato, et al., Enhanced photoelectrocatalytic degradation of an acid dye with boron-doped TiO2 nanotube anodes, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.03.073