V-doped TiO2 anatase: Calculation and experiment

127 JOURNAL OF SCIENCE OF HNUE DOI: 10.18173/2354-1059.2017-0040 Mathematical and Physical Sci. 2017, Vol. 62, Iss. 8, pp. 127-134 This paper is available online at V-DOPED TiO2 ANATASE: CALCULATION AND EXPERIMENT Duong Quoc Van, Le Minh Thu and Nguyen Minh Thuy Fuculty of Physics, Hanoi National University of Education Abstract. To understand the effects of V in the increment of photocatalytic activity of V- doped TiO2, the possible position of V in the TiO2 lattice is an important parameter. This research performed ab-initio calculations to find the possible doped position of V atoms into the lattice of TiO2 anatase and compared that with experimental results. Calculated formation energies of un-doped and doped models prove that V atoms prefer to substitute into Ti positions rather than O positions or interstitial positions, in a good agreement with XPS results. Electronic structures show that doped V atoms lead to the formation of new energy levels near the bottom of conduction bands, reduce band-gap value and increase the photocatalytic activity of TiO2 material. Keywords: V-doped TiO2, doping positions, substitional, DFT, XPS. 1. Introduction Since being discovered by Asahi et al. [1], doped TiO2 have been widely studied and applied in various fields such as water and air treatments, batteries and self-cleaning materials due to its smaller band-gap, compared to un-doped TiO2. Various elements have been doped into TiO2, including metals: Ag, Au, Cu, Co, Cr, Fe, Mn, Mo, Nb, Ni, Ru, V [2-4] and non-metals: B, C, F, I, N, S, P [5-7]. Among all of these elements, V is shown to be a good element to study because V doping can increase the lifetime of carriers [8], extend the absorption range to visible light [9] and improve the photocatalytic activity of TiO2 [10-12]. Photocatalytic activity of V-doped TiO2 were influenced by positions and valences of doped V ions. Long et al. [13] showed that V atoms can be doped into a TiO2 lattice as substitutional defects rather than interstitial defects. Doped V atoms create new energy levels in the band-gap of TiO2, reduce ban gap energy and extend absorption edges to the visible region. In fact, V atoms can be doped into TiO2 as substitutional and interstitial defects and their effects are different. Islam et al. [14] proved that V ions prefer to exist in a TiO2 lattice in the form of V 4+ and V 5+ ions more than V 2+ and V 3+ . Experimental results show that most of the ion V in TiO2 materials have valences of +4, the second most being +5 [11, 15]. However, photocatalytic mechanism of V-doped TiO2 depends on doping position of V atoms in a TiO2 lattice; which has not been asserted. The unsolved problems are where the V atoms will exist in a TiO2 lattice and what valences will V ions have? Received September 7, 2017. Accepted September 30, 2017. Contact Duong Quoc Van, e-mail: vandq@hnue.edu.vn Duong Quoc Van, Le Minh Thu and Nguyen Minh Thuy 128 In this research, V-doped anatase TiO2 has been studied by both theoretical and experimental methods. Different models of V-doped TiO2, wherein V has valence +4, have been built and calculated using software based on density functional theory. The formation energies of models were used to determine the possible doping positions of V atoms in lattice of TiO2. Un-doped and V-doped TiO2 samples were synthesized using hydrothermal methods and analyzed using different techniques to find the doping state of V atoms. The calculated and experimental results are compared to confirm the possible doping positions of V atoms and their valences. 2. Content 2.1 Computational and Experimental Methods 2.1.1 Computational Methods In this study, a 2×2×1 supercell containing 16 Ti atoms and 32 O atoms (labeled as TOO) of pure anatase TiO2 was used to make calculation (Figure 1a). Three modified supercells were created to study the possible doped positions of V atoms into TiO2 a lattice (see Figure 1b – 1d). The TOV-s model ware constructed by substituting one O atom with one V atom whereas the TOV-i models was constructed by embedding one V atom into the interspace, and the TVO-s model was constructed by substituting one Ti atom with one V atom. a. TOO (un-doped). b. TOV-i c. TOV-s d. TVO-s Figure 1. (Color online) The 2x2x1 supercell of pure and V-doped TiO2 models. Calculations based on first principles were performed using the DMol3 modules in Materials Studio. The interactions of electron-ions were modeled using the Vanderbilt ultrasoft pseudopotentials [16] with valence atomic configurations that for Ti are 3s 2 3p 6 3d 2 4s 2 , for O is 2s 2 2p 4 and for V is 3d 3 . The wave functions was expanded through a plane wave basis set with a cutoff energy of 380 eV. The Monkhorst - Pack scheme [17] k-points grid sampling was set at 7×7×3 in the supercells. The convergence threshold for self-consistent iterations was set at 5×10 -7 eV. In the geometry optimization process, the energy change, maximum force, maximum stress and maximum displacement tolerances were set at 5×10 −6 eV/atom, 0.01 eV/Å, 0.02 GPa and 5.10 -4 Å, respectively. 2.1.2 Experimental Methods Doped TiO2 samples were synthesized using the hydrothermal method from initial chemicals: TiCl4, H2O, V2O5, NH4NO3, NH3 10% and HCl. The synthesization procedure can be divided into the following steps: First, 3.0 ml of TiCl4 were added drop by drop into 100 ml of distilled water and V2O5/HCl solution, stirring continuously for 20 minutes. Secondly, NH3 solution was added to adjust the pH value to 7 ÷ 8, then stirred for 30 minutes. Afterwards, distilled water was added and the mixture was transferred into an autoclave and heated at 200 o C Ti O V V-doped TiO2 anatase: calculation and experiment 129 for 5 h. A similar process was used to synthesize an undoped TiO2 sample, using only distilled water instead of a water and V2O5/HCl mixture. Finally, after centrifugation, washing and drying at 120 o C, the sample was calcined at 650 o C for 1 h. The doping concentration of V was calculated using the following equation: %V = nV / (nV + nTi) (1) where nV and nTi were the mole of V and Ti, respectively. In this study, the chosen V concentration is 0.5% at., based on our previous work [18]. Crystal structures of samples were analyzed using X-ray diffraction (XRD) SIEMENS D5005, and the morphology was investigated by scanning electron microscope (SEM) and transmission electron microscope (TEM). The absorption spectra were obtained using a Jasco V670 system, the scanned wavelengths being 200 to 800 nm. The bonding states were characterized by X-ray photoelectron spectroscopy (XPS) in a commercial Microlab 350 XPS system equipped with an Al Kα source, in the ultra-high vacuum (UHV) chamber (approximate 10 - 9 Torr). 2.2. Calculation Results 2.2.1 Formation Energy The possible doped position can be determined by comparing the formation energies of pure and defective models: the lower the formation energy, the easier the models formed. Formation energies (𝐸mod) were calculated using the following formula [19]: Emod = Etot (models) – Etot (pure) - mµV + nµTi + pµO (2) Etot (models) and Etot (pure) are the total energies of V-doped models and pure TiO2; µV, µTi and µO represent the chemical potentials of the V, Ti and O atoms; m, n and p are the numbers of doped vanadium, removed titanium and oxygen atoms in the models. Values of m, n and p for different models are shown in Table 1. Table 1. Values of doped vanadium atoms (m) and removed titanium (n) and oxygen atoms (p) for pure and defective TiO2 models. Models TOO TOV-i TOV-s TVO-s m 0 1 1 1 n 0 0 0 1 p 0 0 1 0 The chemical potential of V is calculated from the energy of a single V atom in bulk material where the chemical potentials of Ti and O are depended on experiment conditions: Ti- rich and O-rich growth conditions. In the Ti-rich growth condition, the chemical potential of Ti is larger than that of O; in the O-rich growth condition, the chemical potential of O is larger than that of Ti. Determination formulas of µTi and µO in different growth conditions are shown in Table 2, where µTi-bulk is the energy of single Ti atoms in bulk material, µTO and µoxy are the total energies of TiO2 and O2 molecules. Table 2. Chemical potentials of Ti and O atoms in different conditions [19]. Ti-rich µTi = µTi-bulk µO = (µTO - µTi)/2 O-rich µO = µoxy/2 µTi = µTO - 2µO Duong Quoc Van, Le Minh Thu and Nguyen Minh Thuy 130 The calculated results of formation energies of defective models are showed in Table 3. The model with the smaller formation energy is more stable than the model with the larger value. The TVO-s model has the smallest formation energy, indicating that the TVO-s model forms easier than the TOV-s or TOV-i models, or the TVO-s is the most stable model. The smallest values of formation energies of TVO-s models in both growth conditions also suggest that the substitution of V atoms into the TiO2 lattice do not depend on experimental conditions. This implies that when doped into TiO2 lattice, V atoms substitute into Ti atoms positions, consistent with previous results [13]. Table 3. Formation energies of defective TiO2 models. Models TOV-i TOV-s TVO-s Ti-rich (eV) 5.929 7.659 0.396 O-rich (eV) 5.929 3.333 1.705 2.2.2 Electronic Structures Total density of states (DOS) and projected density of states (PDOS) of un-doped and doped models were calculated and shown in Figure 2. The calculated band-gap of un-doped model TOO is 3.201 eV, in a good agreement with experimental result (3.2 eV). The calculated band-gap of the TOV-s model is around 1.50 eV, smaller than 3.201 eV of the pure model, mean that V doping lead to the shift of absorption edge to the visible light region. Figure 2 also shows the appearance of V 3d states that lie under the conduction band, leading to the narrowing band-gap of V-doped TiO2, in a good agreement with results from Islam [14]. Figure 2. (Color online) Projected density of states (PDOS) of (a) un-doped and (b) V-doped TiO2 models. The theoretical studies show that the TVO-s model has the smallest energy, indicating that V atoms substitute into Ti atoms positions when doped into TiO2 lattice [13]. The new energy levels produced by V atoms, which just lie under conduction bands, leads to the reduction of band-gap and improvement of photocatalytic activity of TiO2. 2.3 Experimental Results 2.3.1 Sample Characterization Figure 3 presented the XRD patterns of un-doped (labelled as HT) and 0.5%V-doped TiO2 (labelled as HV) samples prepared by hydrothermal methods. For both samples, the XRD patterns show only anatase TiO2 peaks, indicating that the doping of V cannot be detected by diffraction a b V-doped TiO2 anatase: calculation and experiment 131 techniques. This result means that the V concentration is too small to be detected by XRD or V atoms and spreads out in the samples, consistent with a previous study for V-doped TiO2 [20]. Figure 3. (Color online) XRD patterns of HT and HV samples. Figure 4. (Color online) UV-Vis absorption spectra of HT and HV samples. 2.3.2 Absorption Spectra The substitution of V atoms into TiO2 lattice can be predicted by the change in UV-Vis spectrum of the V-doped sample in Figure 4. The un-doped sample HV has an absorption edge at 390 nm, in a good agreement with experimental band-gap value 3.2 eV [21]. The V-doped sample HV exhibit three changes in absorption spectrum: a little redshift of absorption edge around 400 nm, a significant increment of absorbance in the visible range and a new weak absorption peak around 650 nm. These changes can be explained by using the band structure of V-doped TiO2 in Figure 5. The dotted line in the band structure is represented for t2g energy levels, which were created by V 3d electrons. The little redshift around 400 nm belongs to the charge transfers from TiO2 valence band to the t2g level of V, just lies below the conduction band of TiO2 [22]. The increment of absorbance in the range of 400 - 600 nm indicate the transfers from 3d electrons levels of V 4+ to the conduction band of TiO2 [23] while the weak absorption peak at around 650 nm is due to d-d transitions of V 3d electrons [24]. The changes in absorption spectrum of the HV sample are indirect evidence of substitution of V atoms into the TiO2 lattice; the clearest evidence can be seen on the XPS spectra shown in Figure 6. Figure 5. Band structures of V-doped TiO2. Conduction Band dẫn (CB) Valance Band Eg V 3d Duong Quoc Van, Le Minh Thu and Nguyen Minh Thuy 132 2.3.3 XPS Results The XPS spectra of un-doped and doped TiO2 samples and their Gaussian fittings are shown in Figure 6. For the un-doped sample in Figure 6b, O 1s was composed of different oxygen- related bonds like Ti-OH, O-Ti-O [11, 15]. The doped sample O 1s peak shows a notable shift to higher energy, indicates that new bonding peaks appeared in the doped sample. Figure 6c shows that the O 1s peak of HV was composed of three different peaks: a O-Ti-O bonding peak at 530.9 eV, a Ti-OH bonding peak at 532.6 eV and another peak at 534.2 eV, which is due to V-O bonding [25, 26]. XPS results show that only V-O bonding peaks appeared in V-doped samples, evidence that Ti-V bonding peaks does not exist. This indicates that the V atom has been replaced into positions of Ti atoms instead of O atoms, consistent with previous results [25, 26]. The experimental results are consistent with our calculations in section 2.2, confirm that V atoms substitute into Ti atoms when they are doped into an TiO2 lattice. Figure 6. (a) The O 1s XPS spectra of un-doped (HT) and V-doped TiO2 (HV), the Gaussian fitting of O 1s peaks of (b) HT and (c) HV samples. 3. Conclusion The doping positions of V atoms have been researched using both theoretical and experimental methods. The calculated results show that in the doped model Ti atoms substituted by V atoms have smaller formation energy than other models, implying that V atoms substitute into Ti atoms position when doped into a TiO2 lattice. Experimental results show that the substitution of V atoms into TiO2 cannot be observed by XRD patterns due to their small concentration or uniform distribution in the samples. Indirect evidences of the substitution can be found through the change in UV-Vis spectra of the V-doped TiO2 sample. The V-doped sample spectrum show a long tail with a length of 400 - 600 nm - which is assigned to the charge-transfer transitions from d orbitals of V 4+ to the conduction bands of TiO2, and a weak absorption band in the range of 650-700 nm due to d-d transitions of V 3d electrons. The direct evidence can be seen on the XPS spectra. The O 1s peaks of V-doped sample has an unsymmetrical shape, which is due to the V-O peak at the 532.4 eV position. This peak is clearly evidence that V atom has doped into the TiO2 lattice and substituted into the Ti atom position, consistent with theoretical results. REFERENCES [1] R. Asahi, et al., 2001. Visible-Light Photocatalysis in Nitrogen-Doped Titanium Oxides. Science, 293, 5528, pp. 269–271. a b c V-doped TiO2 anatase: calculation and experiment 133 [2] M. Anpo, 2000. Use of visible light. Second-generation titanium oxide photocatalysts prepared by the application of an advanced metal ion-implantation method. Pure and Applied Chemistry, 72, 9, pp. 1787. [3] A. Fuerte, et al., 2001. Visible light-activated nanosized doped-TiO2 photocatalysts. Chemical Communications, 24, pp. 2718-2719. [4] H. Yamashita, et al., 2001. Application of Ion Beam Techniques for Preparation of Metal Ion-Implanted TiO2 Thin Film Photocatalyst Available under Visible Light Irradiation: Metal Ion-Implantation and Ionized Cluster Beam Method. 569-71. [5] T. Ohno, et al., 2003. Photocatalytic Activity of S-doped TiO2 Photocatalyst under Visible Light. Chemistry Letters, 32, 4, pp. 364-365. [6] Y. Liu, et al., 2005. Photocatalytic degradation of azo dyes by nitrogen-doped TiO2 nanocatalysts. Chemosphere, 61, 1, pp. 11-18. [7] J. C. Yu, et al., 2003. Synthesis and Characterization of Phosphated Mesoporous Titanium Dioxide with High Photocatalytic Activity. Chemistry of Materials, 15, 11, pp. 2280-2286. [8] N. Serpone, et al., 1994. Spectroscopic, Photoconductivity, and Photocatalytic Studies of TiO2 Colloids: Naked and with the Lattice Doped with Cr 3+ , Fe 3+ , and V 5+ Cations. Langmuir, 10, 3, pp. 643-652. [9] S. Klosek, D. Raftery, 2001. Visible Light Driven V-Doped TiO2 Photocatalyst and Its Photooxidation of Ethanol. The Journal of Physical Chemistry B, 105, 14, pp. 2815-2819. [10] W.-F. Chen, et al., 2017. Photocatalytic activity of V-doped TiO2 thin films for the degradation of methylene blue and rhodamine B dye solutions. Journal of the Australian Ceramic Society. [11] Y.-H. Lin, 2014. Photocatalytic Degradation of 1, 2-dichloroethane by V/TiO2: The Mechanism of Photocatalytic Reaction and Byproduct. Aerosol and Air Quality Research. [12] J. C. S. Wu, C.H. Chen, 2004. A visible-light response vanadium-doped titania nanocatalyst by sol–gel method. Journal of Photochemistry and Photobiology A: Chemistry, 163, 3, pp. 509-515. [13] R. Long, N. J. English, 2010. First-principles calculation of electronic structure of V-doped anatase TiO2. Chem. Phys. Chem., 11, 12, pp. 2606-2611. [14] M. M. Islam, et al., 2011. Electronic properties of vanadium-doped TiO2. Chem. Phys. Chem, 12, 17, pp. 3467-73. [15] M. Khan, et al., 2013. Effect of V doping concentration on the electronic structure, optical and photocatalytic properties of nano-sized V-doped anatase TiO2. Materials Chemistry and Physics, 142, 1, pp. 148-153. [16] D. Vanderbilt, 1990. Soft self-consistent pseudopotentials in a generalized eigenvalue formalism. Physical Review B, 41, 11, pp. 7892-7895. [17] H. J. Monkhorst, J. D. Pack, 1976. Special points for Brillouin-zone integrations. Phys. Rev. B, 13, 12, pp. 4. [18] N. M. Thuy, et al., 2011. A Visible Light Activity of TiO2 based Photocatalysts. Journal of Science of HNUE, 56, 1, pp. 11-20. Duong Quoc Van, Le Minh Thu and Nguyen Minh Thuy 134 [19] H.-C. Wu, et al., 2012. Effect of Fe Concentration on Fe-Doped Anatase TiO2 from GGA+U Calculations. International Journal of Photoenergy. [20] J. H. An, 2005. Preparation of vanadium - doped TiO2 thin films on glass substrates. Journal of Ceramic Processing Research, 6, 2, pp. 4. [21] W. M. Haynes, et al., 2016. CRC Handbook of Chemistry and Physics. CRC Press. [22] B. Tian, et al., 2009. Flame sprayed V-doped TiO2 nanoparticles with enhanced photocatalytic activity under visible light irradiation. Chemical Engineering Journal, 151, 1-3, pp. 220-227. [23] S. Liu, et al., 2009. Highly active V–TiO2 for photocatalytic degradation of methyl orange. Applied Surface Science, 255, 20, pp. 8587-8592. [24] W. Choi, et al., 1994. The Role of Metal Ion Dopants in Quantum-Sized TiO2: Correlation between Photoreactivity and Charge Carrier Recombination Dynamics. J. Phys. Chem., 98, pp. 13669-13679. [25] B. Liu, et al., 2009. Low temperature fabrication of V-doped TiO2 nanoparticles, structure and photocatalytic studies. Journal of Hazardous Materials, 169, 1–3, pp. 1112-1118. [26] J. Zhu, et al., 2007. Nanocrystalline anatase TiO2 photocatalysts prepared via a facile low temperature nonhydrolytic sol–gel reaction of TiCl4 and benzyl alcohol. Applied Catalysis B: Environmental, 76, 1, pp. 82-91.