Journal of the Korean Physical Society, Vol. 70, No. 11, June 2017, pp. 995∼1000 Effect of Co-Doping and Tri-Doping with Transition Metals and a Nonmetal on Photocatalytic Activity in Visible Light of TiO2 Thin Film Hang Nguyen Thai Phung∗ Department of Physics, Tay Nguyen University, Dak Lak province, Viet Nam Department of Applied Physics, Ho Chi Minh University of Science, Ho Chi Minh city, Viet Nam Van Nguyen Khanh Tran, Phuong Ai Duong and Hung Vu Tuan Le Department of Applied Physics, Ho Chi Minh University of Science, Ho Chi Minh city, Viet Nam Nguyen Duc Truong Department of Applied Physics, Ho Chi Minh City University of Transport, Ho Chi Minh city, Viet Nam (Received 13 March 2017, in final form 28 April 2017) Mono, co- and tri-doped TiO2 thin films with the transition metals (vanadium and/or chrominium) and a nonmetal (nitrogen) have been fabricated by sol-gel method. X-ray diffraction results clearly reveal anatase crystal structure for all obtained samples and doping with any dopants doesn’t change the anatase phase of TiO2 . Compared to TiO2 , three types of doped TiO2 thin films exhibit a red-shift in the absorption edge and have much better photocatalytic properties for methylene blue degradation in visible light region. The maximum visible-photocatalytic performance was achieved for tri-doped TiO2 sample. The mechanism for enhancing visible-photocatalytic activity of co-doped and tri-doped TiO2 thin films was also examined. PACS numbers: 78.66.-w, 78.20.-e, 78.67.-n Keywords: TiO2 , Co-doped, Tri-doped, Degradation, Methylene blue, Photocatalytic, Visible light DOI: 10.3938/jkps.70.995 I. INTRODUCTION TiO2 has been widely applied for the hydrogen production from water splitting and the environment such as water and air purification because of its notable photocatalytic capability [1–3]. However, due to its large band gap, TiO2 can only absorb the ultraviolet (UV) light which accounts for only 5% of the solar energy, so some applications of TiO2 photocatalyst have been limited to UV light region. Therefore, it takes signifficant efforts to make TiO2 photocatalyst reactive under visible light irradiation including doping, sensitization with a small band-gap semiconductor or organic dye. One of these methods is the doping of TiO2 with various elements. Choi et al. experimentally studied the system of 21 transition metals doped TiO2 photocatalysts, and the results showed that doping with single transition metal could enhance visible-photocatalytic activity of TiO2 [4]. Among transition metal elements, V and Cr are considered as interesting dopants and there are many reports related to V or Cr doped TiO2 by using chemical ∗ E-mail: pnthang@ttn.edu.vn pISSN:0374-4884/eISSN:1976-8524 and physical methods [5–7]. Actually, transition metaldoped TiO2 photocatalysts are prevented by the presence of carrier recombination centers and the formation of strongly localized states in its band-gap, which is detrimental to carrier mobility [8]. In order to minimize its role as recombination centers, some researchers synthesized co-doped TiO2 with V or Cr and a nonmetal element [9,10]. At present, the researches on tri-doping, especially relating to V, Cr and N, tri-doped TiO2 thin film have not been studied by scientists yet. In this paper, we prepared mono, co and tri-doped TiO2 thin films with V, Cr and N by sol-gel method and compared their photocatalytic activity in visible light region. The mechanism for enhancing the visible-photocatalytic activity of doped TiO2 is also discussed. II. EXPERIMENTS AND DISCUSSION Tetra butyl orthotitanate (Ti(OBu)4 ) (99.99%) was added to the mixture of diethanolamine (HN(C2 H4 OH)2 ) (99.99%) and ethanol (C2 H5 OH) (99.99%). After stirring for 1 hour, distilled water was -995- c 2017 The Korean Physical Society -996- Journal of the Korean Physical Society, Vol. 70, No. 11, June 2017 added dropwise in this solution with continous stirring for 2 hours to obtain solution A (TiO2 sol). Vanadium III chlorine (VCl3 ) (99.99%) and chrominum III chlorine (CrCl3 ) (99.99%) were completely dissolved in ethanol and the obtained solutions were labeled as solution B and solution C, respectively. Solution D was denoted as a solution in which ethanol and distilled water dissolved urea ((NH2 )2 CO2 ) (99.98%). Adding dropwise in pairs of three solutions (B, C, and D) to solution A with continuous stirring for 1 hour, we obtained co-doped TiO2 sols, which were named as TiO2 :(V, Cr), TiO2 :(V, N) and TiO2 :(Cr, N), respectively. By the similar way, all three solutions (B, C and D) were added to solution A to generate tri-doped TiO2 sol which is denoted as TiO2 :(V, Cr, N). With same procedure, same conditions and same amount of dopants, we alternatively added one of three solutions (B, C, and D) to solution A to get single-doped TiO2 sols which are denoted as TiO2 :V, TiO2 :Cr and TiO2 :N, respectively. Dipping glass substrates into all resultant sols, we obtained single, co- and tri-doped TiO2 thin films which were further baked at 80 ◦ C for 15 minutes and annealed in the air at 500 ◦ C for 2 hours. Names of obtained thin films correspond to the resultant sols. Investigation on crystalline structure of doped TiO2 thin films were carried out by Brucker D8 Advance X-ray Diffraction (XRD) spectrometer with Cu Kα. A Halo RB-10 spectroscopy was used in order to measure the optical absorption property. Degradation rate of methylene blue (MB) was measured to evaluate the photocatalytic activity of the obtained samples. Every 25 × 25 mm2 thin film was dipped into 10 ml of 10 ppm MB solution. The visible light source used in this experiment was a compact lamp. Distance between light source and MB solution was about 5 cm. MB absorptions at 662 nm were measured by a Halo RB-10 spectroscopy at intervals of 30 minutes. Lambert-Beer’s law has been used to calculate the change of MB concentrationat regular irradiation intervals. X-ray photoelectron spectroscopy (XPS) measurements were performed on VG Microlab MK II with Aluminum Kα radiation as the excitation source. Figs. 1 and 2 show XRD patterns of pure, monodoped, co-doped and tri-doped TiO2 thin films. We see that the diffraction peaks of all thin films are ascribed to the peaks of TiO2 anatase phase. Diffraction peaks of all above samples are observed at around 25.4◦ , 37.8◦ , 48.2◦ , and 54.6◦ indexed as (101), (004) and (200) and (211) plane lattices, assigned to the anatase phase of TiO2 , respectively [Anatase XRD JCPDS Card no. 211272]. These observations indicate that mono-doping, co-doping and tri-doping with V, Cr and N have no effects on crystal structure of TiO2 . Furthermore, no characteristic peaks of vanadium oxide, chrominium oxide and nitrogen oxide can be found in XRD spectra. This could be due to the fact that the specific dopants could be considered to be fully incorporated into the TiO2 lattice or that formation of above dopant oxides is below Fig. 1. (Color online) XRD pattern of pure and monodoped TiO2 thin films. Fig. 2. (Color online) XRD pattern of pure, co-doped and tri-doped TiO2 thin films. the XRD detection limit. As shown in the inset of Fig. 1, it is clearly ob- Effect of Co-Doping and Tri-Doping with Transition Metals· · · – Hang Nguyen Thai Phung et al. -997- Fig. 3. (Color online) Uv-Visible absorption spectra of pure and mono-doped TiO2 thin films. Fig. 4. (Color online) Uv-Visible absorption spectra of pure, co-doped and tri-doped TiO2 thin films. served that compared to pure TiO2 , intensities of all mono-doped samples decrease. Furthermore, in comparision with pure TiO2 , TiO2 :Cr and TiO2 :V samples have the slight shifts of (101) anatase peak position to higher diffraction angle. This indicates that Cr and V ions could substitute for some Ti ions in crystal lattice of TiO2 . Because of the difference in atomic radii of the dopants (Cr: 166 pm, V: 171 pm) and Ti atom (176 pm), the lattice parameters of mono-doped TiO2 change slightly. In addition, (101) anatase peak position of TiO2 :N sample significantly shifts to lower diffraction angle. The reason for this could be the incorporation of N into TiO2 crystal lattice by some replacements O sites with N ions. As a atomic radius of N (≈ 56 pm) is larger than O atom (≈ 48 pm), compared to pure TiO2 , there is a change of lattice parameter of TiO2 :N. The inset of Fig. 2 also shows that intensity of all codoped and tri-doped TiO2 counterparts decrease a bit compared to pure TiO2 .(101) anatase diffraction peak position of TiO2 :(V,Cr) is nearly same as that of TiO2 :Cr and TiO2 :V samples. This could be explained by the fact that there isn’t a big difference in atomic radii of Cr, V, and Ti. Compared to TiO2 , the aforementioned peak positions of TiO2 :(V,N), TiO2 :(Cr,N), TiO2 :(V, Cr, N) samples shift toward a higher diffraction angle gradually. This could be the compensation for lattice distortion in co-doped and tri-doped samples. From that, the lattice distortion of tri-doped TiO2 :(V, Cr, N) sample is at least. The optical absroption spectra of mono-doped TiO2 samples are shown in Fig. 3. Compared to TiO2 , we observed a slight red-shift in the absorption edge of TiO2 :V, TiO2 :Cr counterparts, whereas that of TiO2 :N sample significantly extended into the visible light range. The reason for this could be that O ions can be replaced by N ions in the crystal lattice of TiO2 leading to the formation of some acceptor levels on the top of its valence band [11]. Because N2p and O2p states have closed energy levels, overlapping of these states makes TiO2 band-gap narrow. As a result, N-doped TiO2 can absorb visible light. Figure 4 displays the optical absorption spectra of codoped and tri-doped TiO2 samples. The absorption edge of TiO2 :(V, Cr) thin film shifts to lower energy at the wavelength of 400 nm, whereas that of TiO2 :(Cr, N), TiO2 :(V, N), and TiO2 :(V, Cr, N) samples are significantly extended into visible light region at the wavelength range of 450 nm to 500 nm. Before examining the change in the absorption edge as well as band-gap energy of the aforementioned samples in Figs. 7 and 8, we carried out X-ray photoelectron spectroscopy (XPS) measurements to investigate the valence states of all elements in TiO2 :(V,N) and TiO2 :(Cr,N) thin films. The results of XPS spectra of both samples were shown in Figs. 5 and 6. Figs. 5(a) and 6(a) show the high-resolution Ti 2p spectra of the aforementioned counterparts. Two Ti 2p peaks at around 462.3 eV and 468.2 eV were assigned to the 2p3/2 and 2p1/2 electronic states of Ti4+ , respectively [12,13]. The O 1s XPS spectra of both films were exhibited in Figs. 5(b) and 6(b) with the peaks at about 533.6 eV corresponding to an oxidation state of O2− [12, 13]. High-resolution V 2p and Cr 2p spectra of TiO2 :(V, N) and TiO2 :(Cr, N) samples are presented in Figs. 5(c) and 6(c), respectively. The V 2p3/2 and Cr 2p3/2 peaks at about 516.7 eV and 576.9 eV were assigned to V3+ and Cr3+ ions, respectively [14–16]. This indicates that V or Cr atoms substitute Ti sites in the TiO2 crystal lattice and exist in oxidation states of V3+ or Cr3+ . Figures 5(d) and 6(d) present XPS spectra of N 1s of the aforementioned sample and show that N 1s peaks appear at about 399.5 eV. The peak was generally considered for the presence of Ti-N bonds when the oxygen atoms were replaced by nitrogen atoms in the TiO2 crystal lattice -998- Journal of the Korean Physical Society, Vol. 70, No. 11, June 2017 Fig. 5. (Color online) Ti 2p (a), O 1s (b), V 2p3/2 (c) and N 1s (d) XPS spectra of TiO2 :(V, N) thin film. Fig. 6. (Color online) Ti 2p (a), O 1s (b), Cr2p3/2 (c) and N 1s (d) XPS spectra of TiO2 :(Cr, N) thin film. [11,17]. From these results, we concluded that V and Cr ions exist in these samples as V3+ and Cr3+ cases. Figure 7 shows the proposed mechanism for changing band-gap energy of mono-doped TiO2 . For pure TiO2 , valence band consists of O2p states, and conduction band formed mainly by Ti3d states. For TiO2 :N, when some N ions can replace some O sites in TiO2 lattice, N2p states create impurity energy levels located above the top of valence band. Therefore, there is a overlap between O2p states and N2p states which lower band-gap energy of TiO2 . N2p states is located at about 0.3 eV above the valence band edge of TiO2 , so TiO2 :N can absorb visible light [18]. For TiO2 :V or TiO2 :Cr samples, the dopant ions can substitute for Ti sites in TiO2 lattice which forms some impurity states below conduction band edge of TiO2 . Fig. 7. (Color online) Proposed mechanism for changing the band-gap energy of mono-doped TiO2 . Fig. 8. (Color online) Proposed mechanism for changing band-gap energy of co-doped and tri-doped TiO2 . These states act as shallow traps which not only narrows the band-gap and move the absorption edge into visible light region (as can be seen at Fig. 8) but also trap the photo-generated electrons and inhibit the recombination of electron and hole pairs. The lower the recombination rate, the higher the photocatalytic activity (as can be discussed later at Fig. 9). The effects of N and transition metal ions on the change of TiO2 band-gap were displayed in Fig. 8. The co-doping with a transition metal and N makes the bandgap narrow due to the hybridization of O2p states and N2p states. The change in band-gap energy of tri-doped TiO2 is similar to that of co-doped TiO2 The MB decomposition of pure, mono-doped, codoped and tri-doped TiO2 samples was carried out in visible light irradiation as shown in Figs. 9 and 10. The results indicate that doped TiO2 photocatalysts move Effect of Co-Doping and Tri-Doping with Transition Metals· · · – Hang Nguyen Thai Phung et al. Fig. 9. (Color online) Photocatalytic degradation of MB over pure and mono-doped TiO2 thin films under visible light irradiation. the photocatalytic activity into the visible light region. From Fig. 9, all mono-doped TiO2 samples have better photocatalytic activities than that of TiO2 . It takes only 150 minutes to decompose about 45% of MB solution under visible-light irradiation. In addition, Fig. 9 also shows that photocatalytic performance of TiO2 : V is slightly higher than that of TiO2 : Cr. Q. Ming et al. reported that V could be easily incorporated into the TiO2 lattice by replacing Ti because its substitutional energy is less than 1 eV [19]. In the meanwhile, their substitutional energy of Cr is slightly higher (at about 4 eV) than that of V, so Cr ions are more difficult to substitute Ti4+ ions in TiO2 matrix than V ions [19]. Therefore, with a same amount of the dopant, number of V atoms can incorporated into TiO2 lattice more than that of Cr atoms. As a result, TiO2 :V exhibits higher MB decomposition rate than TiO2 :Cr. On the other hand, charge imbalance from valency difference between Ti and transition metal can be compensated by the increased generation of surface hydroxyl groups or the reduction of oxygen vacancies, which effectively retards the recombination of photo-generated electrons and holes. In the case of V, V3+ states located at about 0.43 eV below the bottom of conduction band (Fig. 7), which can reduce the recombination rate of charge carriers and thus improve the photocatalytic activity of pure TiO2 [4]. In the case of Cr, Cr3+ states located at about 0.9 eV above the top of valence band. This is the localized level of t2g of Cr, which is lies in the middle of TiO2 band-gap [4] (Fig. 7). When TiO2 :Cr absorbs visible light, an electron is excited from valence band to t2g level of Cr or from t2g level of Cr to conduction band [20]. As a result, it exits a large amount of electron-hole pairs. From Fig. 10, all co-doped TiO2 thin films show higher photocatalytic activities than that of mono-doped -999- Fig. 10. (Color online) Photocatalytic degradation of MB over pure, co-doped and tri-doped TiO2 thin films under visible light irradiation. samples. TiO2 :(V, Cr), TiO2 :(Cr, N) and TiO2 :(V, N) samples decomposed about 55%, 85% and 89% of MB solution under visible-light irradiation after 150 mininutes, respectively. The tri-doped TiO2 thin film exhibits the highest photocatalytic activity with MB photodegradation rate about 96% after 150 minutes under visible light irradiation. The impurity levels of transition metal ions and N in the band-gap can trap photo-generated electrons and holes, respectively, which improves the utilization of electron-hole pairs efficiently. Therefore, it found that a larger quantity of electrons and holes are in co-doped TiO2 than pure and mono-doped TiO2 . Hence, co-doped TiO2 samples exhibit higher visiblephototcatalytic acitivites than that of pure and monodoped TiO2 photocatalysts. The tri-doped TiO2 thin film has the best photocatalytic activity under visible light irradiation. The first reason for this is same as that of co-doped TiO2 samples. Another reason for this can be the synergistic effect of two transition metals. The synergistic effect of V and Cr can induce a change of surface electronic states, resulting in the promotion of visible-photoinduced carriers and introduce new states around the bottom of conduction band [21]. III. CONCLUSION In this study, mono, co and tri-doped TiO2 with V, Cr and N thin films have been synthesized by sol-gel method. All doped TiO2 photocatalysts have higher photocatalyticactivities than pure TiO2 under visible light irradiation. Co-doped TiO2 samples show higher photocatalytic performances than mono-doped TiO2 while tri-doped TiO2 counterpart has the best visible-photocatalytic property. Compared to pure TiO2 , -1000- Journal of the Korean Physical Society, Vol. 70, No. 11, June 2017 the band-gap energies of series of doped TiO2 thin films are smaller by the formation of N2p levels near the top of the valence band, which is responsible for the enhancement of visible-photocatalytic activity. Furthermore, t2g levels of Cr lie in the middle of the band-gap creates a large amount of electron-hole pairs under visible light irradiation. The increasing of visible-photocatalytic activity of TiO2 :(V, Cr, N) is explained by V impurity states and a synergistic effect of V and Cr. Shallow impurity levels of V inhibit the recombination of electron and hole pairs and synergistic effect of V and Cr forms new states near the conduction band egde. REFERENCES [1] M. A. Shannon, P. W. Bohn, M. Elimelech, J. G. Georgiadis, B. J. Mariñas and A.M. Mayes, Nature 452, 301 (2008). [2] S. Malato, P. Fernández-Ibáñez, M.I. Maldonado, J. Blanco, and W. Gernjak, Catalysis Today 147, 1 (2009). [3] M. N. Chong, B. Jin, C. W. K. Chow, C. Saint, Water research 44, 2997 (2010). [4] W. Choi, A. Termin and M. R. Hoffmann, J. Phys. Chem. 98, 13669 (1994). [5] Y-H. Peng, G-F Huang and W-Q. Huang, Advanced Powder Technology 23, 8 (2012). [6] A. Hajjaji, A. Atyaoui, K. Trabelsi, M. Amlouk, L. Bousselmi, B. Bessais, M. A. E. Khakani and M. Gaidi, American Journal of Analytical Chemistry 5, 473 (2014). [7] J. C-S. Wu, C-H. Chen, Journal of Photochemistry and Photobiology A: Chemistry 163, 509 (2004). [8] J. M. Herrmann, J. Disdier and P. Pichat, Chem. Phys. Lett. 108, 618 (1984). [9] D. Lu, M. Zhang, Z. Zhang, Q. Li, X. Wang and J. Yang, Nanoscale Research Letters 9, 272 (2014). [10] M. E. Kurtoglu, T. Longenbach, K. Sohlberg and Y. Gogotsi, J. Phys. Chem. C 115, 17392 (2011). [11] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga, Science 293, 269 (2001). [12] B. Viswanathan and K. R. Krishanmurthy, Int. J. of Photoenergy, Article ID 269654, 10 pages (2012). [13] D. Morris, R. Dixon, F. H. Jones, Y. Dou et al., Phys. Rev. B 55, 16083 (1997). [14] H. Zhao, S. Bennici, J. Shen and A. Auroux, Appl. Catal. A 356, 121 (2009). [15] G. Silversmit and D. Depla, J. of Electron Spectroscopy and Related Phenomena 135, 167 (2004). [16] Z. Jiao, T. Chen, J. Xiong, T. Wang, G. Lu, J. Ye and Y. Bi, Scientific Reports 3, 2720 (2013). [17] X. B. Chen and C. Burda, J. Phys. Chem. B 108, 15446 (2004). [18] J. Xu, Y. Ao, M. Chen and D. Fu, Applied Surface Science 256, 4397 (2010). [19] Q. Meng, T. Wang, E. Liu, X. Ma, Q. Ge and J. Gong, Phys. Chem. Chem. Physics 15, 9549 (2013). [20] T. Umebayashi, T. Yanaki, H. Itoh and K. Asai, J. Phys. and Chem. of solids 63, 1909 (2002). [21] F. Strobel and D. P. Ridge, Inorg. Chem. 27, 891 (1988).