Theoretical studies of heteroatom-doping in TiO₂ to enhance the electron injection in dye-sensitized solar cells

Abstract

Density functional calculations have been explored to analyze the structural, electronic and charge transfer properties of a doped TiO₂ substrate and catechol–TiO₂ interfaces for dye-sensitized solar cells. The results demonstrate that the dopant W⁶⁺ moves the CB (conduction band) edge downward and introduces 5d-unoccupied orbitals located in the CB bottom of the TiO₂. On the other hand, the dopant Zn²⁺ shifts the CB edge upward with an insertion of 3d-occupied orbitals into the VB (valence band). In catechol–TiO₂ systems, W⁶⁺ enlarges the energy difference between the LUMO of catechol and LUMO of TiO₂, which enlarges the driving force for electron injection in turn and results in an increased short circuit current (J_sc). Our modeling injection dynamics and quantitative Bader analysis of the interfacial charge transfer have revealed that the catechol–TiO₂ doped with W⁶⁺ provides faster and more electron injection for dye sensitized solar cells (DSSCs). While the Zn²⁺ doped system exhibits lower electron efficiency due to the minimized energy difference between the catechol LUMO and TiO₂ CB.

---

1 Introduction

Titanium dioxide (TiO₂) as a functional material has attracted considerable attention in recent decades because of its comprehensive combination of optical and electrical properties and important applications in photovoltaics^1–4 and photocatalysis.^5–8 In fact, the photovoltaic and photocatalysis performance of TiO₂ depends on electronic structure, interfacial properties and charge transport efficiency which are mostly determined by the crystal structure.⁹ Introducing doping metal and non-metal atoms in TiO₂ have been proven as an efficient technology to achieve a higher photoelectric conversion^10–13 and photocatalytic efficiency.^14–17 In particular, doped TiO₂ materials have been recognized as promising candidates due to their size-tunable band gap and easy syntheses for photovoltaics and photocatalysis.

In the last two decades, one of the greatest achievements for clean energy applications is the dye-sensitized solar cells (DSSCs).¹⁸ DSSCs have been considering as an attractive candidate due to their low cost and high efficiency. In DSSC devices, TiO₂ photoanode acts as the electron acceptor in the electron injection process, and also provides direct or tortuous path for electron transport in the electrical circuits.¹⁹ Intensive researches indicate that the driving force for electron injection is generated from the energy difference between the excited states of the dye and the CB (conduction band) edge of TiO₂.^20,21 For single organic molecule, the steady LUMO (Lowest Unoccupied Molecular Orbital) can approximately express the excited states of the dye.^22–24 So tuning of CB edge downward with dopants can enlarge the driving force and results more efficient electron injection,^25–27 which is the key factor for the short-circuit photocurrent density (J_sc) and incident photon-to-electron conversion efficiency (IPCE).²⁸ Dopants also play an important role in affecting the properties of electrical conductivity and surface adsorption of TiO₂.^29–32 Therefore, introduction of dopants in TiO₂ effects the electron injection, the electron transport in the TiO₂ electrode and interface recombination between the injected electron and the oxidized redox mediator, I₃⁻ ions.³³

In order to achieve high-efficiency DSSCs, various types of dopants have been used for the optimization of TiO₂. Many studies have shown that doping nonmetal anions at O sites (e.g. B,^34,35 C,³⁶ N,^37,38 S,³⁹ and I⁴⁰) induced additional electronic states above the valence band (VB) upper edge and scarcely had effect on CB near the bottom edge.^41–43 On the other hand, doping transition metal ions at Ti sites (such as Fe²⁺,⁴⁴ Ce⁴⁺,⁴⁵ Zr⁴⁺,⁴⁶ V⁴⁺ or V⁵⁺,⁴⁷ Nb⁵⁺,⁴⁸ Ta⁵⁺ (ref. 49) and W⁶⁺ (ref. 22)) is also be an effective method to modify the electronic properties of TiO₂. In a TiO₂ system, donor and acceptor-type dopants play different roles in terms of geometry and electronic structure, which affect the device performance in turn. Both of donor and acceptor-type transition metal ions dopants are thought to be very important for DSSC and have been extensively studied. Tungsten doping, as a donor-type dopant, was first reported by Ko et al.;¹⁰ then Wang and coworkers²² found that the CB edge of TiO₂ shifts downward gradually with increasing the tungsten (W⁶⁺) content. Additionally, as an acceptor-type dopant, Zn-doped⁵⁰ TiO₂ enhanced the charge transfer under low-intensity illumination. According to previous studies, there are numbers of researches have reported that the electron injection process can be influenced by different dopants. The detailed mechanism of these influences remains to be elucidated.

In this work, an acceptor (Zn) and a donor dopant (W) in TiO₂ are studied to obtain insight into the mechanism by diverse dopants which affect the electron injection. We have performed a systematic analysis to investigate their effects on geometry and electronic structure, such as CB levels, CB component, density of states, and charge transfer. We have plotted the quantitative Bader analysis and injection dynamics to explore the charge transfer between catechol and the TiO₂ surface. The acceptor (Zn) and donor (W) dopant play different roles to control the electron injection process in DSSCs and the corresponding results should provide valuable guidelines for experimental works.

2 Computational details and modeling

It is well known that anatase nanoparticles appear higher photovoltaic and photocatalytic activities compared with other titania polymorphs.^51–53 TiO₂ (101) surface dominates more than 94% of the external surface in anatase crystal⁵⁴ and anatase (101) is the most well-known model describing the TiO₂ substrate in DSSC.^55–57 To assess the effects of heteroatom-doped TiO₂ electrodes on DSSC device performance, a large anatase (101) surface was constructed with a 15 Å vacuum layer as shown in Fig. 1 (10.21 Å × 11.33 Å, (TiO₂)24, 72 atoms). Plane wave DFT calculations on the slab are performed in the Vienna Ab initio Simulation Package (VASP) code,^58–60 using the projector-augmented plane-wave (PAW) pseudopotentials^61,62 and generalized gradient approximation (GGA) of PBE.^63,64 The cutoff energy is set to 600 eV. Geometries are optimized till the energy and force converged within fair limits (1 × 10⁻⁴ eV and 1 × 10⁻³ eV Å⁻¹ respectively). Density of states (DOS) and the calculated scanning tunneling microscopes (STM) are obtained based on the optimized geometry. The brillouin zone of the surface is sampled using an 8 × 8 × 1 Monkhorst–Pack special k-point mesh to obtain and understand qualitative features on the geometrical and electronic structures of TiO₂ electrode.

Fig. 1 Supercell slab model of TiO₂ (101) surface. (A) Top view (B) side view (purple marked substituted Ti position).

Furthermore, a two-dimensional slab model of sensitized anatase (101) surface is built to describe the charge transfer between dye and the heteroatom-doped TiO₂ electrodes (the slab has dimensions of 10.21 Å × 11.33 Å with a 20 Å vacuum layer). The same pseudopotential and function are utilized to calculate the electronic properties of dye/TiO₂ composite.

3 Results and discussion

3.1 Geometry and electronic properties of anatase (101)

As shown in Fig. 1, heteroatom-doping of TiO₂ substrate has been implemented by substituting the surface 5-fold coordinated Ti (the purple position) with a W⁶⁺ or Zn²⁺ ion. The dopants in semiconductor oxides bring about serial changes in the geometrical and electronic structure. Based on the geometrical analysis, neither of W⁶⁺ and Zn²⁺ remains in the doped Ti site (Ti⁴⁺) due to non-equilibrium charging. At the electron configuration of 3d,¹⁰ Zn²⁺ changes the interatomic spacing and displays stretched distances with surrounding oxygen atoms. In addition, we have noticed that the Zn atoms try to move away from the surface position. In contrast, W⁶⁺ makes coordinate covalent bonds shorter than the corresponding Ti–O bonds because of the greater capacity to attract the electrons. As compared to doped Ti position of the pure TiO₂ electrode, W⁶⁺ and Zn²⁺ both move slightly along the [101] vector (z axis direction). The results demonstrate that z-coordinates are in the order of Zn²⁺ (7.15 Å) > Ti⁴⁺ (6.94 Å) > W⁶⁺ (6.88 Å) in the same coordinate system as shown in Fig. 2.

Fig. 2 Structures of the pure and heteroatom-doped TiO₂ electrodes.

We have also investigated the effect of W⁶⁺ and Zn²⁺ doping on the electronic properties. The scanning tunneling microscopes (STM) images are generated based on the calculated charge density with a maximum z of 2.0 Å above the highest atom.⁶⁵ Contours added to STM images express the main features and different electronic structures of three TiO₂ super cells. For the STM images, F = −1.40 V is used. The calculated STM images of filled states are illustrated in Fig. 3 and some interesting features are also appeared in this image. The STM image of pure TiO₂ for a 1.4 V below the Fermi level provides a simple picture; light and shade are staggered periodically. In the panel (d) of Fig. 3, it is obviously shown that the egg-like shapes of the top layer are brighter than the second layer positions. There are distinct differences between the doped semiconductor with W⁶⁺ and the pure TiO₂. The atom W located at second layer of the slab causes some interesting features in the STM image. A careful look at the panel (e) reveals that darker and slightly smaller patterns of oxygen atoms surrounded the W⁶⁺ substituted site. However, the image still shows nicely pure TiO₂ features for the O atoms far from the W⁶⁺ site over three bonds. Taking a careful look on contour plots, it is can be seen that W⁶⁺ dopant reduced the charge density of the adjacent oxygen atoms. The STM image of Zn²⁺ doped TiO₂ shows the eye-catching bright ball of a top-layer O atom and dark gray oval shapes. The constant density surface, showed by the notch circle in the panel (f), presents reduced charge density except the top O atom adjacent to Zn atom.

Fig. 3 Atomic structures and calculated STM images of TiO₂ (101) surfaces. Panels (a)–(c) represent the top views; (d)–(f) represent the STM images. Colored rounds label the atomic position: blue and fuchsia regions represent Ti and O atom positions in the two top layers of the surface, while cyan and green disks indicate the W and Zn atoms positions.

Furthermore, the DOS have been provided in Fig. 4 for an indepth understanding of the interaction between W⁶⁺/Zn²⁺ dopants and electronic properties. The DOS presents separated CB and VB, which are mainly composed of titanium 3d and oxygen 2p orbitals. Compared to the DOS of pure TiO₂, W⁶⁺ doping in TiO₂ moves the CB edge downward, which in turns enlarges the electron injection driving force and increases the J_sc. Whereas, Zn²⁺ shifts the CB and VB edge upward. The filled lines presented the d orbitals PDOS of W, Zn and corresponding site Ti atoms. It is worth noting that W⁶⁺ introduces some 5d-unoccupied orbitals located in the bottom of the TiO₂ CB, while dopant Zn²⁺, with the configuration of 3d,¹⁰ inserts 3d-occupied orbitals into the VB and the bottom edge of CB remains unchanged. In the ultrafast electron injection, excited electrons are injected into the TiO₂ CB; therefore, the CB component affects the electron injection distribution directly. Additionally, W⁶⁺ shifts the Fermi level negatively, which induces a better charge separation and would result in an enhancement of injection process.

Fig. 4 Calculated DOS (density of states) of TiO₂ (101) substrates. In the right panels, DOS have been zoomed in on (the black line represents the total DOS; red and magenta lines represent the PDOS of titanium 3d and oxygen 2p orbitals; the filled cyan, green and blue lines represent the PDOS of W, Zn and corresponding site Ti d orbitals respectively in three slab models).

3.2 The effect of W⁶⁺/Zn²⁺ dopants on the catechol–TiO₂ systems

3.2.1 The effect of W⁶⁺/Zn²⁺ dopants on electronic structures of catechol–TiO₂ systems. Most researches have focused on the exploration of electron injection mechanisms by employing simple dye.^19,66–69 Catechol (C₆H₆O₂) has been classified as representative examples of type-II dye, which follows a direct electron injection mechanism.¹⁹ As catechol–TiO₂ interface can provide a clear and intuitive display of electron injection distribution, therefore, electronic structures of diverse catechol–TiO₂ were plotted out to analyze the effect of dopant W⁶⁺/Zn²⁺ on the electron injection process.

The density of states are plotted to explore the interactions between catechol and the TiO₂. Compared to the bare TiO₂, the PDOS of TiO₂ adsorbed by catechol indicates a significant upward shift of the CB and VB edges. In Fig. 5, it can be clearly seen that there are strong overlap of the catechol with TiO₂. After adsorption, catechol introduces unoccupied molecular bands in the higher energy of conduction band. Zn²⁺ doped catechol–TiO₂ system shows a significantly overlap near the bottom edge of the conduction band, regarding PDOS of catechol. This minimizes the energy difference between the catechol LUMO and TiO₂ CB, which will induce a slower electron injection process compared with pure catechol–TiO₂ system. Wang and coworkers²² have reported the open-circuit photovoltage (V_oc) remained almost unchanged with less 0.5% W-doping due to the collective effect of the downward shift of the CB and the enhanced electron lifetime. As in Fig. 5, W⁶⁺ shifts the PDOS of catechol away from the CB edge, which enlarges the energy difference between the catechol LUMO and TiO₂ CB. That would increase the driving force of electron injection and the J_sc in the DSSCs. The highest power conversion efficiency (η) was obtained in the experiments, with increasing by 17% in J_sc and by 20% in η.²²

Fig. 5 Calculated DOS of catechol–TiO₂ systems (the black line represents the total DOS; the filled cyan, red and green lines represent the PDOS of catechol in W⁶⁺ doped, pure and Zn²⁺ doped slab models respectively).

The orbital distributions of valence band maximum (VBM) and conduction band minimum (CBM) at the Gamma point are illustrated in Fig. 6. The VBM maps are mostly localized on the catechol, which display diffusion to the adsorption sites at the catechol–TiO₂ interfaces. While the CBM are localized in the TiO₂ moiety, showing Ti-3d and W-5d orbitals characteristics; electron injection process will benefit. Moreover, the orbital distributions differ significantly in CBM as in panel (d), the CBM distribution for pure catechol–TiO₂ mainly resides on the Ti 3d orbitals. With W⁶⁺ doping, as in panel (e), the distribution of CBM concentrates in the region near the W⁶⁺ doping site of TiO₂. In panel (f), Zn does not contribute any atomic orbitals to the CBM; the distribution of CBM mainly localized on the bottom layer of TiO₂ moiety. In the electron injection process, the TiO₂ electrode accepts electrons in conduction band from the excited dyes. It reveals that the effect of W⁶⁺/Zn²⁺ dopants on CBM distributions of catechol–TiO₂ have significant impact on the electron injection process.

Fig. 6 Selected molecular orbitals of three catechol–TiO₂ systems. Left (panels (a)–(c)) and right ((d)–(f)) are the orbital distributions of the VBM and CBM for the Gamma point. (i)–(iii) are the geometries of three catechol–TiO₂ systems.

3.2.2 The effect of W⁶⁺/Zn²⁺ dopants on electron injection through the catechol–TiO₂ interface. Dopants W⁶⁺ and Zn²⁺ differ the distributions and energy levels of CBM significantly, and influence the electron injection process and electron injection distribution further. Based on the Bader charge analysis between catechol and TiO₂ electrodes, an increase in the TiO₂ electrodes charges is observed. In the pure catechol–TiO₂ system, there is 0.934e from the catechol fragment to TiO₂. In case of W⁶⁺ doping, 1.366e charge transfer into W⁶⁺ doped TiO₂ has been observed; conversely, Zn²⁺ doping decreases the charge transfer with 0.443e. The difference in electronic density Δρ is presented in Fig. 7, which is estimated as follows:

Δρ = ρ_total − ρ_catechol − ρ_TiO2

where ρ_total is the total electron density of catechol–TiO₂ system, ρ_catechol and ρ_TiO2 are the electron density of isolated catechol and TiO₂ fragment, respectively. As shown in Fig. 7, the fuchsia areas show where the electron density has been enriched; conversely, the blue regions show where the density has been depleted.

Fig. 7 The diffrence in electronic density between catechol and the TiO₂ (the fuchsia and blue areas refer to the increased and decreased electron density).

In addition, dopants W⁶⁺ and Zn²⁺ also influence the dynamics of electron injection. The interfacial electron transfer has been simulated in the IET (Interfacial Electron Transfer) code;^69,70 the further time-dependent survival probability in three different catechol sensitized TiO₂ are provided in Fig. 8. It exhibits fully consistent behavior with the static charge transfer. Compared to pure catechol–TiO₂ system, W⁶⁺ doping provides faster and more electron injection, in addition to lower recombination. On the one hand doping W⁶⁺ results in a more efficient electron injection, but on the other hand it makes the back electron transfer easier. Conversely, Zn²⁺ doping exhibits a slower electron transfer and more incomplete electron injection.

Fig. 8 The time-dependent survival probability in three different catechol sensitized TiO₂. The dashed lines correspond to elementary steps and the full lines are the exponential fitting (black, blue and green lines correspond to pure, W⁶⁺ and Zn²⁺ doped catechol–TiO₂ systems, respectively).

4 Conclusions

In this research, we report a theoretical investigation of heteroatom-doped TiO₂ electrode by modeling the TiO₂ photoanode and catechol–TiO₂ interface for DSSCs. The structural, electronic, and charge transfer properties of the catechol–TiO₂ interfaces have been considered for electron injection process. According to our study, the dopant W⁶⁺ may benefit the electron injection process by moving the CB edge downward and introducing 5d-unoccupied orbitals located in the CB bottom of the TiO₂. However, dopant Zn²⁺ shifts the CB edge upward with an insertion of 3d-occupied orbitals into the VB; while the bottom edge of CB remains unchanged. The quantitative analysis of charge transfer based on Bader analysis has exposed the critical electronic features underlying interfacial properties and interfacial electron transfers. The calculation reveals strong charge transfer between catechol and pure TiO₂ (0.934e). W⁶⁺ has significantly increased the charge transfer toward the photoanode (1.366e), while doping Zn²⁺ may cause the lowest charge transfer toward the adsorbate catechol with only 0.443e. Furthermore, photo-induced charge transfer dynamics have also been described. It has indicated that the donor dopant W⁶⁺ accelerate the electron injection from dye to semiconductor, while the acceptor dopant Zn²⁺ slower the injection. Hence, it shows that diverse dopants can affect the electron injection process and provide valuable information. We thus envision that this theoretical investigation can provide important insights into the future design of more efficient materials for DSSCs.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 21173096 and 21303068) and the State Key Development Program for Basic Research of China (Grant No.2013CB834801), and Young Scholar Training Program of Jilin University. The wavefunction plots are drawn with the help of the vaspmo program developed by Yang Wang. We also acknowledge Prof. Luis GC Rego for providing a copy of the IET code.

References

1. M. Gratzel, Nature, 2001, 414, 338–344.
2. M. Gratzel, J. Photochem. Photobiol., C, 2003, 4, 145–153.
3. A. Hagfeldt and M. Gratzel, Acc. Chem. Res., 2000, 33, 269–277.
4. F. de Angelis, Acc. Chem. Res., 2014, 47, 3349–3360.
5. A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
6. A. Fujishima, X. Zhang and D. A. Tryk, Surf. Sci. Rep., 2008, 63, 515–582.
7. F. Zuo, K. Bozhilov, R. J. Dillon, L. Wang, P. Smith, X. Zhao, C. Bardeen and P. Feng, Angew. Chem., Int. Ed., 2012, 51, 6223–6226.
8. J. Schneider, M. Matsuoka, M. Takeuchi, J. Zhang, Y. Horiuchi, M. Anpo and D. W. Bahnemann, Chem. Rev., 2014, 114, 9919–9986.
9. G. Liu, H. G. Yang, J. Pan, Y. Q. Yang, G. Q. Lu and H.-M. Cheng, Chem. Rev., 2014, 114, 9559–9612.
10. K. H. Ko, Y. C. Lee and Y. J. Jung, J. Colloid Interface Sci., 2005, 283, 482–487.
11. X. Lu, X. Mou, J. Wu, D. Zhang, L. Zhang, F. Huang, F. Xu and S. Huang, Adv. Funct. Mater., 2010, 20, 509–515.
12. Y.-C. Nah, I. Paramasivam and P. Schmuki, ChemPhysChem, 2010, 11, 2698–2713.
13. K. Lee, A. Mazare and P. Schmuki, Chem. Rev., 2014, 114, 9385–9454.
14. H. Irie, Y. Watanabe and K. Hashimoto, Chem. Lett., 2003, 32, 772–773.
15. T. Ohno, T. Mitsui and M. Matsumura, Chem. Lett., 2003, 32, 364–365.
16. S. In, A. Orlov, R. Berg, F. García, S. Pedrosa-Jimenez, M. S. Tikhov, D. S. Wright and R. M. Lambert, J. Am. Chem. Soc., 2007, 129, 13790–13791.
17. L. Sang, Y. Zhao and C. Burda, Chem. Rev., 2014, 114, 9283–9318.
18. B. O'Regan and M. Gratzel, Nature, 1991, 353, 737–740.
19. W. R. Duncan and O. V. Prezhdo, Annu. Rev. Phys. Chem., 2007, 58, 143–184.
20. A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pettersson, Chem. Rev., 2010, 110, 6595–6663.
21. D. Cahen, G. Hodes, M. Gratzel, J. F. Guillemoles and I. Riess, J. Phys. Chem. B, 2000, 104, 2053–2059.
22. X. Zhang, F. Liu, Q.-L. Huang, G. Zhou and Z.-S. Wang, J. Phys. Chem. C, 2011, 115, 12665–12671.
23. K.-P. Wang and H. Teng, Phys. Chem. Chem. Phys., 2009, 11, 9489.
24. X. Ren, Q. Feng, G. Zhou, C.-H. Huang and Z.-S. Wang, J. Phys. Chem. C, 2010, 114, 7190–7195.
25. C. Klein, M. K. Nazeeruddin, P. Liska, D. Di Censo, N. Hirata, E. Palomares, J. R. Durrant and M. Gratzel, Inorg. Chem., 2005, 44, 178–180.
26. Z.-S. Wang, Y. Cui, Y. Dan-oh, C. Kasada, A. Shinpo and K. Hara, J. Phys. Chem. C, 2008, 112, 17011–17017.
27. G. Zhou, N. Pschirer, J. C. Schöneboom, F. Eickemeyer, M. Baumgarten and K. Müllen, Chem. Mater., 2008, 20, 1808–1815.
28. M. Gratzel, J. Photochem. Photobiol., C, 2003, 4, 145–153.
29. K. Suriye, P. Praserthdam and B. Jongsomjit, Appl. Surf. Sci., 2007, 253, 3849–3855.
30. F. Dong, S. Guo, H. Wang, X. Li and Z. Wu, J. Phys. Chem. C, 2011, 115, 13285–13292.
31. A. Zaleska, Recent Pat. Eng., 2008, 2, 157–164.
32. Z. Zhao, Z. Li and Z. Zou, J. Phys. Chem. C, 2013, 117, 6172–6184.
33. A. J. Frank, N. Kopidakis and J. v. d. Lagemaat, Coord. Chem. Rev., 2004, 248, 1165–1179.
34. H. Tian, L. Hu, C. Zhang, S. Chen, J. Sheng, L. Mo, W. Liu and S. Dai, J. Mater. Chem., 2011, 21, 863–868.
35. A. Subramanian and H.-W. Wang, Appl. Surf. Sci., 2012, 258, 6479–6484.
36. R. Taziwa, E. L. Meyer, E. Sideras-Haddad, R. M. Erasmus, E. Manikandan and B. W. Mwakikunga, Int. J. Photoenergy, 2012, 2012(904321–904329), 904323.
37. T. Ma, M. Akiyama, E. Abe and I. Imai, Nano Lett., 2005, 5, 2543–2547.
38. M. Motlak, M. S. Akhtar, N. A. M. Barakat, A. M. Hamza, O. B. Yang and H. Y. Kim, Electrochim. Acta, 2014, 115, 493–498.
39. Q. Sun, J. Zhang, P. Wang, J. Zheng, X. Zhang, Y. Cui, J. Feng and Y. Zhu, J. Renewable Sustainable Energy, 2012, 4.
40. Q. Hou, Y. Zheng, J.-F. Chen, W. Zhou, J. Deng and X. Tao, J. Mater. Chem., 2011, 21, 3877–3883.
41. X. Chen, C. Burda, J. Guo, K. E. Smith, P.-A. Glans and T. Learmonth, Arabian J. Sci. Eng., Sect. A, 2010, 35, 65–71.
42. X. Chen, P.-A. Glans, X. Qiu, S. Dayal, W. D. Jennings, K. E. Smith, C. Burda and J. Guo, J. Electron Spectrosc. Relat. Phenom., 2008, 162, 67–73.
43. N. Serpone, J. Phys. Chem. B, 2006, 110, 24287–24293.
44. L. Sang, J. L. Gole, J. Wang, J. Brauer, B. Mao, S. M. Prokes and C. Burda, J. Phys. Chem. C, 2013, 117, 15287–15294.
45. J. Zhang, W. Peng, Z. Chen, H. Chen and L. Han, J. Phys. Chem. C, 2012, 116, 19182–19190.
46. M. Durr, S. Rosselli, A. Yasuda and G. Nelles, J. Phys. Chem. B, 2006, 110, 21899–21902.
47. K. Bhattacharyya, S. Varma, A. K. Tripathi, S. R. Bharadwaj and A. K. Tyagi, J. Phys. Chem. C, 2008, 112, 19102–19112.
48. A. K. Chandiran, F. d. r. Sauvage, M. Casas-Cabanas, P. Comte, S. M. Zakeeruddin and M. Graetzel, J. Phys. Chem. C, 2010, 114, 15849–15856.
49. J. Liu, H. Yang, W. Tan, X. Zhou and Y. Lin, Electrochim. Acta, 2010, 56, 396–400.
50. K.-P. Wang and H. Teng, Phys. Chem. Chem. Phys., 2009, 11, 9489–9496.
51. D. C. Hurum, A. G. Agrios, K. A. Gray, T. Rajh and M. C. Thurnauer, J. Phys. Chem. B, 2003, 107, 4545–4549.
52. K. Woan, G. Pyrgiotakis and W. Sigmund, Adv. Mater., 2009, 21, 2233–2239.
53. B.-M. Kim, S.-G. Rho and C.-H. Kang, J. Nanosci. Nanotechnol., 2011, 11, 1515–1517.
54. M. Lazzeri, A. Vittadini and A. Selloni, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 63, 155409.
55. L.-M. Liu, S.-C. Li, H. Cheng, U. Diebold and A. Selloni, J. Am. Chem. Soc., 2011, 133, 7816–7823.
56. F. Schiffmann, J. VandeVondele, J. Hutter, R. Wirz, A. Urakawa and A. Baiker, J. Phys. Chem. C, 2010, 114, 8398–8404.
57. J. Yu, J. Fan and K. Lv, Nanoscale, 2010, 2, 2144–2149.
58. G. Kresse and J. Hafner, Phys. Rev. B: Condens. Matter Mater. Phys., 1993, 47, 558–561.
59. G. Kresse and J. Furthmüller, Phys. Rev. B: Condens. Matter Mater. Phys., 1996, 54, 11169–11186.
60. G. Kresse and J. Furthmuller, Comput. Mater. Sci., 1996, 6, 15–50.
61. P. E. Blochl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 17953–17979.
62. G. Kresse and D. Joubert, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59, 1758–1775.
63. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
64. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1997, 78, 1396.
65. D. E. P. Vanpoucke and G. Brocks, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 77, 241308.
66. L.-M. Liu, S.-C. Li, H. Cheng, U. Diebold and A. Selloni, J. Am. Chem. Soc., 2011, 133, 7816–7823.
67. R. Sanchez-de-Armas, M. A. San-Miguel, J. Oviedo and J. F. Sanz, Comput. Theor. Chem., 2011, 975, 99–105.
68. R. O. Sanchez-de-Armas, J. Oviedo, M. A. N. San Miguel and J. F. Sanz, J. Phys. Chem. C, 2011, 115, 11293–11301.
69. L. G. C. Rego and V. S. Batista, J. Am. Chem. Soc., 2003, 125, 7989–7997.
70. W. Li, L. G. C. Rego, F.-Q. Bai, J. Wang, R. Jia, L.-M. Xie and H.-X. Zhang, J. Phys. Chem. Lett., 2014, 5, 3992–3999.

---