Study on Titanium Dioxide Nanocrystals with Specific Crystal Facet on Surface for High Performance Photocatalyst and Dye-Sensitized Solar Cells

Transcription

1 Study on Titanium Dioxide Nanocrystals with Specific Crystal Facet on Surface for High Performance Photocatalyst and Dye-Sensitized Solar Cells Changdong Chen Kagawa University Japan December 2014

2 Contents CHAPTER I General Introduction Crystal Structures and Physical Properties of Titanium Dioxide Polymorphs Photocatalytic Reactions of Titanium Dioxides and Their Applications Photocatalytic Reactions Mechanism Visible Light Active Photocatalysis of TiO Application of Photocatalysis to Decomposition of H 2 O and Organic Pollution Dye-Sensitized Solar Cells Structure and Principle of Dye-Sensitized Solar Cell Materials for Dye-Sensitized Solar Cell Performance Characteristic Parameters and Equivalent Circuit of DSSC Titanium Dioxide Nanocrystals for Dye-Sensitized Solar Cell Synthesis of Anatase Titanium Dioxides with Specific Facet on Surface and Their Photocatalytic Activities Surface Structures of {101}, {010}, {001}, and {111} Facets of Anatase Synthesis of Anatase TiO 2 Nanoparticles with Specific Facet on the Surface Photocatalytic and DSSCs Performances of Anatase Titanium Dioxides with Specific Facet on the Surface Layered Titanate Compounds Microwave Hydrothermal Process for Synthesis of Titanium Dioxides Purposes of Present Study References 36 i

3 CHAPTER II Synthesis of Titanium Dioxide Nanocrystals from H 1.07 Ti 1.73 O 4 Layered Titanate Nanosheets by Normal Hydrothermal Process and Their Dye-Sensitized Solar Cell Performance Introduction Experimental Section Preparation of Layered Titanate Nanosheet Colloidal Solutions Preparation of TiO 2 Nanoparticles from PA-HTO Solution Preparation of TiO 2 Nanoparticle Paste and TiO 2 Photoelectrodes Fabrication and Characterization of Dye-Sensitized Solar Cell Adsorption of N719 Dye on TiO 2 Nanoparticles Physical Analysis Results and Discussion Synthesis of TiO 2 Nanoparticles from PA-HTO Nanosheet Solution TEM Study on PA-HTO Nanosheets and TiO 2 Nanoparticles Transformation Reaction Mechanism from PA-HTO Nanosheets to TiO 2 Nanocrystals DSSC Performance of {010}-Faceted TiO 2 Nanoparticles N719 Dye Adsorption Behavior on TiO 2 Nanocrystals with Different Facets on Surfaces Effect of Crystal Facets on DSSC Performance Effect of Light-Scattering Layer on DSSC Performance Conclusion References 75 CHAPTER III ii

4 Synthesis of Titanium Dioxide Nanocrystals from H 1.07 Ti 1.73 O 4 Layered Titanate Nanosheets by Microwave Hydrothermal Process for High Performance Photocatalyst and Dye-Sensitized Solar Cells Introduction Experimental Section Preparation of HTO Layered Titanate Nanosheet Colloidal Solutions Microwave Hydrothermal Treatment of PA-HTO Solution Photocatalytic Characterization Fabrication and Characterization of Dye-Sensitized Solar Cells Physical Analysis Results and Discussion Microwave-Assisted Conversion of HTO Nanosheets to TiO 2 Nanocrystals Nanostructural Study on Conversion Reaction from HTO Nanosheets to TiO 2 Nanocrystals Electronic Band Structure and Photocatalytic Response of TiO 2 Nanocrystals DSSCs Performance of TiO 2 Nanocrystals Conclusion References 106 CHAPTER IV Synthesis of [111]- and {010}-Faceted Anatase TiO 2 Nanocrystals from Tri-Titanate Nanosheets and Their Photocatalytic and DSSCs Performances Introduction Experimental Section 112 iii

5 Preparations of Na 2 Ti 3 O 7 and H 2 Ti 3 O 7 Samples Preparation of H 2 Ti 3 O 7 Nanosheet Colloidal Solution and TiO 2 Nanocrystals Photocatalytic Characterization Fabrication and Characterization of Dye-Sensitized Solar Cells Physical Analysis Results and Discussion Preparations of Tri-Titanate H 2 Ti 3 O 7 Nanosheet Solution Synthesis of TiO 2 Nanocrystals from TMA-HTO Nanosheet Colloidal Solution Nanostructural Study of TiO 2 Nanocrystals and Transformation Reaction Mechanism from TMA-HTO Nanosheets to TiO 2 Nanocrystals Electronic Band Structure and Photocatalytic Activity of TiO 2 Nanocrystals DSSC Performance of [111]-Faceted TiO 2 Nanocrystals Conclusion References 138 CHAPTER V Summary PUBLICATIONS ACKNOWLEDAMENT iv

6 CHAPTER I General Introduction The energy sources consumption is always one of the challenges that the mankind faces. With a constantly increasing population and improved daily living standard, more and more energy sources are needed. The main energy sources used today are the fossil fuels, such as coal, oil, and natural gas. However, the fossil fuels have two main disadvantages, one is the environmental pollution, and the other is non-renewable. To solve these problems, a large number of studies are proceeding to develop clean and sustainable energy sources. The solar energy which radiates light and heat from the sun, such as solar heating, solar photovoltaics, solar thermal energy, and artificial photosynthesis, attracts considerable attentions due to a range of clean and sustainable characteristics. Today the most used solar cell is silicon based solar cell. However, the high quality silicon materials for the solar cells are high cost and heavy pollution in making process. The dye-sensitized solar cell (DSSC) base on the titanium dioxide (TiO 2 ) materials attracts considerable attentions about its potentially low cost, relatively high energy conversion efficiency, safety, and non-pollution. The titanium dioxide (TiO 2 ) is excellent photovoltaic and photocatalytic materials. The synthesis processes of high photovoltaic and photocatalytic performance TiO 2 are related to technologies of chemistry, physics, material, chemical engineering and so on. In this chapter, a general introduction for the TiO 2 functional materials, including the crystal structure and physical properties of TiO 2, photocatalytic reactions of TiO 2 nanocrystals and their applications, the DSSC system and the application of TiO 2 nanocrystals in DSSC system, synthesis of TiO 2 nanocrystals with specific facet, 1

7 layered titanate compounds, and the microwave hydrothermal process. Furthermore, the purposes of this dissertation are clarified. Figure 1.1. Crystal structures of (a) rutile phase, (b) anatase phase, (c) brookite phase, and (d) TiO 2 -B phase Crystal Structures and Physical Properties of Titanium Dioxide Polymorphs As an excellent functional material, a large number of studies have been carried out on the structures, physical and chemical properties, and applications of titanium dioxide (TiO 2 ) materials. And seven kinds of the TiO 2 polymorphs have been reported as far as we know. 1-6 There are four kinds of the main polymorphs, rutile, anatase, brookite, and TiO 2 -B, which were found in nature. Furthermore, the TiO 2 -ІІ and TiO 2 -ІІІ polymorphs are occasionally occurred in nature, and mostly synthesized at high pressure process. 4 The TiO 2 -H and TiO 2 -R polymorphs are metastable phases, which usually are synthesized from titanate precursor by topochemical structural 2

8 conversion reactions. 5, 6 In recent years, the more and more studies have related on rutile and anatase polymorphs. They have a coincident basic blocks consist of the TiO 6 octahedron, where a titanium atom occupies at the center of the TiO 6 octahedron, which is constituted with six of the oxygen atoms. 7, 8 The difference of anatase and rutile crystal structures is caused by the distortion level of each TiO 6 octahedron and the assembly pattern of the each octahedron chains. The crystal structures of the main four TiO 2 polymorphs are shown in Figure 1.1. In the rutile structure, the TiO 6 octahedron is not regular, which shows a slight distortion. 7 And the each TiO 6 octahedron contacts with ten neighbor TiO 6 octahedrons, where two sharing edge oxygen atoms and eight sharing corner oxygen atoms as shown in Figure 1.1(a). The Ti-Ti bond lengths are nm and nm, the Ti-O bond lengths are nm and nm in the rutile structure. 2 In the anatase structure, the TiO 6 octahedron of constituent structure is significantly distorted and the each TiO 6 octahedron contacts with neighbor eight TiO 6 octahedrons as shown in Figure 1.1(b), in which four sharing corner oxygen atoms and four sharing edge oxygen atoms, respectively. And the Ti-Ti bond lengths are nm and nm, which are bigger than that of rutile phase, the Ti-O bond lengths are nm and nm, which are smaller than that of rutile phase in the anatase structure. 1, 9 Moreover, in the crystal structure of the brookite phase, the TiO 6 octahedrons share both edges and corners, forming an orthorhombic structure as shown in Figure 1.1(c). 10, 11 The Ti-Ti bond lengths are nm and nm, the Ti-O bond lengths are nm in the brookite structure. 12 And the TiO 2 -B has a monoclinic structure, which is composed of corrugated sheets of edge- and corner-sharing TiO 6 octahedrons as shown in Figure 1.1(d). 13, 14 3

9 Figure 1.2. XRD patterns of rutile, anatase, and brookite phases. The structural parameters of the main TiO 2 polymorphs, rutile, and anatase, brookite, and TiO 2 -B phases are summarized in Table 1.1. The diffraction techniques, including X-ray powder diffraction (XRD) and selected area electron diffraction (SAED), are important and useful method for the structural studies of the TiO 2 powder samples. The typical XRD patterns of rutile, anatase, and brookite phases are shown in Figure 1.2. The rutile phase with tetragonal crystal system has characteristic diffraction peaks of (110), (101), (200), (111), (210), (211), and (220). The anatase phase with tetragonal crystal system exhibits diffraction peaks of (101), (103), (004), (112), (200), (105), and (211). The brookite phase with orthorhombic system has many diffraction peaks and the main peaks are (210), and (111), and (210). Furthermore, TiO 2 -B phase belong to monoclinic system, which exhibits the main peaks of (110), (002), (-113), and (020). 14 4

10 Table 1.1. Structural parameters and physical properties of main TiO 2 phases. Crystal phase Rutile Anatase Brookite TiO 2 -B Crystal system Tetragonal Tetragonal Orthorhombic Monoclinic Space group P4 2 /mnm I4 1 /amd Pbca C2/m Lattice constants (nm) a = a = a = a = c = c = b = b = c = c = β = o Mass density (g cm -3 ) Refractive index n ω = 2.61 n ω = 2.56 n α = n ε = 2.91 n ε = 2.49 n β = 2.58 n γ = 2.70 Dielectric constant (ε) Band gap energy (ev) ~3.0 ~ Mohs scale The different crystal structures of the TiO 2 polymorphs causes their different physical and chemical properties, such as mass densities, stabilities, thermodynamic properties, electronic band structure, dielectric constant and so on The some physical and chemical properties are shown in Table 1.1. Rutile is a thermodynamic stable phase, and anatase is metastable phase which transforms to the rutile phase above 600 o C. The mass density, refractive index, dielectric constant, and mohs scale of rutile phase are larger than which of anatase phase. And the band gap energy of 21, 22, 23, 24 rutile phase is smaller than that of the anatase phase. 5

11 1.2. Photocatalytic Reactions of Titanium Dioxides and Their Applications Since Fujishima and Honda reported a breakthrough research of photocatalytic water splitting on TiO 2 with the UV-light irradiating in 1972, the heterogeneous photocatalysis has been a focus of considerable attention. 25, 26 After that, a large number of studies have been reported on the photocatalytic reactions and their applications. In this section, a review on the photocatalytic reactions of TiO 2 and their applications is given Photocatalytic Reactions Mechanism The phenomenon of heterogeneous photocatalysis occur when a semiconductor (photocatalyst) is irradiated by visible or UV light. The semiconductor has a void region called bandgap which extends from the top of the filled valence band (VB) to the bottom of the vacant conduction band (CB). When the semiconductor is irradiated under visible or UV light with light energy equal to or greater than its bandgap energy (E g ), an electron (e - ) is excited from the VB to the CB, and a hole (h + ) is created in the VB. The excited electron in CB and created hole in VB can act as reductant and oxidant, respectively, and they can cause redox reaction on the semiconductor surface. Such photochemical reaction is called photocatalytic reaction and the semiconductor acts as photocatalyst in the reaction

12 Figure 1.3. Schematic illustrating of the photocatalytic reactions on a semiconductor in heterogeneous photocatalysis. A more detailed photocatalytic reaction process is illustrated in Figure 1.3. When the photocatalyst absorbs the irradiated visible or UV light, the e - -h + pairs (exciton) can be generated in the photocatalyst. The excited electron in CB can transfer to the photocatalyst surface (pathway A), and reduce an electron acceptor adsorbed on the surface, such as oxygen in an aerated solution. The O 2 adsorbed on the surface can be reduced to superoxide (O 2 ), and then O 2 can further react with other species. The generated hole in VB also can migrate to the photocatalyst surface (pathway B), and combine with an electron from an electron donor species, and oxidize the donor species adsorbed on the photocatalyst surface. For example, the holes can oxidize OH - adsorbed on the surface to OH radical in an aqueous solution, and then the OH radical formed can further react with other species. Furthermore, the excited e - and generated h + can recombine on the photocatalyst surface and inside of the photocatalyst (pathway C). And also the excited e - and generated h + can be trapped by 7

13 the lattice defect of photocatalyst crystal, and then recombine together. 31 Therefore, the inhibition of the e - -h + charge recombination and promoting the charge separation are effective to enhance the photocatalytic efficiency. The decreasing crystal defect or increasing crystallinity of the photocatalyst is an effective method to diminish the charge recombination. The surface modification with metals (Pt, Au, Ag, etc.) or semiconductors (CdS, CdSe, etc.) can promote the charge separation as shown in Figure 1.4, where the excited electrons can transfer to the metal side and generated hole to the semiconductor side. Since the photocatalytic reactions occur on the photocatalyst surface, the enlargement of photocatalyst surface area can enhance also enhance the photocatalytic efficiency Figure 1.4. Schematic illustrating the charge separation process by the TiO 2 surface modification with metal (Pt) and semiconductor (CdS). The redox potentials of photogenerated electrons and holes at photocatalyst are depended on the bandgap energy positions of the photocatalyst. The relationship between the band energy level and redox potential of electrochemical reaction is shown in Figure 1.5 to illustrate the thermodynamic limitations of the redox reactions that can be carried out with photogenerated electrons and holes. 31 The photogenerated 8

14 electrons by the photocatalyst can reduce thermodynamically an electron acceptor species, if the reduction potential of the acceptor species locates below the conduction band of the photocatalyst. If the oxidation potential of an electron donor species locates above the valence band of the photocatalyst, the photogenerated holes can oxidize the donor species thermodynamically. Figure 1.5. Band-edges and positions of valence band and conduction band of several semiconductors in contact with chemical reduction potential (NHE), and the H 2 /H 2 O and O 2 /H 2 O reduction potentials in an aqueous solution at ph 1. The TiO 2 is n-type semiconductor materials. Similar to other semiconductor materials, TiO 2 can absorb light whose energy is equal or larger than the bandgap energy of TiO 2, and lead to inject the electrons from the valence band to the conduction band. The anatase phase and rutile phase have slightly different bandgap energies of 3.2 and 3.0 ev, which correspond to the wavelength of absorbance band edges of 410 and 393 nm, respectively. The band energy levels of anatase and rutile are also different as given in Figure 1.5. The bandgap energy and band-edge position are dependent not only on chemical composition, but also on the structure The TiO 2 polymorphs exhibit different photocatalytic behavior because they have different bandgap energies and band-edge positions. In the TiO 2 polymorphs, anatase exhibits 9

15 highest photocatalytic activity Visible Light Active Photocatalysis of TiO 2 The bandgap energies of TiO 2 polymorphs indicate that the UV light with a wavelength less 410 or 393 nm is necessary for photocatalytic reactions on anatase or rutile, but it only has 4% of UV light in the solar spectrum contains. It means the efficiencies of photocatalytic reactions of the TiO 2 polymorphs are low on the sun light irradiating conditions. Therefore, the development of visible light active TiO 2 photocatalyst is significant in the viewpoint of exploitation of solar energy because of the more than 90% sunlight energy is in visible light region. The main method is shifting absorption threshold of TiO 2 semiconductor towards visible region. The doping technology of TiO 2 semiconductor has been reported for the preparation of visible light active TiO 2 photocatalyst, such as metal element doped TiO 2 semiconductor (Nd 3+, Cr 3+, Fe 3+, Co 2+, V 5+ ), non-metal element doped TiO 2 semiconductor (N), and oxygen vacancy TiO 2 semiconductor Figure 1.6. Schematic diagram of an electrochemical photocell developed by Fujishima and Honda. (1) n-type TiO 2 electrode; (2) platinum black counter electrode; (3) ionically conducting separator; (4) gas burette; (5) load resistance and (6) 10

16 voltmeter Application of Photocatalysis to Decomposition of H 2 O and Organic Pollution The photodecomposition of H 2 O effect is similar with the photosynthesis of plants. The plants can obtain the energy and produce oxygen through oxidizing H 2 O and reducing CO 2 with sunlight. 25, 26 On the basis of the principle of photosynthesis, Fujishima and Honda have found that H 2 O can be split using the TiO 2 semiconductor material with the UV-light irradiating. 25 The device of splitting H 2 O is shown in Figure 1.6. The TiO 2 electrode and the platinum (Pt) counter electrode are connected with an electrical load. Once the UV light with light consisting of wavelengths shorter than 415 nm irradiates on the surface of the TiO 2 electrode, the photogeneration e - can flow to the Pt electrode from the TiO 2 electrode through the external circuit. In other words, the photocurrent flows from the Pt counter electrode to the TiO 2 electrode through the external circuit. This process is very similar with electrochemical decomposition of water, and the distinction is the process of electrochemical decomposition of H 2 O needs to add an electric potential difference of more than 1.23 V between anodic electrode and cathodic electrode, and the photodecomposition of H 2 O process needs to absorb UV light with wavelength less than about 400 nm. The direction of the photocurrent expounds that an oxidation reaction can occur at the TiO 2 electrode and a reduction reaction can occur at the Pt electrode, and the oxygen and hydrogen can be obtained from this process. The photodecomposition reaction of H 2 O can accord to the following reaction mechanism from (1.1) to (1.4). 26 TiO 2 + 2hv 2h + + 2e - (excitation of TiO 2 by UV light) (1.1) H 2 O + 2h + (1/2)O 2 + 2H + (at the TiO 2 electrode) (1.2) 2H + + 2e - H 2 (at the Pt electrode) (1.3) H 2 O + 2hv (1/2)O 2 + H 2 (overall reaction) (1.4) 11

17 Frank and Bard have reported the photodecomposition of cyanide in solution by TiO 2 suspension, and it starts the new direction to aiming at apply TiO 2 photocatalysis to decompose the organic pollution, such as the photocatalytic decompositions of aldehydes, carboxylic acids, chlorophenols, dyes, phenolics, ethers, ketones and so on. In this photocatalytic reaction, the organic pollution can be decomposed by the oxidation reaction. The various forms of active oxygen like O 2, HO 2, OH, O, and H 2 O 2 can be produced These active oxygen species have fortissimo oxidizability and most of organic pollution can be essentially oxidized completely to CO Dye-Sensitized Solar Cells In 1991, a new type of photovoltaic cell fabricated by using dye sensitized effect, which is known as a dye-sensitized solar cell (DSSC), has be developed by Grӓtzel group. This DSSC can successfully convert visible light energy to electrical energy and achieve 7.9% solar energy conversion efficiency, and the incident photon to electrical current conversion efficiency is nearly 80%. 67 After that, a larger number of intensive studies on DSSCs have been carried out among chemistry, physics, material, chemical engineering and so on, and the highest conversion efficiency of DSSC has been renewed about 12.3%. 81 Many kinds of excellent factors led the DSSC to gained extensive attentions, such as low production cost, flexibility, lightweight, transparency, multicolor options, and indoor applications

18 Structure and Principle of Dye-Sensitized Solar Cell Figure 1.7. Schematic structure of a typical dye-sensitized solar cell. A typical DSSC is composed of three main parts: the dye sensitized TiO 2 electrode, the liquid electrolyte containing I - 3 /I - redox couple, and the platinum (Pt) electrode with sandwich structure as shown in Figure The dye sensitized TiO 2 electrode usually composes of a sintered TiO 2 film on conducting glass coated with fluorine-doped tin oxide (FTO), the thickness of sintered TiO 2 film is around 10 µm, and the porosity of film is around 50-60%. 67, 85 A photosensitizer with the anchoring groups (-COOH) can be anchored onto the TiO 2 nanoparticle surface. After photo exciting, the oxidized photosensitizer can be restored to ground state by electron transfer in the liquid electrolyte containing the I - 3 /I - redox couple. Furthermore, the formed I 3 - ions can be reduced to I - ions on the cathode with a coated platinum catalyst. 65, To achieve the best performance, the charge transfer resistance of platinum electrode is less than 1Ω cm The operation principle of DSSC is shown in Figure 1.8, where the desired pathways of the electron transfer processes are described as path 1 to 4. When the sun light irradiate on the dye sensitized TiO 2 electrode surface, the photogenerated e - 13

19 can be excited from the ground state to excited state of dye as the path 1, and the photogenerated hole can be left to form a oxidized dye. And then the photogenerated e - can inject into the conduction band of TiO 2 as the path 2, and then the photogenerated e - can be collected on the FTO substrate through the TiO 2 film. The collected photogenerated e - can transfer to the platinum counter electrode through outside circuit as the path 3. In the meantime, the oxidized dyes on the TiO 2 electrode can be regenerated by 3I - from the electrolyte as the path 4. And 3I - is then further regenerated at the Pt electrode through the reduction of I 3 - by the photogenerated e -. Furthermore, the photogenerated e - loss processes are described as path 5 to 7. The photogenerated e - can recombine with the hole in the oxidized dye as the path 5. The photogenerated e - injected into the conduction band of TiO 2 can also recombine with the hole in the oxidized dye as the path 6 and with the electron acceptors of I 3 - ions in the electrolyte as the path 7, respectively. Usually, we call these photogenerated e - loss processes as the back current, and it degrades the 65, performance of DSSC. Figure 1.8. Operation principle of DSSC and main electron paths in the cell. 14

20 For the photogenerated e - can transfer from dye to TiO 2 semiconductor, successfully, the potential of electronically excited state (S * ) needs lower than the potential of CB of TiO 2 semiconductor. Furthermore, the maximum output photovoltage of the solar cell corresponds to the difference between Fermi level of semiconductor and the redox potential of electrolyte, which can be expressed by equation (1.5). V max = E Fn E F0 (1.5) Where the V max value is the maximum output photovoltage, the E Fn value is Fermi level of semiconductor, and E F0 value is the redox potential of electrolyte, respectively. For TiO 2 semiconductor and I 3 - /I - redox pair, the maximum output photovoltage or largest open-circuit photovoltage (V oc ) is about 0.9 V Materials for Dye-Sensitized Solar Cell A large improvement in DSSCs in 1991 is the use of a mesoporous TiO 2 electrode with a high surface area in internal TiO 2 electrode. 67 The high surface area of mesoporous TiO 2 electrode can adsorb more dye molecules on the surface in order to improve the efficiencies. The TiO 2 nanocrystal materials for DSSC will be described in detail in Section 1.4. Except TiO 2 nanocrystal materials, other metal oxide materials and ternary oxides, such as ZnO, Nb 2 O 5, SnO 2 and Zn 2 SnO 4, also have been tested for the dye-sensitized electrode Another important material in DSSC is the sensitizer, which can adsorb sun-light and provide the photogenerated electrons for DSSC system. 99, 100 The sensitizer for DSSC should satisfy several basic characteristics: (1) the sensitizer should be able to absorb the sun-light with the whole visible region and even the part of the near-ir, it means the absorption spectrum of the sensitizer should be lower than the wavelength of near-ir (920 nm); (2) the sensitizer should have the anchoring groups such as 15

21 -COOH and -H 2 PO 3 and be able to strongly anchor on the TiO 2 nanoparticle surface; (3) the excited state level of the sensitizer should be higher than the conduction band edge of the TiO 2 in energy, and determine the photogenerated electrons can inject into conduction band of the TiO 2 from the excited dye; (4) the oxidized state level of the sensitizer should be more positive than the redox potential of electrolyte for regeneration process of the sensitizer; (5) the sensitizer should be stable at electrochemical and thermal reaction process. Based on these conditions, many different types of sensitizers have been designed and developed, such as metal complexes, porphyrins and phthalocyanines, and organic dyes. Furthermore, the electrolyte and counter electrode are also the important parts in DSSC. The electrolyte can collect electrons at counter electrode and regenerate the sensitizer as described above; the commonly used electrolytes include liquid redox electrolytes, gel and polymer electrolytes, and ionic liquid electrolytes And the counter electrode are usually prepared by sprayed and heated platinum on the conducting glass substrate surface. Moreover, carbon materials, cobalt sulfide, and conducting polymers are also used to prepare the counter electrode Performance Characteristic Parameters and Equivalent Circuit of DSSC There are four main characteristic parameters, short-circuit photocurrent density (J sc ), open-circuit photovoltage (V oc ), fill factor (ff), and photoelectric power conversion efficiency (η) for the characterization of solar cells performance. These parameters can be evaluated from the photocurrent-voltage characteristic curve of the solar cells as shown in Figure 1.9. The short-circuit photocurrent density is a current density without any external applied voltage, namely current density in the short electric circuit system or maximum current density of the solar cell. The open-circuit 16

22 photovoltage is a voltage in the open circuit system, namely maximum voltage of the solar cell. The maximum V oc of DSSC corresponds to the difference between Fermi level of semiconductor and the redox potential of electrolyte. The fill factor provides an easy comparison for the performance of a solar cell compared to the theoretical maximum performance. The ff value can be calculated by equation (1.6), and the photoelectric power conversion efficiency (η) can be calculated by equation (1.7), where the W MAX value is the maximum output power and I s value is the intensity of the incident light, respectively. Therefore, enhancements of J sc, V oc, and ff are 67, 105 important to enhance solar cell performance η. ff = WMAX Jsc Voc η = Jsc Voc ff Is (1.6) (1.7) Figure 1.9. A typical photocurrent-voltage characteristic curve of solar cell and relations between the photocurrent-voltage characteristics and cell parameters. The electrochemical impedance spectrum (EIS) can be used to measure and analyze internal resistance in DSSC An equivalent circuit and a typical impedance spectrum of DSSC are shown in Figure In the equivalent circuit, R s 17

23 is electron transport resistance in the circuit, R rc and C 1 represent the charge transfer resistance in the charge recombination process and the chemical capacitance of the TiO 2 film and electrolyte interface, respectively, R D is the diffusion resistance of electrolyte, and R Pt and C 2 are the charge transfer resistance and the interface capacitance of Pt and electrolyte interface. Small values of R s, R D, and R Pt, and large value of R rc can improve the DSSC performance. The typical impedance spectrum of DSSC shows a linear feature and three semicircles, respectively. In this spectrum, with decreasing the frequency, firstly a linear feature corresponding R s, a small semicircle resulting from R Pt and C 2, a big semicircle due to R rc and C 1, and then a small semicircle resulting from R D are observed. Therefore the parameters of R s, R Pt, R rc, C 1 and C 2 can be evaluated from the impedance spectrum. Usually the semicircle corresponding to R D cannot be observed even in the very low-frequency region such as until to 0.1 Hz. Figure Equivalent circuit and a typical impedance spectrum of a dye-sensitized solar cell. 18

24 1.4. Titanium Dioxide Nanocrystals for Dye-Sensitized Solar Cell As described above, the dye sensitized TiO 2 electrode is a most significant part in DSSC system. The dye sensitized TiO 2 electrode is fabricated by coating a conducting glass plate with TiO 2 nanoparticles, and adsorbing dye molecules on the TiO 2 nanoparticle surface. Therefore, the performance of dye sensitized TiO 2 electrode is strongly affected by the properties of the TiO 2 nanoparticles. The structure, particle size and morphology, crystallinity, crystal facet exposing on the surface of the TiO 2 nanoparticles can influence the DSSC performance similar to their photocatalytic activity Usually it is considered that anatase nanoparticles exhibit higher performance that rutile particles. Although the anatase phase and rutile phase are all constructed with the TiO 6 octahedron and have a tetragonal lattice system, they show the different DSSC performance and photocatalytic reactivity. Because of the different connection type of these TiO 6 octahedrons and the surface atomic and electronic structures are revealed in the crystal system. 64 And another reason is that anatase nanoparticles with size less than 30 nm are difficult to be synthesized. The high performance TiO 2 electrode has a double-layer structure as shown in Figure The nanocrystalline TiO 2 layer is fabricated using TiO 2 with particle size less 30 nm and the optimum layer thickness is about 10 µm. The nanocrystalline TiO 2 layer is mesoporous layer that provide a large TiO 2 surface area for adsorption of large amount of dye molecules on the dye sensitized electrode, and also increase of the contacting area between the dye molecules and the liquid electrolyte. The increases dye loading amount causes enhancement of the efficiency of sun light harvesting. Therefore, usually the anatase nanoparticles with the size of less than 30 nm are selected. Since the nanocrystalline TiO 2 layer has higher transmittance, the sun light can partially pass through the nanocrystalline TiO 2 layer, which causes 19

25 lower sun light harvesting. To solve this problem, the nanocrystalline TiO 2 layer is coated by a light-scattering layer fabricated using TiO 2 particles with a size of about 114, nm. This layer can reduce the transmission loss of light. Figure Structures of (a) single-layer and (b) double-layer TiO 2 mesoporous electrodes for DSSC. The crystallinity of the TiO 2 nanocrystals also can affect the efficiency of DSSC; because of high crystallinity of the TiO 2 nanocrystal can reduce the probability of captured photogenerated electrons into the defect of the TiO 2 nanocrystal. 82 Furthermore, the effect of morphology of the TiO 2 nanocrystals on the DSSC performance has been studied using spherical nanoparticles, nanorods, nanowires, nanosheets, and nanotubes. 116 The effect of enlarging electron conductivity of the TiO 2 mesoporous electrode using nanorod and nanowire improves the DSSC performance. In recent years, the effect of crystal facet exposing on TiO 2 nanocrystal surface has been reported. 117 It indicate that the different facets exposing on TiO 2 nanocrystal surface can affect the behavior of the dye molecules adsorption, which can affect the efficiency of DSSC. 20

26 1.5. Synthesis of Anatase Titanium Dioxides with Specific Facet on Surface and Their Photocatalytic Activities Since Fujishima and Honda have reported the photocatalytic reactions of TiO 2 to split H 2 O under the UV-light irradiating in 1972, 25 and Grӓtzel group has reported the high performance DSSC in 1991, 67 a large number of studies have been the focus of considerable attention on TiO 2 and their applications. These studies refer to application range of chemistry, physics, material, chemical engineering and so on in order to optimize the photocatalytic reactivity and DSSCs performance. As described above, the TiO 2 has two main phases of anatase and rutile, and the anatase phase is more suitable for the photocatalyst and DSSCs than the rutile phase. 64 Usually, the sol-gel processes are used to the synthesis of anatase TiO 2 nanoparticles. The titanium salts are firstly hydrolyzed in solution, and the products are recrystallized. 118 However, the reaction temperature range is limited; therefore crystal size and morphology are not easy to be controlled through the sol-gel processes. The hydrothermal process is another method to prepare TiO 2 nanoparticles, which can synthesize TiO 2 nanoparticles under high temperature and pressure conditions above 100 o C and 1 atm from a precursor. 119 In recent years, more and more studies have reported to synthesize TiO 2 nanoparticles by hydrothermal process. The anatase, rutile, and brookite particles can be synthesized using the hydrothermal and solvothermal processes The hydrothermal and solvothermal soft chemical processes are also effective for synthesis and particle morphology control of anatase nanoparticles from layered titanate compounds. Recently, the synthesis of anatase nanoparticles with specific crystal facet on the particle surface has become a hot topic in the photocatalytic and DSSC studies. The crystal facet on the surface strongly affects the photocatalytic activity and dye 21

27 molecule adsorption due to different surface energies, arrangements of the atoms on surface, and electronic structures. Wulff et al. have constructed the crystal morphology model of anatase and predicted to the percentage of {101} facet on the crystal surface is about 94% in a normal anatase nanoparticle due to that the {101} facet is a thermodynamic stable facet with low surface energy. 123 Furthermore, they have calculated to the surface energy of different facets in an order of {101} (0.44 J m -2 ) < {010} (0.53 J m -2 ) < {001} (0.90 J m -2 ). The large number of studies has started after this theoretical study of surface energy. Barnard et al. have studied carefully the effects of surface energy and predicted the morphology of anatase nanoparticles in the acidic and alkaline conditions as shown in Figure As a result, the {101}, {010} and {001} facet can be shown on nanoparticle surface in different of acidic and alkaline conditions. This result is quite important to synthesize experimentally the specific facet on nanoparticle surface based on surface chemical principle. Figure Predicted morphology of anatase nanoparticles in acidic and alkaline conditions. Wen et al. have reported first experimental study on the synthesis of the anatase 22

28 TiO 2 nanoparticle with high surface energy of the {010} facet on the surface from a lepidocrocite-like H 1.07 Ti 1.73 O 4 layered titanate nanosheets by hydrothermal reaction, and found that {010}-faceted anatase TiO 2 nanoparticles exhibit higher photocatalytic activity than the normal spherical nanocrystals without specific facet on the nanoparticle surface in The results spur researchers to study the synthesis of TiO 2 nanocrystals with specific facet. And then Yang et al. have synthesized high percentage of {001}-faceted anatase nanoparticles using hydrothermal method in a hydrofluoric acid solution in Furthermore, their results of first-principle quantum chemical calculation reveals that the {001} facet of anatase is one of reactive facets for the photocatalytic reactions. Very recently, Xu et al. have calculated the surface energy of {111} facet and resulted that the {111} facet (1.61 J m -2 ) has higher surface energy than the {010} facet (0.90 J m -2 ) in 2013, 125 they also have given a schematic illustration of {101}, {010}, {001} and {111} facet in anatase structure and these facets on anatase nanoparticle surface as shown in Figure Figure Schematic illustration of {101}, {010}, {001}, and {111} facets of anatase TiO 2 nanoparticle. 23

29 Surface Structures of {101}, {010}, {001}, and {111} Facets of Anatase Figure Surface structures of {101}, {010}, {001}, and {111} facets. The surface structures of {101}, {010}, (001), and {111} are illustrated in Figure As described above, the {101} facets have the low surface energy, the surface of bulk-terminated {101} facets expose unsaturated Ti 5c and saturated Ti 6c titanium atoms as well as O 2c and O 3c oxygen atoms. Because of the anatase TiO 2 is the tetragonal system, the bulk-terminated {010} and {100} facets have the same surface structure, which expose unsaturated Ti 5c titanium atoms, O 2c and O 3c oxygen atoms. Comparing with the {101} facets, high percentage of Ti 5c titanium atoms on the {010} surface results in a higher surface energy than the {101} facets. The {001} facets expose unsaturated Ti 5c titanium atoms, O 2c and O 3c oxygen atoms on the surface. As 24

30 the same with the {010} facets, the unsaturated Ti 5c and O 2c atoms can lead the facets to have high surface energy of 0.90 J m -2. The surface of {111} facets have the unsaturated Ti 5c and Ti 3c atoms. Xu et al. have used the density functional theory (DFT) to calculate surface energy of anatase, and gave the surface energies for {101}, {010}, {001} and {111} facets increasing in an order of {101} (0.43 J m -2 ) < {010} (0.57 J m -2 ) < {001} (0.95 J m -2 ) < {111} (1.61 J m -2 ), which means the {111} facets may have high photocatalytic activity Synthesis of Anatase TiO 2 Nanoparticles with Specific Facet on the Surface The {101} facets can be formed on small particle surface, and usually the large particles grew by Ostwald ripening expose mainly the low surface energy {101} facts on the surface. Penn et al. have reported the slightly truncated octahedron crystals with dominant of {101} facets can be obtained from sol-gel route under hydrothermal conditions using the pristine TiO 2 nanoparticles as a precursor. The percentage of grew {101} can be change via changing growth directions under different hydrothermal conditions. 126 The result finds that the particles can rapidly grow along [001] direction in an acidic solution comparing with deionized water, and the area reduction of high surface energy facets, resulting in stabilization, is the driving force of particle growth. Amano et al. have synthesized the octahedral anatase single crystals with {101} facets from titanate nanowires under hydrothermal conditions. 127 Less studies have been reported on the synthesis of {010}-faceted anatase particles. Wen et al. the first reported the synthesis of the {010}-faceted anatase TiO 2 nanoparticle from a lepidocrocite-like H 1.07 Ti 1.73 O 4 layered titanate nanosheets by hydrothermal reaction. 117 After that, Wu et al. have synthesized the rhombic anatase TiO 2 nanoparticles with a large percentage of {010} facets and anatase TiO 2 25

31 nanosheet with exposed {001} facets by a nonaqueous synthetic process. 113 Pan et al. have obtained {010}-faceted anatase TiO 2 particles through hydrothermal treatment of TiOSO 4 aqueous solution precursor in different concentrations of TiOSO 4 -HF aqueous solutions and reaction conditions. 128 Furthermore, they also have prepared anatase TiO 2 rods with {010} facets from lepidocrocite-like layered titanate Cs 0.68 Ti 1.83 O 4 /H 0.68 Ti 1.83 O 4 precursor. 129 Yang et al. have synthesized high percentage of {001}-faceted anatase particles using titanium tetrafluoride and hydrofluoric acid solution under hydrothermal conditions. 112 And then a large number of studies have been reported on synthesis of the {001}-faceted anatase before its high surface energy and promising high photocatalytic performance. Han et al. have used tetrabutyl titanate and hydrofluoric acid to synthesize anatase nanosheets with {001} facets, the result reveals the photocatalytic reactivity is higher than P25 nanoparticles. 130 Furthermore, Liu et al. have obtained the dominant {001} facets of anatase TiO 2 sheets from TiB 2 powder precursor and aqueous solution of HF under hydrothermal treatment. 131 Xu et al. have reported that single-crystalline anatase with the {111} facets can be synthesized from a gel-like precursor by heating treatment in a NH 3 -HF solution. 125 However, they mistake the facet assignment. Actually, the synthesized anatase particles have a facet vertical to the [111]-direction on the surface, and we call it [111]-facet. The [111]-facet is different to {111}-facet in the tetragonal lattice system. Therefore, the synthesis of {111}-faceted anatase particles have not been achieved. Du et al. have prepared the [111]-faceted anatase nanocrystals from exfoliated K 2 Ti 4 O 9 nanosheets under hydrothermal conditions

32 Photocatalytic and DSSCs Performances of Anatase Titanium Dioxides with Specific Facet on the Surface Many studies have been reported on the photocatalytic activity of anatase particles with specific facet on the surface. Wen et al. have reported the first photocatalytic study on the anatase nanocrystals with the {010} facets on the surface, and found that {010}-faceted anatase TiO 2 nanoparticles exhibit higher photocatalytic activity than the normal spherical nanocrystals without specific facet on the nanoparticle surface. 117 Amano et al. have reported the octahedral anatase single crystals with {101} facets exhibited relatively high photocatalytic activity for decomposition of organic compounds and low activity for hydrogen evolution, it may be due to the characteristics of the anatase {101} surface. 127 Han et al. have reported the anatase nanosheets with {001} facets revealed the higher photocatalytic reactivity than P25 nanoparticles by comparing the degradation efficiency of methyl orange. 130 Furthermore, Liu have reported the anatase TiO 2 single crystals with dominant {001} facets also have higher photocatalytic reactivity than P25 nanoparticles by degrading of methylene blue dye and indicated the high density of unsaturated Ti 5C has rather high reactive for dissociative absorption of dye molecules. 132 Very recently, Du et al. have reported that the [111]-faceted anatase nanocrystals exhibit higher photocatalytic activity than that of spherical nanoparticles without specific facet on the surface (non-facet), but lower photocatalytic activity than the {010}-faceted anatase nanocrystals. The results of photocatalytic studies suggest that the photocatalytic activity is not only dependent on the surface energy but also on the surface crystal structure and electronic structure. 122 Although a large number of studies have been reported on synthesis and photocatalytic activity of the TiO 2 with specific facet on the particle surface, unfortunately, only less number of studies have been reported on the facet effect on 27

33 the DSSC performance, it may be due to the synthesis processes are difficult to provide enough amounts of nanocrystals with the specific facet for the DSSC characterizations, and/or the nanocrystals sizes are too large for high performance DSSC. Actually, the most of anatase particle with specific facet on the surface are synthesized by crystal growth processes based on Ostwald ripening mechanism in the solutions containing facet directing agents such as HF, NH 3, and organic compounds as described above. These processes can provide the anatase particles with particle size of about 1 μm, 130 however, it is difficult to prepare nanoparticles with size of about 20 nm for high performance DSSCs. Wen et al. have reported the first study on the crystal facet effect on the DSSC performance, and found that {010}-faceted anatase nanocrystals exhibit specially high J sc value due to strong adsorption of sensitizer dye molecules on the {010} facets, which improves injecting photoelectrons from LUMO (lowest unoccupied molecular orbital) of the molecule to the conduction band of TiO 2. 79, 133 Wu et al. have reported that the bipyramid rod-like anatase nanocrystals with high percentage of {101} faces can capture the electrons injected from the dye photoexcited state to the anatase conducting band, decrease the annihilation of electrons, and increase the electron concentration of the TiO 2 photoelectrode compared to the spherical anatase nanocrystals. 134 Furthermore, Wang et al. have studied the DSSC performance of anatase nanosheets-based microspheres with dominant high surface energy {001} facets, which exhibited the superior cell performance comparing with that of P25. And they concluded that the TiO 2 nanosheets-based microspheres have high surface area to increase amount of absorbed dye molecules, and it also can enhance light harvesting. 135 Very recently, we have studied the DSSC performance of {010}-faceted, [111]-faceted, and non-faceted anatase nanoparticles with size about 20 nm and compared with P25. 28

34 1.6. Layered Titanate Compounds The layered titanate compounds can be applied to photocatalyst, fuel cells electrolyte, and biomedical materials The layered titanate compounds are constructed by stacking negatively charged structural unit layers and cations such as the alkali and alkaline earth metal ions in the interlayer spaces. The structure of negatively charged layer is consisted with corner and edge shared TiO 6 octahedrons. Depending on the content of the alkali or alkaline metal and the assembly pattern of the corner and edge shared TiO 6 octahedron, the layered titanate compounds have two kinds of main structural formula A 2 Ti n O 2n+1 and A m Ti 2-m/3 B m/3 O 4, respectively. For A 2 Ti n O 2n+1 system, the structural unit of layered titanate compound is the number n of edge shared TiO 6 octahedrons on the same plane and the adjacent structure unit can share corner, and form corrugated structure of layers, namely they have step-like layered structures. For A m Ti 2-m/3 B m/3 O 4 system, it shows the structure related to lepidocrocite-type structure. This layer structure is composed of edge-shared TiO 6 octahedrons which construct a flat TiO 6 octahedral layer. The different structures of the layered titanate compounds usually exhibit different properties, such as layer charge density and exfoliation behavior. 139 The Na 2 Ti 3 O 7 and K 2 Ti 4 O 9 layered titanates all belong to A 2 Ti n O 2n+1 system as 140, 141 shown in Figure 1.15, which have been reported firstly by Anderson in They are all monoclinic system, the sodium ions and potassium ions locate in the interlayers and the structural unit numbers are 3 and 4, respectively, there are 3 and 4 of edge shared TiO 6 octahedrons on the same plane. These structural units can connect with the adjacent structural units using the same corner. This kind of layered titanate can grow along b-axis and form wire or ribbon morphology of particles. Furthermore, different structures of Na 2 Ti 3 O 7 and K 2 Ti 4 O 9 layered titanates occur the different 29

35 charge densities of the host layers, which can affect the ion-exchange, ions intercalated and exfoliation reactions. Figure Structures of Na 2 Ti 3 O 7 and K 2 Ti 4 O 9 layered titanates. The K 0.8 Ti 1.73 Li 0.27 O 4 and Cs 0.68 Ti 1.83 O 4 layered titanates belong to A m Ti 2-m/3 B m/3 O 4 system as shown in Figure , 143 They belong to orthorhombic system and have the lepidocrocite-like layered structure. The K 0.8 Ti 1.73 Li 0.27 O 4 layered structure is composed of host layers of edge-shared TiO 6 octahedrons and interlayer exchangeable potassium ions compensating for the minus charge of the TiO 6 octahedral layers. 30

36 Lithium ions occupy the titanium (IV) octahedral sites in the TiO 6 octahedrons. Because of the lower charge density of the K 0.8 Ti 1.73 Li 0.27 O 4 layered titanate, it is often used as a precursor to synthesize other titanate compounds. Furthermore, the K 0.8 Ti 1.73 Li 0.27 O 4 layered titanate can grow along a-axis and c-axis, and form sheet and plate morphology of particles. The Cs 0.68 Ti 1.83 O 4 layered titanate has a similar structure to the K 0.8 Ti 1.73 Li 0.27 O 4 layered titanate. There are some vacancies in the titanium (IV) octahedral sites, and its properties are also similar to which of the K 0.8 Ti 1.73 Li 0.27 O 4 layered titanate. Figure Structure of K 0.8 Ti 1.73 Li 0.27 O 4 layered titanate Microwave Hydrothermal Process for Synthesis of Titanium Dioxides As described above, the hydrothermal processes are useful for the synthesis of TiO 2 nanoparticles, and the crystal structure, crystallinity, crystal size, crystal morphology and crystal plane can be controlled by changing the reaction conditions, such as reaction time, reaction temperature, concentration, and so on in the reaction system. However, usually the normal hydrothermal processes needs to cost much long reaction time for reaction, owing to tardy heat transmission from outside to inside as 31

37 shown in Figure Microwave hydrothermal process is a unique and useful method, in which the solvent molecules and reactants can be heated directly by increasing their rotational speeds when solvent molecules and reactants absorb the microwave (Figure 1.17). This unique property can shorten reaction time and change the property of product Figure Modes of hydrothermal bomb recrystallizations under (a) normal hydrothermal and (b) microwave hydrothermal conditions. Komarneni et al. have reported the first synthesis of TiO 2 from a diluted titanium tetrachloride precursor solution using the microwave hydrothermal method and found that the microwave hydrothermal process can lead to quite rapid crystallization, and the particle size, morphology, and polymorph can be controlled under the microwave hydrothermal conditions. 144 Immediately following, Wilson et.al have used hydrolysis of titanium isopropoxide as a precursor and synthesized nanocrystalline anatase TiO 2 colloids, the result revealed that the microwave hydrothermal treatment can obtain high crystalline product with a smaller size and a more regular shape in a short reaction time and low energy comparing with the normal hydrothermal processes

38 1.8. Purposes of Present Study Titanium dioxide is a significant functional material which has wide applications to white pigment, catalyst supports, photocatalyst, and dye-sensitized solar cell (DSSC). In these applications, TiO 2 has several advantages, such as, inexpensiveness, high stability, high oxidizability of photogenerated holes, and high reducing ability of photogenerated photoelectrons. TiO 2 also is an excellent metal oxide semiconductor for DSSC due to its excellent ability to adsorb dye molecules and high conductivity to transport photogenerated electrons. The photocatalytic and DSSC performances of TiO 2 materials are strongly dependent on the structure, surface area, crystallinity, and crystal facet exposing on crystal surface. The present study aims (1) to synthesize the superior titanium dioxide nanoparticles with the controllable crystal structure, crystal size, crystal morphology, and facet on the surface from layered titanate nanosheet precursors using normal hydrothermal process and microwave hydrothermal process, (2) to study the mechanisms of formation process from the layered titanate nanosheet precursors with different crystal structures to titanium dioxide nanoparticles in the normal and microwave hydrothermal processes, (3) to clarify the facet effects on the photocatalytic reaction and DSSC performance and the superior titanium dioxide nanoparticles for the photocatalyst and DSSC. In Chapter II, the formation and characterization of anatase TiO 2 nanoparticles from H 1.07 Ti 1.73 O 4 layered titanate nanosheets under hydrothermal conditions and their DSSC performances compared with non-faceted anatase nanocrystals and commercial P25 with partially [111]-faceted anatase nanocrystals are described. The {010}-faceted anatase TiO 2 nanocrystals with various morphologies are formed by the hydrothermal reaction of the H 1.07 Ti 1.73 O 4 layered titanate nanosheets. There are two 33

39 kinds of reactions in the formation process of the anatase nanocrystals from the H 1.07 Ti 1.73 O 4 nanosheets. One is an in situ topotactic structural transformation reaction, and another is a dissolution-deposition reaction. The results of DSSC study suggested that the DSSC performance increases in an order of non-faceted spherical nanocrystals < [111]-faceted nanocrystals < {010}-faceted nanocrystals, which concludes that the facet on the nanoparticle surface significantly impacts on the DSSC performance and the {010}-faceted nanocrystals are promising for the high performance DSSCs. Furthermore, the results of light-scattering layer effect reveal that leaflike anatase TiO 2 nanoparticles can enhance light harvesting in the TiO 2 electrode and improve DSSCs performances. In Chapter III, a microwave hydrothermal process for the synthesis of TiO 2 nanocrystals from H 1.07 Ti 1.73 O 4 layered titanate nanosheet solution and the characterization of the synthesized TiO 2 nanocrystals are described. The {010}-faceted anatase nanocrystals with controllable crystal size and morphology can be synthesized by microwave hydrothermal treatment of layered titanate nanosheet solutions. The crystal size and morphology of the anatase nanocrystals synthesized using the microwave hydrothermal process are more uniform than which synthesized using the normal hydrothermal process described in Chapter II. The photocatalytic behavior and DSSC performance of the synthesized {010}-faceted anatase nanocrystals are studied and compared with P25 nanocrystals and non-faceted anatase nanocrystals. The microwave hydrothermal process is suitable to control the structural conversion reaction for the uniform the crystal size and morphology due to its uniform heating mechanism. The UV-visible spectrum results revealed that the bandgap of the anatase nanocrystals enhance in an order of non-faceted nanocrystal < [111]-faceted nanocrystal < {010}-faceted nanocrystal, which corresponded to their photocatalytic activities. The DSSC performance also enhanced in the same order, namely the 34

40 {010}-faceted nanocrystals show the highest DSSC performance and photocatalytic activity. In Chapter IV, the synthesis and characterization of TiO 2 nanocrystals with specific facets on the surface from H 2 Ti 3 O 7 layered titanate nanosheets by hydrothermal process are described. The anatase, rutile, and brookite TiO 2 nanocrystals can be synthesized from H 2 Ti 3 O 7 layered titanate nanosheets. The crystallinity, particle size and morphology, and facet on the surface of anatase TiO 2 nanocrystals are strongly depended on hydrothermal conditions. The [111]-faceted and {010}-faceted anatase nanocrystals are obtained by controlling the hydrothermal reaction conditions. The DSSC performance and the photocatalytic activity of anatase TiO 2 nanocrystals with [111]-faceted and {010}-faceted on the surface are investigated, and results suggest that the DSSCs performance and the photocatalytic activity are dependent on the facet and crystal size of the nanocrystals. The increasing order of non-faceted nanocrystal < [111]-faceted nanocrystal < {010}-faceted nanocrystal for the photocatalytic activity and DSSC performance is confirmed by using the [111]-faceted anatase nanocrystals synthesized from H 2 Ti 3 O 7 layered titanate nanosheets. In Chapter V, a summary for this dissertation is given. The results of the synthesis of anatase nanocrystals in this study indicate a direction for the synthesis TiO 2 nanocrystals with specific facet on the surface, and this process will be suitable also to synthesize other faceted anatase nanocrystals and other TiO 2 nanocrystals with specific facet on the surface, and also to synthesize other metal oxide nanocrystals. The conclusions of the photocatalytic and DSSC studies give a new viewpoint for the developments of high performance photocatalytic and DSSC materials using crystal facet effects. 35

41 1.9. References [1]. Grant, F. A. Rev. Mod. Phys. 1959, 31, [2]. Czanderna, A. W.; Clifford, A.F.; Honig, J. M. J. Am. Chem. Soc. 1957, 79, [3]. Banfield, J. F.; Veblen, D. R.; Smith, D. J. Am. Miner. 1991, 76, [4]. El Goresy, A.; Chen, M.; Dubrovinsky, L.; Gliiet, P.; Graup, G. Science 2001, 293, [5]. Latroche, M.; Brohan, L.; Marchand, R.; Toumoux, M. J. Solid State Chem. 1989, 81, [6]. Akimoto, J.; Gotoh, Y.; Oosawa, Y.; Nonose, N.; Kumagai, T.; Aoki, K.; Takei, H. J. Solid State Chem. 1994, 113, [7]. Diebold, U. Surf. Sci. Rep. 2003, 48, [8]. Landmann, M.; Rauls, E.; Schmidt, W. G. J. Phys.: Condens. Matter 2012, 24, [9]. Liu, L.; Chen, X. Chem. Rev. 2014, 114, [10]. Li, Y.; White, T. J.; Lim, S. H. J. Solid State Chem. 2004, 177, [11]. Reyse-Coronado, D.; Rodriguez-Gattomo, G.; Espinosa-Pesqueira, M.E.; Cab, C.; De Coss, R.; Oskam, G. Nanotechnology 2008, 19, [12]. Meagher, E.P. Can. Mineral. 1979, 17, [13]. Su, X.; Wu, Q.; Zhan, X.; Wu, J.; Wei, S.; Guo, Z. J. Mater Sci. 2012, 47, [14]. Liu, Z.; Andreev, Y. G.; Armstrong, A. R.; Brutti, S.; Ren, Y.; Bruce, P. G. Progress in Natural Science: Materials International 2013, 23, [15]. Cronemeyer, D. C. Phys. Rev. 1952, 87, [16]. Turchi, C.S.; Ollis, D. F.; J. Catal. 1990, 122,

42 [17]. Reinhardt, P.; Hess, B. A. Phys. Rev. B 1994, 50, [18]. Asahi, R.; Taga, Y.; Mannstadt, W.; Freeman, A. J. Phys. Rev. B - Condensed Matter and Materials Physics 2000, 61, [19]. Wachs, I. E.; Chen, Y.; Jehng, J. M.; Briand, L. E.; Tanaka, T. Catal. Today 2003, 78, [20]. Chen, X.; Mao, S. S. Chem. Rev. 2007, 107, [21]. Sanjinés, R.; Tang, H.; Berger, H.; Gozzo, F.; Margaritondo, G.; Lévy, F. J. Appl. Phys. 1994, 75, [22]. Martin, S. T.; Herrmann, H.; Hoffmann, M. R. J. Chem. Soc. Faraday Trans. 1994, 90, [23]. Umebayashi, T.; Yamaki, T.; Itoh, H.; Asai, K. Appl. Phys. Lett. 2002, 81, [24]. Scanlon, D. O.; Dunnill, C. W.; Buckeridge, J.; Shevlin, S. A.; Logsdail, A. J.; Woodley, S. M.; Catlow, C. R. A.; Powell, M. J.; Palgrave, R. G.; Parkin, I. P.; Watson, G. W.; Keal, T. W.; Sherwood, P.; Walsh, A.; Sokol, A. Nat. Mater. 2013, 12, [25]. Fujishima, A.; Honda, K. Nature 1972, 238, [26]. Fujishima, A.; Rao, T. N.; Tryk, D. A. J. Photoch. Photobio. C 2000, 1, [27]. Fujishima, A.; Kobayakawa, K.; Honda, K. J. Electrochem. Soc. 1975, 122, [28]. Fujishima, A.; Rao, T. N.; Tryk, D. A. Electrochim. Acta 2000, 45, [29]. Hashimoto, K.; Irie, H.; Fujishima, A. Jpn, J. Appl. Phys. 2005, 44, [30]. Nakata, K.; Fujishima, A. J. Photoch. Photobio. C 2012, 13, [31]. Linsebigler, A. L.; Lu, G.; Yates, J. T. Chem. Rev. 1995, 95, [32]. Anpo, M.; Takeuchi, M. J. Catal. 2003, 216, [33]. Subramanian, V.; Wolf, E. E.; Kamat, P. V. J. Am. Chem. Soc. 2004, 126, 37

43 [34]. Tada, H.; Mitsui, T.; Kiyonaga, T.; Akita, T.; Tanaka, K. Nat. Mater. 2006, 5, [35]. Robel, I.; Subramanian, V.; Kuno, M.; Kamat, P. V. J. Am. Chem. Soc. 2006, 128, [36]. Sakai, N.; Ebina, Y.; Takada, K.; Sasaki, T. J. Am. Chem. Soc. 2004, 126, [37]. Nagaveni, K.; Hegde, S. M.; Ravishankar, N.; Subbanna, G. N. Madras, Giridhar. Langmuir 2004, 20, [38]. Guisbiers, G.; Overschelde, O. V.; Wautelet, M. Appl. Phys. Lett. 2008, 92, [39]. Beranek, R. Advances in Physical Chemistry 2011, 2011, [40]. Kalathil, S.; Khan, M. M.; Ansari, S. A.; Lee, J.; Cho, M. H. Nanoscale 2013, 5, [41]. Luttrell, T.; Halpegamage, S.; Tao, J.; Kramer, A.; Sutter, E.; Batzill, M. Sci. Rep.-UK 2014, 4, [42]. Klosek, S.; Raftery, D. J. Phys. Chem. B 2002, 105, [43]. Chen, D.; Jiang, Z.; Geng, J.; Wang, Q.; Yang, D. Ind. Eng. Chem. Res. 2007, 46, [44]. Teoh, W. Y.; Amal, R.; Mädler, L.; Pratsinis, S. E. Catal. Today 2007, 120, [45]. Fan, X.; Chen, X.; Zhu, S.; Li, Z.; Yu, T.; Ye, J.; Zou, Z. J. Mol. Catal. A-Chem. 2008, 284, [46]. Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Science 2001, 293, [47]. Yu, J. C.; Yu, J.; Ho, W.; Jiang, Z.; Zhang, L. Chem. Mater. 2002, 14,

44 [48]. Lindgren, T.; Mwabora, J. M.; Avendaño, E.; Jonsson, J.; Hoel, A.; Granqvist, C. G.; Lindquist, S. E. J. Phys. Chem. B 2003, 107, [49]. Valentin, C. D.; Pacchioni, G.; Selloni, A. Phys. Rev. B 2004, 70, [50]. Nakamura, R.; Tanaka, T.; Nakato, Y. J. Phys. Chem. B 2004, 108, [51]. Livraghi, S.; Paganini, M. C.; Giamello, E.; Selloni, A.; Di Valentin, C.; Pacchioni, G. J. Am. Chem. Soc. 2004, 126, [52]. Sathish, M.; Viswanathan, B.; Viswanath, R. P.; Gopinath, C. S. Chem. Mater. 2005, 17, [53]. Di Valentin, C.; Finazzi, E.; Pacchioni, G.; Selloni, A.; Livraghi, S.; Paganini, M.C.; Giamello, E. Chem. Phys. 2007, 339, [54]. Chen, X.; Burda, C.; J. Am. Chem. Soc. 2008, 130, [55]. Wang, J.; Tafen, D. N.; Lewis, J. P.; Hong, Z.; Manivannan, A.; Zhi, M.; Li, M.; Wu, N. J. Am. Chem. Soc. 2009, 131, [56]. Hensel, J.; Wang, G.; Li, Y.; Zhang, J. Z. Nano Lett. 2010, 10, [57]. Yang, G.; Jiang, Z.; Shi, H.; Xiao, T.; Yan, Z. J. Mater. Chem. 2010, 20, [58]. Serpone, N. J. Phys. Chem. B 2006, 110, [59]. Jing, L.; Xin, B.; Yuan, F.; Xue, L.; Wang, B.; Fu, H. J. Phys. Chem. B 2006, 110, [60]. Pan, X.; Yang, M. Q.; Fu, X.; Zhang, N.; Xu, Y. J. Nanoscale 2013, 5, [61]. Frank, S. N.; Bard, A. J. J. Am. Chem. Soc. 1977, 99, [62]. Schwarz, H. A.; Dodson, R. W. J. Phys. Chem. 1984, 88, [63]. Lawlee, D.; Serpone, N.; Meisel, D. J. Phys. Chem. 1991, 95, [64]. Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W. Chem. Rev. 1995, 95,

45 [65]. Mills, A.; Le Hunte, S. J. Photoch. Photobio. A 1997, 108, [66]. Gaya, U. I.; Abdullah, A. H. J. Photoch. Photobio. C 2008, 9, [67]. O Regan, B.; Grätzel, M. Nature 1991, 335, [68]. Hagfeldt, A.; Didriksson, B.; Palmqvist, T.; Lindström, H.; Södergren, S.; Rensmo, H.; Lindquist, S. E. Sol. Energ. Mat. Sol. Cells 1994, 31, [69]. Smestad, G.; Bignozzi, C.; Argazzi, R. Sol. Energ. Mat. Sol. Cells 1994, 32, [70]. Kay, A.; Grätzel, M. Sol. Energ. Mat. Sol. Cells 1996, 44, [71]. Huang, S. Y.; Schlichthörl, G.; Nozik, A. J.; Grätzel, M.; Frank, A. J. J. Phys. Chem. B 1997, 101, [72]. Bach, U.; Lupo, D.; Comte, P.; Moser, J. E.; Weissörtel, F.; Salbeck, J.; Spreitzer, H.; Grätzel, M. Nature 1998, 395, [73]. Park, N. G.; Van De Lagemaat, J.; Frank, A. J. J. Phys. Chem. B 2000, 104, [74]. Plass, R.; Pelet, S.; Krueger, J.; Grätzel, M.; Bach, U. J. Phys. Chem. B 2002, 106, [75]. Wang, Q.; Moser, J. E.; Grätzel, M. J. Phys. Chem. B 2005, 109, [76]. Nazeeruddin, M. K.; De Angelis, F.; Fantacci, S.; Selloni, A.; Viscardi, G.; Liska, P.; Ito, S.; Takeru, B.; Grätzel, M. J. Am. Chem. Soc. 2005, 127, [77]. Gao, F.; Wang, Y.; Shi, D.;, Zhang, J.; Wang, M.; Jing, X.; Humphry-Baker, R.; Wang, P.; Zakeeruddin, S. M.; Grätzel, M. J. Am. Chem. Soc. 2008, 130, [78]. Feldt, S. M.; Gibson, E. A.;, Gabrielsson, E.; Sun, L.; Boschloo, G.; Hagfeldt, A. J. Am. Chem. Soc. 2010, 132, [79]. Wen, P.; Xue, M.; Ishikawa, Y.; Itoh, H.; Feng, Q. ACS Appl. Mater. Interfaces 2012, 4,

46 [80]. Sewvandi, G. A.; Tao, Z.; Kusunose, T.; Tanaka, Y.; Nakanishi, S.; Feng, Q. ACS Appl. Mater. Interfaces 2014, 6, [81]. Yella, A.; Lee, H. W.; Tsao, H. N.; Yi, C.; Chandiran, A. K.; Nazeeruddin, M. K.; Diau, E. W. G.; Yeh, C. Y.; Zakeeruddin, S. M.; Grätzel, M. Science 2011, 334, [82]. Grätzel, M. Prog. Photovolt. Res. Appl. 2000, 8, [83]. Hagfeldt, A.; Grätzel, M. Acc. Chem. Res. 2000, 33, [84]. Grätzel, M. J. Photoch. Photobio. C 2003, 4, [85]. Grätzel, M. Nature 2001, 414, [86]. Grätzel, M. Inorg. Chem. 2005, 44, [87]. Mishra, A.; Fischer, M. K. R.; Büuerle, P. Angew. Chem. Int. Edit. 2009, 48, [88]. Papageorgiou, N.; Maier, W. F.; Grätzel, M. J. Electrochem. Soc. 1997, 144, [89]. Frank, A. J.; Kopidakis, N.; Lagemaat, J. V. D.; Coordin. Chem. Rev. 2004, 248, [90]. Gong, J.; Liang, J.; Sumathy, K. Renew. Sust. Energ. Rev. 2012, 16, [91]. Takagi, K.; Magaino, S.; Saito, H.; Aoki, T.; Aoki, D. J. Photoch. Photobio. C 2013, 14, [92]. Barnes, P. R. F.; Miettunen, K.; Li, X.; Anderson, A. Y.; Bessho, T.; Grätzel, M.; O Regan, B. Adv. Mater. 2013, 25, [93]. Zhang, S.; Yang, X.; Qin, C.; Numata, Y.; Han, L. J. Mater. Chem. A 2014, 2, [94]. Yum, J. H.; Baranoff, E.; Kessler, F.; Moehl, T.; Ahmad, S.; Bessho, T., Marchioro, A.; Ghadiri, E.; Moser, J. E.; Yi, C., Nazeeruddin, M. K.; Grätzel, M. Nat. Commun. 2012, 3,

47 [95]. Redmond, G.; Fitzmaurice, D.; Grätzel, M. Chem. Mater. 1994, 6, [96]. Lenzmann, F.; Krueger, J.; Burnside, S.; Brooks, K.; Grätzel, M.; Gal, D.; Rühle, S.; Cahen, D. J. Phys. Chem. B 2001, 105, [97]. Nang Dinh, N.; Bernard, M. C.; Hugot-Le Goff, A.; Stergiopoulos, T.; Falaras, P. C. R. Chimic 2006, 9, [98]. Tan, B.; Toman, E.; Li, Y.; Wu, Y. J. Am. Chem. Soc. 2007, 129, [99]. Robertson, N. Angew. Chem. Int. Edit. 2006, 45, [100]. Hagfeldt, A.; Boschioo, G.; Sun, L.; Kloo, L.; Pettersson, H. Chem. Rev. 2010, 110, [101]. Haque, S. A.; Palomares, E.; Cho, B. M.; Green, A. N. M.; Hirata, N.; Klug, D. R.; Durrant, J. R. J. Am. Chem. Soc. 2005, 127, [102]. Stergiopoulos, T.; Ghicov, A.; Likodimos, V.; Tsoukleris, D. S.; Kunze, J.; Schmuki, P.; Falaras, P. Nanotechnology 2008, 19, [103]. Lai, Y. H.; Chiu, C. W.; Chen, J. G.; Wang, C. C.; Lin, J. J.; Lin, K. F.; Ho, K. C. Sol. Energ. Mater. Sol. C. 2009, 93, [104]. Seo, D. W.; Sarker, S.; Nath, N. C. D.; Choi, S. W.; Ahammad, A. J. S.; Lee, J. J.; Kim, W. G. Electrochim. Acta 2010, 55, [105]. Koide, N.; Han, L. Rev. Sci. Instrum. 2004, 75, [106]. Bisquert, J. J. Phys. Chem. B 2002, 106, [107]. Han, L.; Koide, N.; Chiba, Y.; Mitate, T. Appl. Phys. Lett. 2004, 84, [108]. Han, L.; Koide, N.; Chiba, Y.; Islam, A.; Mitate, T. C. R. Chimic 2006, 9, [109]. Adachi, M.; Sakamoto, M.; Jiu, J.; Ogate, Y.; Isoda, S. J. Phys. Chem. B 2006, 110, [110]. Bijarbooneh, F. H.; Zhao, Y.; Kim, J. H.; Sun, Z.; Malgras, V.; Aboutalebi, S. H.; Heo, Y. U.; Ikegami, M.; Dou, S. X. J. Am. Ceram. Soc. 2013, 96, 42

48 [111]. Wen, P.; Itoh, H.; Tang, W.; Feng, Q. Lamgmuir. 2007, 23, [112]. Yang, H. G.; Sun, C. H.; Qiao, S. Z.; Zou, J.; Liu, G.; Smith, S. C.; Cheng, H. M.; Lu, G. Q. Nature 2008, 453, [113] Wu, B.; Guo, C.; Zheng, N.; Xie, Z.; Stucky, G. D. J. Am. Chem. Soc. 2008, 130, [114]. Hore, S.; Vetter, C.; Kern, R.; Smit, H.; Hinsch, A. Sol. Energ. Mater. Sol. C. 2006, 90, [115]. Chou, C.; Guo, M.; Liu, K.; Chen, Y. Appl. Energ. 2012, 92, [116]. Baraton, M. I. Recent Patents on Nanotechnology 2012, 6, [117]. Wen, P.; Itoh, H.; Tang, W.; Feng, Q. Lamgmuir. 2007, 23, [118]. Miao, L.; Tanemura, S.; Toh, S.; Kaneko, K.; Tanemura, M. J. Cryst. Growth 2004, 264, [119]. Cheng, H.; Ma, J.; Zhao, Z.; Qi, L. Chem. Mater. 1995, 7, [120]. Wen, P.; Itoh, H.; Tang, W.; Feng, Q. Microporous Mesoporous Mat. 2008, 116, [121]. Wen, P.; Ishikawa, Y.; Itoh, H.; Feng, Q. J. Phys. Chem. C 2009, 113, [122]. Du, Y,; Feng, Q.; Chen, C.; Tanaka, Y.; Yang, X. ACS Appl. Mater. Interfaces 2014, 6, [123]. Lazzeri, M.; Vittadini, A.; Selloni, A. Phys. Rev. B: Condens. Matter 2001, 63, [124]. Barnard, A. S.; Curtiss, L. A. Nano Lett. 2005, 5, [125]. Xu, H.; Ouyang, S.; Li, P.; Kako, T.; Ye, J. ACS Appl. Mater. Interfaces 2013, 5, [126]. Penn, R. L.; Banfield, J. F. Geochim. Cosmochim. Acta 1999, 63,

49 [127]. Amano, F.; Yasumoto, T.; Prieto-Mahaney, O. O.; Uchida, S.; Shibayama, T.; Ohtani, B. Chem. Commun. 2009, 17, [128]. Pan, J.; Liu, G.; Lu, G.; Cheng, H. Angew. Chem. Int. Ed. 2011, 50, [129]. Pan, J.; Wu, X.; Wang, L.; Liu, G.; Lu, G.; Cheng, H. Chem. Commun. 2011, 47, [130]. Han, X.; Kuang, Q.; Jin, M.; Xie, Z.; Zheng, L. J. Am. Chem. Soc. 2009, 131, [131]. Liu, G.; Yang, H. G.; Wang, X. W.; Cheng, L. N.; Lu, H. F.; Wang, L.Z.; Lu, G. Q.; Cheng, H. M. J. Phys. Chem. C 2009, 113, [132]. Liu, M.; Li, H.; Zeng, Y.; Huang, T. Appl. Surf. Sci. 2013, 174, [133]. Wen, P.; Tao, Z.; Ishikawa, Y.; Itoh, H.; Feng, Q. Appl. Phys. Lett. 2010, 97, [134]. Wu, J.; Hao, S.; Lin, J.; Huang, M.; Huang, Y.; Lan, Z.; Li, P. Cryst Growth Des. 2008, 1, [135]. Wang, Y.; Yang, W.; Shi, W. Ind. Eng. Chem. Res. 2011, 50, [136]. Choy, J. H.; Lee, H. C.; Jung, H.; Huang, S. J. J. Mater. Chem. 2001, 11, [137]. Corcoran, D. J. D.; Tunstall, D. P.; Irvine, J. T. S. Solid State Ionics 2000, , [138]. El-Naggar, I. M.; Mowafy, E. A.; Ali, I. M.; Aly, H. F. J. Int. Adsorp. Soc. 2002, 8, [139]. Bavykin, D. V.; Walsh, F. C. Eur. J. Inorg. Chem. 2009, [140]. Andersson, S.; Wadsley, A. D. Acta Chem. Scand. 1961, 15, [141]. Andersson, S.; Wadsley, A. D. Acta Cryst. 1961, 14, [142]. Roth, R. S.; Parker, H. S.; Brower, W. S. Mat. Res. Bull. 1973, 8, [143]. Sasaki, T.; Watanabe, M.; Michiue, Y.; Komatsu, Y.; Izumi, F.; Takenouchi, S. 44

50 Chem. Mater. 1995, 7, [144]. Komarneni, S.; Rajha, R. K.; Katsuki, H. Mater. Chem. Phys. 1999, 61, [145]. Wilson, G. J.; Will, G. D.; Frost, R. L.; Montgomery, S. A. J. Mater. Chem. 2002, 12, [146]. Wilson, G. J.; Matijasevich, A. S.; Mitchell, D. R. G.; Schulz, J. C.; Will, G. D. Langmuir 2006, 22,

51 CHAPTER II Synthesis of Titanium Dioxide Nanocrystals from H 1.07 Ti 1.73 O 4 Layered Titanate Nanosheets by Normal Hydrothermal Process and Their Dye-Sensitized Solar Cell Performance 2.1. Introduction Dye-sensitized solar cells (DSSCs) based on mesoporous nanocrystalline TiO 2 film have attracted considerable attention due to their potentially low cost, relatively high energy conversion efficiency, safety, and non-pollution. 1-3 Common DSSCs consist of three main parts including a sensitized TiO 2 or ZnO photoelectrode, a Pt counter electrode, and a redox electrolyte. 4, 5 The TiO 2 photoelectrode is the most important part that strongly affects the DSSCs performances. The TiO 2 photoelectrode is fabricated by a coating conducting glass plate with a mesoporous TiO 2 nanoparticle film, and adsorption of sensitizer dye molecules on the nanoparticles surfaces. Therefore, a lot of effort has been made on the synthesis of TiO 2 nanoparticles to develop the high performance TiO 2 photoelectrodes It is commonly considered that anatase-type TiO 2 nanoparticles with small particle size about 20 nm and high crystallinity exhibit high performance. 11 Usually spherical nanocrystals are used to fabricate the TiO 2 photoelectrode, while addition fibrous and rod-like particles into the spherical particles can increase the conductivity of TiO 2 film, which improves cell performance Since dye adsorption reaction occurs on crystal surfaces, it is easy to be understood that the adsorption reaction would be strongly dependent on the crystal 46

52 facet on the particle surface. However, to study the effect of crystal facet on the DSSCs performances, it is necessary to prepare the TiO 2 nanoparticles preferentially exposing a specific crystal facet on the surface but it is not easy. Wen et al. have reported a pioneering study on the hydrothermal soft chemical synthesis of anatase TiO 2 nanocrystals with specific crystal facet on the surface, and found that {010}-faceted anatase TiO 2 nanocrystals exhibit higher photocatalytic activity than the normal spherical nanocrystals without specific facet on the surface for the first time. 15, 16 The results spur researchers to study the synthesis of TiO 2 nanocrystals with specific crystal facet, and also the surface energy and photocatalytic activity of the crystal facets Wu et al. have reported syntheses of {001}- and {010}-faceted anatase nanocrystals and their photocatalytic performances in degradation of methyl orange, and found that the {010}-faceted nanocrystals exhibit higher photocatalytic activity than the {001}-faceted nanocrystals and commercial P25 sample. 24 Han et al. have reported that the {001}-faceted anatase nanosheets exhibit the higher photocatalytic activity than the P25 sample. 25 The Wulff construction and theoretically calculated surface energy predicts that the surface energy of anatase increases in an order of {101} facet (0.44 J/m 2 ) < {010} facet (0.53 J/m 2 ) < {001} facet (0.90 J/m 2 ) < {111} facet (1.61 J/m 2 26, 27 ). The higher photocatalytic activity of the {010} facet than that of {001} facet and {101} facet is owing to the superior electronic band structure of {010} facet for the photocatalytic reactions. 28 However, up to now, most studies on the TiO 2 facets have focused on their synthesis, surface energy, and photocatalytic activity. Recently, Wen et al. have reported studies on the effects of the crystal facet on DSSC performance and sensitizer dye adsorption behavior for the first time. 29, 30 The results suggested that the dye adsorption behavior is strongly affected by the crystal surface properties of the 47

53 TiO 2 nanocrystals and the dye molecules are strongly adsorbed on {010} facet which results enhancement of short-circuit current density. After that, other studies have also 7, 28 confirmed that the crystal facets on the surfaces affect the DSSCs performances. Very recently, it has been reported that {001}-faceted anatase TiO 2 nanocrystals exhibit a high performance for DSSCs due to the enhancement in light scattering effect and suppressed electron recombination, 31 and also the strong adsorption of dye molecules on the {001} facet facilitates the electron transport from dye molecules to the conduction band of TiO 2. 32, 33 Unfortunately, only limited numbers of studies have been reported on the facet effect on the DSSCs performance, maybe due to the synthesis processes are difficult to provide enough amounts of nanocrystals with the specific facet for the DSSCs characterizations, and/or the nanocrystals sizes are too large for high performance DSSCs. In this chapter, we describe synthesis of morphology controllable {010}-faceted anatase TiO 2 nanocrystals from exfoliated layered titanate nanosheets with lepidocrocite-like structure. Regular rhombic, tetragonal, and leaflike anatase nanocrystals, and platelike anatase mesocrystals constructed from oriented anatase nanocrystals can be achieved using this process. In this process, n-propylamine (PA) is chose as the exfoliating agent because it is much cheaper than other exfoliating agents, such as tetrabutylammonium hydroxide and tetramethylammonium hydroxide, therefore, the mass production of the TiO 2 nanoparticles is possible using PA exfoliating agents. The formation reaction mechanism of the anatase nanocrystals and mesocrystals is also investigated by a nanostructural analysis. The reaction mechanism is significant to understand the phase conversion and particle morphology evolution in the formation process of the {010}-faceted TiO 2 nanoparticles from layered titanate nanosheets, which would give a guide for controlling particle morphology of the TiO 2 nanoparticles with specific facet on the surface. Furthermore, 48

54 we characterized the DSSCs performances of the {010}-faceted anatase nanocrystals in detail, and compare their performances with which of commercial P25 nanocrystals containing [111]-faceted (facet vertical to [111]-direction) anatase nanocrystals and spherical anatase nanocrystals without specific facet on the surface. The {010}-faceted anatase nanocrystals exhibit the highest short-circuit current density that results the highest energy conversion efficiency. The results reveal the significant effect of the crystal facet on the DSSC performance Experimental Section Preparation of Layered Titanate Nanosheet Colloidal Solutions The layered titanate nanosheet colloidal solutions were prepared from layered titanate K 0.8 Ti 1.73 Li 0.27 O 4 (KTLO). A mixture of 5.1 g of KOH, 0.6 g of LiOH H 2 O, 6.9 g of TiO 2 (anatase form) and 25 ml of distilled water were sealed in Hastelloy-C-lined vessel with internal volume of 45 ml, and then heated at 250 o C for 24 hrs under stirring conditions. The hydrothermally treated samples were washed with distilled water up to neutral and dried at room temperature to obtain the layered titanate KTLO as the precursor. The KTLO sample (10 g) was acid-treated with a 0.2 M HNO 3 solution (1L) for 24 hrs under stirring conditions to exchange K + ions and Li + ions with H + ions to obtain an H + -form layered titanate H 1.07 Ti 1.73 O 4 (HTO). The acid treatment was repeated twice to complete the ion-exchange reaction. The HTO layered titanate was washed with distilled water several times and dried using a freeze drier. The layered titanate nanosheet colloidal solution was prepared by treating HTO sample (10 g) in 0.1 M n-propylamine (PA) (1L) solution under stirring conditions at room temperature for 24 hrs, and the nanosheet colloidal solution was named as PA-HTO. 49

55 Preparation of TiO 2 Nanoparticles from PA-HTO Solution TiO 2 nanoparticles were prepared by hydrothermal treatment of the PA-HTO nanosheet colloidal solution after adjusting to a desired ph value with a 3 M HCl solution in a ph range below 11.7, and a 1M KOH solution in a ph range above about The 40 ml of ph adjusted PA-HTO solution was sealed in a Teflon-lined stainless steel vessel with internal volume of 80 ml, and then hydrothermally reacted at a desired temperature for 24 hrs. After the hydrothermal treatment, the samples were washed with distilled water several times, and finally dried using a freeze drier. The obtained TiO 2 sample was named as PA-X-Y, where X and Y were the temperature of hydrothermal treatment and ph value of the nanosheet solution used for the hydrothermal treatment, respectively Preparation of TiO 2 Nanoparticle Paste and TiO 2 Photoelectrodes A TiO 2 nanoparticle paste was prepared by mixing 0.5 g of TiO 2 nanoparticle sample, 2.5 g of ethanol, 2.0 g of α-terpineol, 1.4 g of 10 wt% solution of ethyl-cellulose 10 (8-14mPas), and 1.1 g of 10 wt% solution of ethyl-cellulose 45 (45-65mPas). The mixture was dispersed by ultrasonication for 30 min, and then ball-milling for 72 hrs. After ball-milling, the ethanol was removed from the mixture using a rotary-evaporator. The obtained pastes contain 18 wt% TiO 2 nanoparticles, 9 wt% ethyl-cellulose and 73 wt% α-terpineol. P25 and ST20 nanocrystals pastes were also prepared using similar method. The P25 nanocrystals sample was purchased from Degussa. The ST20 nanocrystals sample was prepared by hydrothermal treatment of ST01 (anatase phase with particle size of about 7 nm) sample purchased from Ishihara Sangyo at 200 o C and 12 hrs. The TiO 2 nanoparticles paste was applied repetitively to the fluorine-doped tin 50

56 oxide (FTO) conducting glass plate as follows. The FTO glass plate was cleaned in distilled water and acetone by ultrasonication for 10 min, respectively. Then the FTO glass plate was dipped in 0.1M titanium tetraisopropoxide (TTIP) solutions for 1 min and washed with distilled water and ethanol, dried at room temperature and calcined at 480 o C for 1 h to coat the FTO glass surface with a dense TiO 2 thin film. The TiO 2 paste was coated (10 10 mm) on the TTIP-treated FTO glass plates by screen printing technique and kept in an ethanol box until the TiO 2 film surface smoothly and then dried at 120 o C for 10 min. This process was repeated to obtain a desired thickness of TiO 2 film. After the TiO 2 paste coating, the TiO 2 film was calcined at 450 o C for 30 min to obtain a single-layer TiO 2 porous electrode. A double-layer porous TiO 2 electrode was fabricated by coating PA nanoparticle paste onto the single-layer TiO 2 electrode as a light scattering layer using the screen printing technique process same as the single-layer TiO 2 electrode. The TiO 2 porous electrodes were dipped in the 0.1M TTIP solution for 1 min, and washed with distilled water and ethanol, dried at room temperature and calcined at 480 o C for 1 h again. After cooling to 80 o C, the TiO 2 porous electrodes were soaked in a M N719 dye (di-tetrabutylammonium cis-bis(isothiocyanato)bis(2,2'-bipyridyl-4,4'-dicarboxylato) ruthenium(ii)) solution for 24 hrs at room temperature, and then wash with a t-butyl alcohol and acetonitrile (v/v=50%:50%) mixed solvent Fabrication and Characterization of Dye-Sensitized Solar Cell The DSSCs were comprised of a dye-adsorbed TiO 2 electrode as an anode, a Pt-coated FTO glass as a cathode, and an electrolyte solution between the anode and the cathode. The electrolyte solution contains 0.1M LiI, 0.01M I 2, 0.6M of 1-butyl-3-n-propylimidazolium iodide (BMII), 0.4M 4-tert-butylpyridine (TBP) and 51

57 0.1M guanidine thiocyanate (GT) in acetonitrile and valeronitrile (v/v=85%:15%). The photocurrent voltage characteristic curves for the DSSCs were measured using a Hokuto-Denko BAS100B electrochemical analyzer under irradiation with simulated sunlight of AM 1.5 (100mW/cm 2 ), using a sunlight simulator (YSS-E40, Yamashita Denso). A light-passing mask was fixed on the surface of FTO glass of the anode to set the effectively irradiating area on the cell as 0.25 cm 2. Electrochemical impedance spectroscopic analyses (EIS) were performed with an impedance analyzer (Solartron SI 1260) under the dark condition in a two-electrode configuration. The impedance spectra were recorded in a frequency range of 0.1 Hz to 1 MHz with alternate current (AC) amplitude of 10 mv at an applied direct current (DC) bias of -0.7 V Adsorption of N719 Dye on TiO 2 Nanoparticles Adsorption experiment of dye was carried out by a batch method. A TiO 2 nanocrystal sample (10 mg) was added into an ethanol solution (5 ml) of desired concentration N719 dye and stirred at room temperature for 72 hrs. After the adsorption, the liquid phase was separated from the solid phase by centrifuge under rpm for 10 min, and then the concentration of the remaining N719 dye in the liquid phase was analyzed using a SHIMADZU UV-2450 spectrophotometer. The amount of the N719 dye adsorbed was determined from the change of dye concentration in ethanol solution before and after the absorption. The TiO 2 nanocrystal samples were calcined at 450 C for 30 min before the adsorption experiment Physical Analysis The crystal structure of the sample was investigated using a powder X-ray 52

58 diffractometer (Shimadzu, Model XRD-6100) with Cu Kα (λ= nm) radiation. The size and morphology of the samples were characterized by field emission scanning electron microscopy (FE-SEM) (Hitachi, Model S-900). Transmission electron microscopy (TEM) observation and selected-area electron diffraction (SAED) were performed on a JEOL Model JEM-3010 system at 300 kv. Nitrogen gas adsorption was carried out on a QUANTACHROME AUTOSORB-1-MP apparatus. The specific surface area of TiO 2 nanoparticles was calculated from the adsorption data using the Brunauer Emmett Teller (BET) method (Autosorb-1) Results and Discussion Synthesis of TiO 2 Nanoparticles from PA-HTO Nanosheet Solution Figure 2.1. XRD patterns of PA-HTO sample and products obtained by hydrothermal treatment of PA-HTO nanosheet solution at 120 o C for 24 hrs. (a) PA-HTO, (b) PA , (c) PA , (d) PA , and (e) PA : PA-HTO phase; : rutile phase; : anatase phase; : K-HTO phase. The PA-HTO nanosheet colloidal solutions with different ph values were 53

59 hydrothermally treated at various temperatures to synthesize TiO 2 nanoparticles. Figure 2.1 shows XRD patterns of PA-HTO nanosheet sample and products obtained at 120 o C. PA-HTO has a lepidocrocite-like layered structure with strong (020) peak at d=1.07 nm. It indicates that the layered structure has a basal spacing of 1.07 nm, which is larger than 0.92 nm of H 1.07 Ti 1.73 O 4 (HTO). 16 This suggests that CH 3 (CH 2 ) 2 NH 2 (PA) molecules are intercalated into the interlayer of HTO, and form PA-H + with H + in the interlayer. The intercalation of PA causes the swelling of the basal spacing of HTO layered titanate, and finally exfoliation reaction of the HTO layered structure into its structural elementary layers, PA-HTO nanosheets. A single phase of rutile-type of TiO 2 (JCPDS File No , tetragonal system, a=b=0.459 nm, c=0.296 nm) was formed at ph 0.5, and a single phase of anatase-type of TiO 2 (JCPDS File No , tetragonal system, a=b=0.380 nm, c=0.951 nm) was formed at ph 3.0. A mixture of PA-HTO phase and anatase phase was formed at ph 11.5, which means that PA-HTO nanosheets are partially changed to anatase TiO 2. PA sample has a layered structure which corresponds to K 0.8 Ti 1.73 Li 0.27 O 4 (KTLO) structure (JCPDS File No , orthorhombic system, a=0.382 nm, b=1.559 nm, c=0.298 nm) with a basal spacing of 0.87 nm. It can be explained by a restacking reaction of HTO nanosheets in KOH solution. The restacking reaction accompanies an ion-exchange reaction of K + and PA-H + in PA-HTO interlayer. The PA-HTO nanosheet colloidal solutions were also hydrothermally treated at other temperatures conditions, and the XRD patterns of products obtained under different conditions are shown in Figure 2.2. The dependences of the products on the reaction temperature and ph value are summarized in Figure 2.3. The products are strongly dependent on the ph value of the PA-HTO nanosheet solution. In a range of ph < 1, the PA-HTO nanosheets are transformed to rutile phase. In a range of 1 < ph < 13, anatase phase is formed, where the temperature of formation of anatase phase 54

60 increases with increasing the ph value. The HTO layered phase is stable in a range of ph > 13, where K-HTO is formed. Figure 2.2. XRD patterns of products obtained by hydrothermal treatment at 100 o C, 135 o C and 155 o C for 24 hrs. : PA-HTO phase, : rutile phase, : anatase phase and : K-HTO phase, respectively. Figure 2.4 shows FE-SEM images of products obtained by the hydrothermal reaction of PA-HTO solutions. PA sample (Figure 2.4(a)) exhibits platelike particle morphology with small amount of nanoparticles on the surface. This sample has a mixed phase of PA-HTO and anatase (Figure 2.3). The small nanoparticles correspond to anatase phase, and platelike particles maybe correspond to mixture of PA-HTO nanosheet and anatase phases. Because the thickness about 50 nm of the platelike particles is much larger than that PA-HTO nanosheet (about 1 nm), a restacking reaction of PA-HTO nanosheets occurs under the hydrothermal conditions. The nanosheetlike particles and leaflike nanoparticles were observed in PA The PA-HTO nanosheets are partially transformed to leaflike anatase nanoparticles. 55

61 Figure 2.3. Dependences of the products on the reaction temperature and ph value. : rutile TiO 2 phase; : anatase TiO 2 phase; : PA-HTO and anatase TiO 2 phase; : K-HTO phase. In PA , platelike particles of anatase phase were formed and the platelike particles are constructed from small nanoparticles with size of about 20 nm. In PA , regular rhombic (70%) and tetragonal (30%) nanoparticles of anatase phase with size about 20 nm were formed. In PA , the rhombic and tetragonal anatase nanoparticles with size about 30 nm were also formed, but the content of tetragonal nanoparticles increased to about 70% and the content of rhombic nanoparticles reduced to about 30%. In PA , the regular tetragonal anatase nanoparticles with size of about 60 nm were formed. These results reveal that in the ph range of 3.5 to 9.5 the percentage of tetragonal nanocrystal and its crystal size increase with increasing of ph value. In PA , leaflike anatase nanoparticles with size of about 300 nm in length and 30 nm in width were obtained. In PA , sheetlike particles with thickness of a size about 1 µm in width and 50 nm in thickness were obtained. The sheetlike particles were formed by restacking HTO nanosheets in KOH solution and kept the lepidocrocite-like layered structure after the hydrothermal reaction (see Figure 2.2). 56

62 Figure 2.4. FE-SEM images of (a) PA , (b) PA , (c) PA , (d) PA , (e) PA , (f) PA , (g) PA , (h) PA and (i) PA samples. In PA , rhombic and tetragonal nanoparticles of anatase phase were formed, but the particle sizes were larger than which in PA , suggesting that the crystal growth occurs under the high temperature conditions. The FE-SEM results reveal that anatase nanocrystals can be prepared by hydrothermal treatment of HTO nanosheet solution, and the crystal size and morphology of anatase nanocrystals can controlled by reaction temperature and solution ph value. 57

63 TEM Study on PA-HTO Nanosheets and TiO 2 Nanoparticles Figure 2.5. TEM images and SAED patterns of (a), (b) HTO, (c), (d) PA , (e) PA , (f) PA , (g) PA , and (h) PA samples. The nanostructures of the hydrothermal reaction products were investigated using TEM and SAED. Figure 2.5 shows TEM images and SAED patterns of PA-HTO nanosheets and hydrothermally-treated samples. PA-HTO nanosheet has nanosheetlike particle morphology and shows a single crystal SAED pattern, where the b-axis is vertical to the basal plane of nanosheet. PA sample consists of a mixture of anatase and PA-HTO (Figure 2.1(d)). It has nanosheetlike particle morphology and shows simultaneously two sets of the SAED spot patterns corresponding to anatase phase with diffractions of (101) plane (d=0.38 nm) and (200) plane (d=0.185 nm), and HTO phase with diffractions of (400) plane (d=0.088 nm) and (004) plane (d=0.073) in one nanosheetlike particle. This result suggests the HTO nanosheet is partially transformed to anatase phase, namely HTO and anatase phases coexist in one nanosheetlike particle, which reveals that the HTO structure is transformed to anatase structure by a topotactic mechanism under the hydrothermal reaction conditions. There is a crystallographic relationship between HTO structure 58

64 and anatase structure in the topotactic structural transformation reaction. The [010]-direction of anatase phase corresponds to the [010]-direction of HTO phase that is vertical to the basal plane of the nanosheet. And the a- and c-axis orientations of HTO structure are transformed to the a- and c-axis orientations of anatase structure by rotating an angle of 60 o on the (010) plane. The result reveals that {010}-faceted nanosheetlike anatase particles can be formed from the PA-HTO nanosheets by the hydrothermal reaction. Although the platelike particles of PA sample were constructed from anatase nanocrystals with a size of about 20 nm, it shows a SAED pattern similar to the single crystal of anatase (Figure 2.5(e)). It suggests that almost all the anatase nanocrystals in one platelike particle align in the same crystal-axis orientation directions, namely the platelike anatase particles are a mesocrystal. 34 Furthermore, the platelike anatase mesocrystal shows a [010]-axis orientation, and the basal plane of the platelike particle corresponds to the (010) facet that is vertical to the [010] orientation. The orientation of the platelike mesocrystals corresponds to the orientation of the nanosheetlike anatase particles, revealing that the platelike anatase mesocrystals are also formed by the topotactic structural transformation mechanism. Since the thickness of about 50 nm of the mesocrystals are much larger than the that of HTO nanosheet (about 1 nm), a restacking reaction of HTO nanosheets occurs in the low ph (<2.5) range, which accompanies an ion-exchange of H + with PA-H + in PA-HTO interlayer. The lattice images of the rhombic, tetragonal, and leaflike anatase nanocrystals synthesized in the ph range of are clearly observed (Figures 2.5(f)-(h)), indicating that these nanoparticles are the single nanocrystals with high crystallinity. About 90% of these anatase nanocrystals present the {010} facet on their basal plane. In the HR-TEM image of the rhombic nanocrystal, the lattice fringes of (101) 59

65 (d=0.352 nm) and (004) (d=0.231) are observed, where two sides of the rhombic nanocrystal are parallel to (101) facet, other two sides are parallel to (-101) facet, and two corners are 68 o and 112 o, respectively. The lattice fringes of (101) facet (d=0.352 nm) is observed in the HR-TEM image of the tetragonal nanocrystal, and diffraction spots of (101), (200), and (004) planes are observed in its SAED pattern. The result reveals that two sides of the tetragonal nanocrystal are parallel to the (101) facet and other two sides parallel to facet vertical to (101) facet. The lattice fringes of (-101) and (101) facets are observed in the HR-TEM image of the leaflike anatase nanocrystal. The axis direction of the leaflike nanocrystal corresponds to the c-axis of the anatase structure Transformation Reaction Mechanism from PA-HTO Nanosheets to TiO 2 Nanocrystals On the basis of the above results, we propose a reaction model for the formation of anatase nanocrystals from PA-HTO nanosheets in the hydrothermal reaction process, as shown in Figure 2.6. In the ph range of 1 < ph < 3, PA-H + ions adsorbed on HTO nanosheet surface are ion-exchanged with H + ions in the acidic solution, and then the restacking reaction of the HTO nanosheets occurs, which causes formation of platelike HTO particles. Under the hydrothermal conditions, the platelike HTO particles are transformed to the platelike anatase particles by an in situ topotactic structural transformation reaction, which retains the platelike particle morphology. In the topotactic structural transformation reaction, the TiO 6 octahedra in the HTO structure shift from the positions of the lepidocrocite-like layered structure to the positions of anatase structure. There is a crystallographic relationship between the structure of the precursor and the structure of the product in the topotactic structural 60

66 transformation reaction In the present case, the [010]-direction of lepidocrocite-like layered structure corresponds to the [010]-direction of the anatase phase, and the a- and c-axes of layered structure transform to a- and c-axes of anatase structure by rotating an angle of 60 o on the (010) plane, as shown in Figure 2.5 (d). Therefore, {010}-faceted platelike anatase particles are formed by this reaction. The cracks are formed in the anatase platelike anatase particles, and split the platelike anatase particles to the small anatase nanoparticles by dissolution-deposition reaction similar to normal hydrothermal reaction, but it retains the platelike particle morphology and the {010}-faceted orientation, resulting formation of {010}-faceted anatase mesocrystal (Figure 2.5(e)). Figure 2.6. Transformation reaction mechanism from PA-HTO nanosheets to TiO 2 nanocrystals. 61

67 In the ph range of 3 < ph < 13, firstly the PA-HTO nanosheets are transformed to anatase phase by the similar topotactic structural transformation reaction described above, resulting formation of nanosheetlike anatase particles with (010) facet on the nanosheetlike particle surface, and then the nanosheetlike anatase particles are split into the rhombic, tetragonal, and leaflike nanocrystals by formation of cracks on the nanosheetlike particles owing to the dissolution-deposition reaction under hydrothermal conditions. At around ph 3.5, the cracks are formed mainly along (101) and (-101) planes, and therefore, the main product is the rhombic nanocrystals with two sides parallel to (101) facet; other two sides parallel to (-101) facets, therefore, two corners of 68 o and 112 o are formed. The angles of 68 o and 112 o are consistent with the angles between (101) and (-101) planes (Figure 2.5(f)). The crack formation along (-101) plane decreases and that along the plane vertical to (101) facet increases with increasing the ph value in the ph range of Therefore, the percentage of rhombic nanocrystals decreases and that of tetragonal nanocrystals increases in the products. The two sides of tetragonal nanocrystals are consistent with (101) facet; other two sides are consistent with the facet vertical to (101) facets. In the range of 10 < ph < 13, the cracks are formed mainly along the (100) and (001) planes, which slit the nanosheetlike anatase particle to rodlike anatase nanocrystals with axis of rodlike crystal directing to the c-axis of anatase structure. And then the leaflike anatase nanocrystals are formed by dissolution of the edges of the rodlike nanocrystals. The above results reveal that two types of reactions can occur simultaneously in the formation reaction processes of the anatase nanocrystals from PA-HTO nanosheets. 16, One is the in situ topotactic transformation reaction, in which the structure of PA-HTO nanosheets is transformed to anatase structure, and this reaction retains the morphology of the precursor after the reaction. Another is the dissolution-deposition reaction on the surface of the PA-HTO nanosheets, which is similar to the normal 62

68 hydrothermal reaction, and the nanocrystal morphology and size can be changed by this reaction. In PA-HTO nanosheet hydrothermal reaction system, rutile phase is also formed in the ph range below 1. The rutile phase shows irregular particle morphology without correlation with that of PA-HTO nanosheet precursor (see Figure 2.7). This result suggests that the rutile phase is formed by dissolution-deposition reaction mechanism similar to the normal hydrothermal reaction, but not by topotactic transformation reaction. It has been reported that rutile phase is preferentially formed in the acidic solution under hydrothermal conditions Figure 2.7. FE-SEM images of PA samples. This sample is single rutile phase DSSC Performance of {010}-Faceted TiO 2 Nanoparticles We used five kinds of anatase nanocrystal samples, PA , PA , PA , PA , and PA , for the DSSC study. These samples have {010} facet on the basal plane of the nanocrystals, but their crystal morphologies and particle sizes are different. Current-voltage characteristics and cell parameters for DSSCs fabricated using these anatase nanocrystal samples are shown in Figure 2.8 and Table 2.1, respectively. For comparison, the thicknesses of the TiO 2 porous films were kept almost same around 8 µm. The short-circuit current density (J sc ) increases in an order of PA < PA < PA < PA < PA , 63

69 which is consistent with the decreasing order of the crystal size. While the open-circuit voltage (V oc ) increases in an order consistent with the increasing of the crystal size. The fill factor (ff) shows a similar trend to the V oc values. The energy conversion efficiency (η) increases in the order similar to the increasing order of the J sc values, namely PA sample shows the highest performance for DSSC. Because the smaller particles have larger surface area, they can absorb more dye molecules on the surface, which enhance the J sc value. Figure 2.8. Current-voltage characteristic curves of DSSCs fabricated using anatase nanocrystals synthesized by hydrothermal treatment of PA-HTO nanosheet solutions. Table 2.1. Cell performance parameters and EIS parameters of DSSCs fabricated using anatase nanocrystals synthesized by hydrothermal treatment of PA-HTO nanosheet solutions. Sample Thickness J sc V oc ff η R s R rc R Pt (μm) (ma/cm 2 ) (V) (%) (Ω) (Ω) (Ω) PA PA PA PA PA

70 An electrochemical impedance spectroscopy (EIS) study was carried out on the DSSCs (see Figure 2.9). The EIS parameters of the electron transport resistance (R s ), the charge transfer resistances related to recombination of electrons at the TiO 2 / electrolyte interface (R rc ) and reduction reaction of I 3 - at the Pt counter/electrode interface (R Pt ) can been evaluated from the EIS spectra, as shown in Table The R rc value decreases with increasing crystal size of TiO 2 because of increase in the surface area of TiO 2 electrode contacting to the electrolyte solution, which enhance electron recombination from TiO 2 conduction band to the I - 3 ions in the electrolyte solution. The electron recombination at TiO 2 / electrolyte interface lowers the V oc 42, 43 value. Figure 2.9. Electrochemical impedance spectrum of DSSCs prepared using the products obtained by hydrothermal treatment of PA-HTO nanosheet solution at 135 o C for 24 hrs N719 Dye Adsorption Behavior on TiO 2 Nanocrystals with Different Facets on Surfaces To study the effect of the crystal facets on the DSSCs performances, we used three kinds of TiO 2 nanocrystal samples, PA , P25, and ST20 which have similar 65

71 particle sizes of about 20 nm and different crystal facets on the nanocrystals surfaces. The PA sample contains mainly rhombic nanocrystals which have the two basal planes corresponding to {010} facet and four planes corresponding to {101} facets (Figure 2.6). The PA sample is used as the {010}-faceted nanocrystal sample because of its highest DSSC performance (η) in the {010}-faceted nanocrystals samples synthesized from PA-HTO nanosheets. P25 is commercial TiO 2 nanocrystal sample commonly used in DSSCs, which is a mixture of anatase phase (80%) and rutile phase (20%). Although P25 is commonly used in the DSSC and photocatalytic studies, the nanostructures, especially the facet exposes on the nanocrystal surface, are still unclear yet. Therefore, a nanostructural study on P25 nanocrystals was carried out before the DSSC study. Figure FE-SEM images of (a) P25 and (b) ST20 samples, (c) HR-TEM image and (d) model of crystal facets on surface of tetragonal anatase nanocrystal of P25 sample. 66

72 Figure 2.10 shows the FE-SEM and TEM images of P25 nanocrystals. This sample contains about 80% of nanocrystals with a size of about 20 nm and about 20% of nanocrystals with a size about 80 nm. The nanocrystals with a size of about 20 nm are anatase phase, while which with a size of about 80 nm are rutile phase (Figure 2.10 (a)). There are two kinds of particle morphologies for the anatase nanocrystals. One is near-spherical morphology (content of about 40%) and other is tetragonal morphology (content of about 40%). The near-spherical nanoparticles have not specific facet on the crystal surface, while tetragonal nanocrystals have specific facet on the crystal surface. In the HR-TEM image of the tetragonal nanocrystal, the anatase lattice fringes of (101) and (011) facets (d=0.352 nm) are observed (Figure 2.10 (c)). The HR-TEM image reveals that the basal plane of the tetragonal particle correspond to a facet vertical to [111]-direction (we call it as [111]-facet), and other four planes correspond to the {101} facet as shown in Figure 2.10 (d). This is the first time to determine the facets on the P25 nanoparticles surfaces. The ST20 sample is anatase phase with spherical particle morphology without specific facet on the surface and has a crystal size of about 20 nm (Figure 2.10 (b)). Figure Adsorption isotherms for N719 dye on TiO 2 nanoparticles samples of P25, PA , and ST20. 67

73 We investigated N719 dye adsorption isotherms on these three kinds of TiO 2 nanocrystals because the dye adsorption behavior significantly affects DSSCs performance. The relationships between the concentration of the N719 dye and the dye adsorbed densities are show in Figure The adsorption isotherms data fitted with the Langmuir isotherm for these three kinds of nanoparticles, revealing Langmuir type monomolecular layer adsorption on the TiO 2 nanoparticles. The Langmuir equation can be represented as follows: C s / Q s = 1 / (Q. m K ad ) + C s / Q m (2.1) Where C s (mol/dm 3 ) is the equilibrium concentration of N719 dye in the solution, Q s (mol/m 2 ) is the dye adsorption density, K ad (dm 3 /mol) is the adsorption constant, and Q m (mol/m 2 ) is the maximum adsorption density. The adsorption constant and the maximum adsorption density were estimated by plotting C s /Q s against C s and the results were listed in Table 2.2. The adsorption constant K ad increases in an order of ST20 < P25 < PA that corresponds to the binding energy increasing order of the dye molecules adsorbed on the TiO 2 surfaces because K ad is a thermodynamic equilibrium constant. The maximum adsorption density Q m increases an order of ST20 < PA < P25. In this study, we used 0.3 mm N719 solution in the fabrications of DSSCs; therefore, the adsorption densities Q 0.3 at this concentration were estimated from the adsorption isotherms and listed also in Table 2.2. The dye adsorption density at 0.3 mm N719 concentration also increases in the order of ST20 < PA < P25. The dye adsorption results suggest that the TiO 2 nanoparticles with different facet on the surface exhibit different adsorption behavior. The different dye adsorption can be attributed to that different kinds of lattice spaces and surface energy have a different tendency to adsorb dye molecules. 44, 45 68

74 Table 2.2. N719 dye adsorption parameters and surface areas for TiO 2 nanoparticles. Sample K ad Q m Q 0.3 S BET ( 10 4 L/mol) ( 10-6 mol/m 2 ) ( 10-6 mol/m 2 ) (m 2 /g) P PA ST Effect of Crystal Facets on DSSC Performance To study the effect of the crystal facets on the DSSC performances, we used PA , P25, and ST20 TiO 2 nanoparticle samples to fabricate DSSC cells and investigated their cell performances by I-V characteristics and EIS. Figure 2.12 shows dependences of DSSC cell parameters on the TiO 2 film thickness of TiO 2 electrodes. The J sc values increase, reach the maximum values, and then decrease with increasing in the TiO 2 film thickness for all the samples. The PA , P25, and ST20 show the maximum values at around 8, 10, and 9 μm, respectively. The maximum J sc value increases in an order of ST20 (10.6 ma/cm 2 ) < P25 (13.3 ma/cm 2 ) < PA (14.2 ma/cm 2 ). The V oc values decrease linearly with increasing in the film thickness for all the samples, and the V oc values increase in an order of PA < ST20 < P25 when their thicknesses are same. The ff values of the samples exhibit nearly same trend that slightly decreases with increasing in the film thickness. The η values exhibit a similar behavior to the J sc values, which increase, reach the maximum values, and then decrease with increasing in the film thickness for all the samples. The maximum η values are 5.5% at around 6 μm for PA , 4.9% at around 10 μm for P25, and 4.1% at around 9 μm for ST20, respectively. These results reveal that the η values are 69

75 strongly dependent on the J sc values, and PA with the highest J sc value gives the highest η value. Figure Photovoltaic characteristics related to the thickness of TiO 2 single-layer electrodes. : PA ; : P25; : ST20. (a) short-current density, (b) open circuit voltage, (c) fill factor, and (d) power conversion efficiency. The above results suggest that the {010}-faceted nanoparticles (PA ) exhibit higher DSSC performance (η) than that of the [111]-faceted nanocrystals (P25), furthermore, the spherical nanocrystals (ST20) without specific facet on the surface exhibit lowest the DSSC performance, which suggests that the facet on the anatase nanocrystal surface affects the DSSC performance. The above results also reveal that the {010}-faceted PA exhibits higher J sc than that of [111]-faceted P25 and ST20 without specific facet on the surface. This fact may be explained by the larger dye adsorption constant K ad on {010} facet than which on [111]-facet and spherical 70

76 nanocrystal surface without specific facet. The maximum J sc increasing order of ST20 < P25 < PA corresponds to the increasing order of adsorption constant K ad, suggesting that the large K ad, namely, the strong anchoring of the dye molecule onto the TiO 2 surface can enhance J sc. The strong anchoring can accelerate the injection rate of excited photoelectrons from the dye molecule into the TiO 2 surface. 30 The dye adsorption amount can also affect the DSCCs performances. The increase of dye adsorption amount can enhance the light harvesting, which also enhances the J sc. 13 Figure Electrochemical impedance spectrum for DSSCs prepared using PA , P25 and ST20 samples. An EIS study was carried out on DSSCs fabricated using PA , ST20, and P25 with the same TiO 2 film thickness (see Figure 2.13). Their I-V characteristics are showed in Figure 2.14, the cell parameters and the EIS parameters are given in Table 2.3. The P25 and ST20 provide almost same R rc value, and PA provides a lower R rc than P25 and ST20. This result is consistent of their V oc values, namely the increase of R rc causes decrease of the charge recombination from TiO 2 conduction band to I - 3 in the electrolyte solution at the TiO 2 /electrolyte interface, which enhances V oc. 13 It has been reported that the increase of dye adsorption density can enhance V oc values because reduction of charge recombination at the TiO 2 /electrolyte interface

77 Figure Current-voltage characteristic curves of DSSCs fabricated using PA , P25 and ST20 without and with PA light-scattering layers (LSL). In the present study, the dye adsorption density Q 0.3 increasing order of ST20 < PA < P25 is little different from the V oc increasing order of PA < ST20 < P25. This result implies that except the dye adsorption density, other factors, such as particle morphology, surface energy, and surface electronic band structure maybe also affect the V oc. 28, 46 It has been reported that the surface electronic band structure of {010} facet is suitable for the photocatalytic reaction. 27 We think the surface electronic band structure also can affect the DSSCs performances. Figure Structure of DSSC with (a) single-layer TiO 2 electrode, and (b) double-layer TiO 2 electrode. 72

78 Effect of Light-Scattering Layer on DSSC Performance Although the PA sample shows a low η value than other samples because of its large particle size (Table 2.1). However, such large size TiO 2 particles can be used as light-scattering layer to improve light harvesting in the TiO 2 electrode (see Figure 2.15). 38 therefore, a light-scattering layer can be used to improve the DSSC performance. Figure 2.14 shows the I-V characteristics of the DSSCs of PA , ST20, and P25 electrodes without and with PA light-scattering layer, and their cell parameters and EIS parameters obtained from EIS spectra (see Figure 2.16), are summarized in Table 2.3. Compared with the DSSCs of PA , ST20, and P25 without the light-scattering layer, the DSSCs of PA /LSL, ST20/LSL, and P25/LSL with the light-scattering layer show enhanced J sc values. Although slightly decreases of V oc and ff are observed after coating the light-scattering layer, but the large J sc enhancement results the η improvement. The EIS results reveal that the introduction of the light-scattering layer causes a decrease in the R rc value because of increase in the surface area of TiO 2 /electrolyte interface, an increase in the R s value because of increase in the thickness of the TiO 2 film. Figure Electrochemical impedance spectra of DSSCs fabricated using PA , P25 and ST20 with PA light-scattering layers (LSL). 73

79 Table 2.3. Cell performance parameters and EIS parameters of DSSCs fabricated using PA , P25 and ST20 without and with PA light-scattering layers (LSL). Sample J sc V oc ff η R s R rc R Pt (ma/cm 2 ) (V) (%) (Ω) (Ω) (Ω) PA P ST PA /LSL P25/LSL ST20/LSL Conclusion The {010}-faceted anatase TiO 2 nanocrystals can be synthesized by hydrothermal treatment of PA-HTO nanosheets. The particles sizes and morphologies of the TiO 2 nanocrystals depend on the hydrothermal reaction temperature and ph value of reaction solutions. There are two kinds of reactions in the formation process of the anatase nanocrystals. One is the in situ topotactic structural transformation reaction, and another is the dissolution-deposition reaction. The DSSCs results for the {010}-faceted anatase nanoparticles reveal that the J sc and η values increase with decreasing crystal size, while the V oc and ff values increases with the increasing of the crystal size. Furthermore, the DSSCs results for anatase nanocrystals with different facets on surface suggest that the J sc and η values increase in the order of spherical nanocrystals without specific facet < [111]-faceted nanocrystals < {010}-faceted nanocrystals. Namely the {010}-faceted nanocrystals are promising for the high 74

80 performance DSSCs. The leaflike anatase particles with large particle size are effective as the light-scattering layer in the porous TiO 2 electrode to enhance light harvesting in the TiO 2 electrode, which improves DSSCs performances References [1]. O Regan, B.; Grätzel, M. Nature 1991, 353, [2]. Grätzel, M. Nature 2001, 414, [3]. Grätzel, M. J. Photochem. Photobiol. C 2003, 4, [4]. Zhang, S.; Yang, X.; Qin, C.; Numata, Y.; Han, L. J. Mater. Chem. A 2014, 2, [5]. Zhang, Q.; Chou, T. O.; Russo, B.; Jenekhe, S. A.; Cao, G. Angew. Chem. Int. Ed. 2008, 47, [6]. Tan, B.; Wu, Y. J.Phys. Chem. B 2006, 110, [7]. Yang, W.; Wang, Y.; Shi, W. CrystEngComm 2012, 14, [8]. Zhang, G., Pan, K.; Zhou, W.; Qu, Y.; Pan, Q. Dalton Trans. 2012, 41, [9]. Yan, X.; Feng, L.; Jia, J.; Zhou, X.; Lin, Y. J. Mater. Chem. A 2013, 1, [10]. Wang, H; Miyauchi, M.; Ishikawa, Y.; Pyatenko, A.; Koshizaki, N.; Li, Y.; Li, L.; Li, X.; Bando, Y.; Golberg, D. J. Am. Chem. Soc. 2011, 133, [11]. Grätzel, M. Prog. Photovolt. Res. Appl. 2000, 8, [12]. Chen, D.; Huang, F.; Cheng, Y.; Caruso, R. A. Adv. Mater. 2009, 21, [13]. Usami, A. Sol. Energy Mat. Sol. Cells 2000, 64, [14]. Kang, S. H.; Choi, S. H.; Kang, M. S.; Kim, J. Y.; Kim, H. S.; Hyeon, T.; Sung, Y. E. Adv. Mater. 2008, 20, [15]. Wen, P.; Itoh, H.; Feng, Q. Chem. Lett. 2006, 35,

81 [16]. Wen, P.; Itoh, H.; Tang, W., Feng, Q. Langmuir 2007, 23, [17]. Dinh, C. T.; Nguyen, T. D.; Kleitz, F.; Do, T. O. ACS Nano 2009, 3, [18]. Li, J.; Xu, D. Chem. Commun. 2010, 46, [19]. Pan, J.; Liu, G.; Lu, G. Cheng, H. Angew. Chem. Int. Ed. 2011, 50, [20]. Barnard, A. S.; Cyrtiss, L. A. Nano let. 2005, 5, [21]. Wang, D.; Kanhere, P.; Li, M.; Tay, T. Y.; Huang, Y.; Sum, T.; Mathews, N.; Sritharan, T.; Chen, Z. J. Phys. Chem. C 2013, 117, [22]. Amano, F.; Yasumoto, T.; Prieto-Mahaney, O. O.; Uchida, S.; Shibayama, T.; Ohtani, B. Chem. Commun. 2009, 17, [23]. Yang, X.; Li, Z.; Sun, C.; Yang, H.; Li, C. Chem. Mater. 2011, 23, [24]. Wu, B.; Guo, C.; Zheng, N.; Xie, Z.; Stucky, G. D. J. Am. Chem. Soc. 2008, 130, [25]. Han, X.; Kuang, Q.; Jin, M.; Xie, Z. Zheng, L. J. Am. Chem. Soc. 2009, 131, [26]. Lazzeri, M.; Vittadini, A.; Selloni, A. Phys. Rev. B. 2001, 63, [27]. Xu, H.; Reunchan, P.; Ouyang, S.; Tong, H.; Umezawa, N.; Kako, T.; Ye, J. Chem. Mater. 2013, 25, [28]. Pan, J.; Wu, X.; Wang, L.; Liu, G.; Lu, G.; Cheng, H. Chem. Commun. 2011, 47, [29]. Wen, P.; Tao, Z.; Ishikawa, Y.; Itoh, H.; Feng, Q. Appl. Phys. Lett. 2010, 97, [30]. Wen, P.; Xue, M.; Ishikawa, Y.; Itoh, H.; Feng, Q. ACS Appl. Mater. Interfaces 2012, 4, [31]. Wu, D.; Gao, Z.; Xu, F.; Chang, J.; Gao, S.; Jiang, K. CrystEngComm 2013, 15, [32]. Zhang, W.; Xie, Y.; Xiong, D.; Zeng, X.; Li, Z.; Wang, M.; Cheng, Y, B.; Chen, 76

82 W.; Yan, K.; Yang, S. ACS Appl. Mater. Interfaces 2014, 6, [33] W. Wang, H. Zhang, R. Wang, M. Feng and Y. Chen, Nanoscale, 2014, 6, [34]. Cölfen, H.; Antonietti, M. Angew. Chem. Int. Ed. 2005, 44, [35]. Wen, P.; Itoh, H.; Tang, W.; Feng, Q. Microporous Mesoporous Mat. 2008, 116, [36]. Wen, P.; Ishikawa, Y.; Itoh, H.; Feng, Q. J. Phys. Chem. C. 2009, 113, [37]. Yuan, H.; Besselink, R.; Liao, Z.; Elshof, J. E. T. Scientific Reports 2014, 4, [38]. Li, G.; Gray, K. A. Chem. Mater. 2007, 19, [39]. Feng, X.; Zhai, J.; Jiang, L. Angew. Chem., Int. Ed. 2005, 44, [40]. Zheng, Y.; Shi, E.; Chen, Z.; Li, W.; Hu, X. J. Mater. Chem. 2001, 11, [41]. Han, L.; Koide, N.; Chiba, Y.; Islam, A.; Mitate, T. C. R. Chimie. 2006, 9, [42]. Wang, Q.; Ito, S.; Grätzel, M. J. Phys. Chem. B. 2006, 110, [43]. Adachi, M.; Sakamoto, M.; Jiu, J.; Ogata, Y.; Isoda, S. J. Phys. Chem. B. 2006, 110, [44]. Langmuir, I. J. Am. Chem. Soc. 1918, 40, [45]. Yu, J.; Fan, J.; Lv, K. Nanoscale 2010, 2, [46]. Meng, Q. B.; Takahashi, K.; Zhang, X. T.; Sutanto, I.; Rao, T. N.; Sato, O.; Fujishima, A.; Watanabe, H.; Nakamori, T.; Uragami, M. Langmuir 2003, 19,

83 CHAPTER II (Synthesis of titanium dioxide nanocrystals from H 1.07 Ti 1.73 O 4 layered titanate nanosheets by normal hydrothermal process and their dye-sensitized solar cell performance) was reprinted (adapted) with permission from (Changdong Chen, Galhenage Asha Sewvandi, Takafumi Kusunose, Yasuhiro Tanaka, Shunsuke Nakanishi, and Qi Feng, Synthesis of {010}-Faceted Anatase TiO 2 Nanoparticles from Layered Titanate for Dye-sensitized Solar Cells, CrystEngComm, 2014, 16 (37), ) Royal Society of Chemistry. 78

84 CHAPTER III Synthesis of Titanium Dioxide Nanocrystals from H 1.07 Ti 1.73 O 4 Layered Titanate Nanosheets by Microwave Hydrothermal Process for High Performance Photocatalyst and Dye-Sensitized Solar Cells 3.1. Introduction Titanium dioxide (TiO 2 ) is one of the most important functional materials and has attracted great attentions in the last decades due to its excellent photocatalytic 1-3 and dye-sensitized solar cells (DSSCs) performances, 4-6 low cost, and nontoxicity. To improve photocatalytic performance of TiO 2 materials, a large number of studies have been implemented on the syntheses of the nanocrystals and nanostructured materials, extend light absorption region to visible light by nitrogen-doping, 7 and synthesis of TiO 2 heterojunction materials to improve charge separation ability. 8 TiO 2 as a primary material impacts on the dye adsorption and photoelectron transport which strongly affect the performance of DSSCs. 9 In fact, the crystal structure, crystallinity, crystal size, morphology, and crystal surface structure, such as crystal facet exposing on surface of TiO 2 materials, can affect photocatalytic and photovoltaic performances. 10 Recently the effect of crystal facet on the photocatalytic and photovoltaic 11, 12 performances attracted much attention. We have reported the first study on photocatalytic performance of anatase TiO 2 nanocrystals with specific crystal facet on the surface, and found that {010}-faceted anatase TiO 2 nanocrystals exhibit higher photocatalytic activity than the normal spherical nanocrystals without specific facet 79

85 on the surface. 13, 14 The results animated researchers to study the effect of crystal facet on TiO 2 photocatalytic properties. Yang et al. have synthesized high percentage of {001}-faceted TiO 2 particles using hydrofluoric acid solution, their results of the experimental study and first-principle quantum chemical calculation reveal that the {001}-facet of anatase is one of reactive facets for the photocatalytic reactions. 15 After that, Wu et al. have synthesized {010}-faceted and {001}-faceted anatase TiO 2 nanocrystals, and found that the {010}-faceted nanocrystals exhibit higher photocatalytic activity than the {001}-faceted nanocrystals and commercial P25 sample in degradation of methyl orange. 16 The P25 sample is considered as a standard TiO 2 nanocrystal sample with high photocatalytic activity. Han et al. have reported that the {001}-faceted anatase TiO 2 nanosheets exhibit the higher photocatalytic activity than the P25 sample. 17 Amano et al. have reported {101}-faceted perfect octahedral anatase crystallites exhibit higher photocatalytic activity than the P25 sample. 18 Therefore, these results indicate that in addition of bandgap and bulk band structure, the photocatalytic activity also depends on the crystal facet that relates both surface atomic structure and surface electronic band structure The crystal facet also affects the DSSCs performance. We have reported the first study on the crystal facet effect on the DSSCs performance, and found that {010}-faceted anatase nanocrystals exhibit specific high short-circuit current density 23, 24 (J sc ) due to strong adsorption of sensitizer dye molecules on the {010} facet, which improves injecting photoelectrons from LUMO (lowest unoccupied molecular orbital) of the molecule to the conduction band of TiO Wu et al. have reported that the bipyramid rod-like anatase nanocrystals with high percentage of {101} facets can capture the electrons injected from the dye photoexcited state to the anatase conducting band, decrease the annihilation of electrons, and increase the electron concentration of the TiO 2 photoelectrode compared to the spherical anatase 80

86 nanocrystals. 26 The crystal facet effect on the DSSCs performance has been confirmed 20, also by other studies. Generally a hydrothermal process is used for the synthesis of the TiO 2 nanocrystals with a specific facet on the surface. The {001}-, {010}-, and {101}-faceted anatase nanocrystals have been synthesized by hydrothermal treatments of the titanium precursors in solutions containing the hydrofluoric acid as a crystallographic 15, 17 controlling agent, We have reported a unique and useful hydrothermal soft chemical process for the synthesis of {010}-faceted anatase nanocrystals by the 13, 14, 30, 31 topochemical conversion of layered titanate nanosheet precursor to anatase. However, the normal hydrothermal processes usually cost much long time, owing to tardy heat transmission in the normal hydrothermal processes. Microwave hydrothermal process is a unique and useful method, in which the solvent molecules and reactants can be heated directly by increasing their rotational speeds when solvent molecules and reactants absorb the microwave. 32 Some studies on microwave hydrothermal synthesis of TiO 2 nanocrystals have been reported. This unique heating process can shorten reaction time and increase the crystallization of products In the most cases, the TiO 2 nanocrystals with controllable crystal size and morphology were synthesized by microwave hydrothermal treatment of the colloidal TiO 2 32, solutions prepared by hydrolysis of the titanium compound. In this chapter, we describe synthesis TiO 2 nanocrystals from exfoliated layered titanate nanosheet precursor with lepidocrocite-like structure using microwave hydrothermal process for the first time. The microwave hydrothermal process can provide {010}-faceted anatase TiO 2 nanocrystals with uniform controllable sizes and morphologies. This gives us an opportunity to study the effects of crystal facet, size, and morphology on the bandgap, surface electronic band structure, photocatalytic activity, and DSSCs performance of TiO 2 nanocrystals. The results revealed that the 81

87 {010}-faceted anatase nanocrystals exhibit higher photocatalytic activity due to its larger bandgap, and excellent DSSCs performance than P25 nanocrystals which contain [111]-faceted (facet vertical to [111]-direction) anatase nanocrystals and spherical anatase nanocrystals without specific facet on the surface Experimental Section Preparation of HTO Layered Titanate Nanosheet Colloidal Solutions The layered titanate nanosheet colloidal solutions were prepared from layered titanate K 0.8 Ti 1.73 Li 0.27 O 4 (KTLO). A mixture of 5.1 g of KOH, 0.6 g of LiOH H 2 O, 6.9 g of TiO 2 (anatase form) and 25 ml of distilled water were sealed in Hastelloy-C-lined vessel with internal volume of 45 ml, and then heated at 250 o C for 24 hrs under stirring conditions. The hydrothermally treated samples were washed with distilled water up to neutral and dried at room temperature to obtain the layered titanate KTLO as the precursor. The KTLO sample (10 g) was acid-treated with a 0.2 M HNO 3 solution (1L) for 24 hrs under stirring conditions to exchange K + ions and Li + ions with H + ions to obtain an H + -form layered titanate H 1.07 Ti 1.73 O 4 (HTO). The acid treatment was repeated twice to complete the ion-exchange reaction. The HTO layered titanate was washed with distilled water several times and dried using a freeze drier. The layered titanate nanosheet colloidal solution was prepared by treating HTO sample (10 g) in 0.1 M n-propylamine (PA) (1L) solution under stirring conditions at room temperature for 24 hrs, and the nanosheet colloidal solution was named as PA-HTO Microwave Hydrothermal Treatment of PA-HTO Solution TiO 2 nanocrystals were prepared by microwave hydrothermal treatment of the 82

88 PA-HTO nanosheet colloidal solution. Before the microwave hydrothermal treatment, the PA-HTO nanosheet colloidal solution was adjust to a desired ph value with a 3 M HCl solution in a ph range below 11.7 or a 1M KOH solution in a ph range above The 40 ml of ph adjusted PA-HTO solution was sealed into a Teflon vessel with internal volume of 80 ml, and then it was microwaved at a desired temperature for 2 hrs. After the microwave hydrothermal treatment, the obtained samples were washed with distilled water several times, and finally dried using a freeze drier. The obtained TiO 2 sample was named as MW-X-Y, where X and Y were the temperature of microwave treatment and ph value of the nanosheet solution used for the microwave hydrothermal reaction, respectively Photocatalytic Characterization The 20 mg of TiO 2 sample was added into a 10 ppm methylene blue aqueous solution (MB, 100mL) and stirred for 2 hrs to disperse the sample well in the solution and to reach MB adsorption equilibrium on TiO 2 nanocrystals surface under dark conditions without UV-irradiation. And then the suspension was illuminated by a 100 W ultraviolet lamp with nm wavelength (UVA, Asahi Spectra, LAX-Cute) at room temperature, and the lamp located at 20 cm away from the MB solution under continuously stirring conditions. At every 20 min interval, 3 ml solution was drawn from the suspension and immediately centrifuged to separate the TiO 2 nanocrystals from the solution under rpm for 10 min. The MB concentration in the solution was determined by using a Shimadzu UV-2450 spectrophotometer. Degradation efficiency of MB by the photocatalytic reaction was calculated from the variations of the MB concentration in the solution before and after UV light irradiation. 83

89 Fabrication and Characterization of Dye-Sensitized Solar Cells A TiO 2 nanocrystal paste was prepared by mixing TiO 2 nanocrystal sample (0.5 g), ethanol (2.5 g), α-terpineol (2.0 g), 10 wt% solution of ethyl-cellulose 10 (8-14mPas, 1.4 g), and 10 wt% solution of ethyl-cellulose 45 (45-65mPas, 1.1 g). The mixture was dispersed by ultrasonication for 30 min, and then ball-milling for 72 hrs. After ball-milling, the ethanol was removed from the mixture using a rotary-evaporator. P25 and ST20 nanocrystals pastes were also prepared using the similar method. The P25 nanocrystals sample was purchased from Degussa. The ST20 nanocrystals sample was synthesized by hydrothermal treatment of a commercial anatase nanoparticle sample (ST01, Ishihara Sangyo) with crystal size of 7 nm at 200 o C for 12 hrs. The TiO 2 photoelectrode was prepared as follows. The fluorine-doped tin oxide (FTO) conducting glass plate was cleaned in distilled water and acetone by ultrasonication for 10 min, orderly. Then the FTO glass plate was dipped in 0.1M titanium tetraisopropoxide (TTIP) solutions for 1 min and washed with distilled water and ethanol, dried at room temperature and calcined at 480 o C for 1 h to coat the FTO glass surface with a dense TiO 2 thin film. The prepared TiO 2 paste was coated (10 10 mm) on the TTIP-treated FTO glass plates by screen printing technique and kept in an ethanol box until the TiO 2 film surface smoothly and then dried at 120 o C for 10 min. This process was repeated to obtain a desired thickness of TiO 2 film. After the TiO 2 paste coating, the TiO 2 film was calcined at 450 o C for 30 min to obtain a TiO 2 porous electrode. The TiO 2 porous electrodes were dipped in the 0.1M TTIP solution for 1 min, and washed with distilled water and ethanol, dried at room temperature and calcined at 480 o C for 1 h again. After cooling to 80 o C, the TiO 2 porous electrodes were soaked in a M N719 dye (di-tetrabutylammonium cisbis(isothiocyanato) bis(2,2'-bipyridyl-4,4'-dicarboxylato)ruthenium(ii)) solution for 24 hrs at room temperature, and then washed with a t-butyl alcohol and acetonitrile (v/v=50%:50%) 84

90 mixed solvent. The DSSCs were comprised of a dye-adsorbed TiO 2 electrode as an anode, a Pt-coated FTO glass as a cathode, and an electrolyte solution between the anode and the cathode. The electrolyte solution contains 0.1M LiI, 0.01M I 2, 0.6M of 1-butyl-3-n-propylimidazolium iodide (BMII), 0.4M 4-tert-butylpyridine (TBP) and 0.1M guanidine thiocyanate (GT) in acetonitrile and valeronitrile (v/v=85%:15%). The photocurrent voltage characteristic curves for the DSSCs were measured using a Hokuto-Denko BAS100B electrochemical analyzer under irradiation with simulated sunlight of AM 1.5 (100mW/cm 2 ), using a sunlight simulator (YSS-E40, Yamashita Denso). A light-passing mask was fixed on the surface of FTO glass of the anode to set the effectively irradiating area on the cell as 0.25 cm Physical Analysis The crystal structure of the samples was investigated using a powder X-ray diffractometer (Shimadzu, Model XRD-6100) with Cu Kα (λ= nm) radiation. The size and morphology of the samples was characterized by field emission scanning electron microscopy (FE-SEM) (Hitachi, Model S-900). Transmission electron microscopy (TEM) observation and selected-area electron diffraction (SAED) was performed on a JEOL Model JEM-3010 system at 300 kv. Nitrogen gas adsorption was carried out on a QUANTACHROME AUTOSORB-1-MP apparatus. The specific surface area of samples was calculated from the adsorption data using the Brunauer-Emmett-Teller (BET) method Results and Discussion Microwave-Assisted Conversion of HTO Nanosheets to TiO 2 Nanocrystals 85

91 Figure 3.1. XRD patterns of PA-HTO nanosheet and products obtained by microwave hydrothermal treatment of PA-HTO nanosheet solution with ph 11.5 at different temperatures for 2 hrs. : PA-HTO phase, and : anatase phase. The PA-HTO nanosheet colloidal solutions with different ph values were hydrothermally microwaved at various temperatures to synthesize TiO 2 nanocrystals. The XRD patterns of samples prepared by hydrothermally microwaving PA-HTO nanosheet solution with ph 11.5 at dissimilar temperatures are show in Figure 3.1. Before the microwave hydrothermal treatment, PA-HTO has a lepidocrocite-like layered structure with a basal spacing of 1.09 nm (see Figure 3.2), indicating CH 3 (CH 2 ) 2 NH + 3 (PA-H + 13, 14 ) ions are intercalated into the interlayer space of HTO. After the microwave hydrothermal treatment at 95 o C, a mixture of anatase phase and PA-HTO phase was obtained. With the temperature increasing, the proportion of anatase phase increases and that of the PA-HTO phase decreases. At 165 o C, PA-HTO nanosheets were transformed to anatase phase completely and the single-phase anatase TiO 2 was obtained above this temperature. Meanwhile, the crystallinity of formed anatase phase increases gradually with the temperature increasing. 86

92 Figure 3.2. XRD patterns of (a) KTLO, (b) HTO, and (c) PA-HTO samples. Figure 3.3 provides XRD patterns of products obtained by microwave hydrothermal treatment of PA-HTO nanosheet solutions with different ph values at 175 o C for 2 hrs. A mixture of rutile, brookite, and anatase phases was obtained at ph 0.5. In a ph range of 1.5 to 11.5, single-phase anatase TiO 2 was formed. The HTO layered phase is stable in a ph range of above 13, where PA-H + ions in the interlayer space and surface of HTO are exchanged with K + ions and then a K + -form HTO phase with a basal spacing of 0.88 nm is formed, but it retains the lepidocrocite-like layered structure. 14 On the basis of the XRD results (see Figure 3.4), the dependence of the products on the reaction temperature and ph value are summarized in Figure 3.5. The rutile and brookite phases are formed preferentially below ph 1, and anatase phase is formed preferentially in the ph range of 1 to 12. The lepidocrocite-like layered structure is stable in the ph range of above 13, where K + -form HTO is formed. The result is different from normal hydrothermal treatment of HTO nanosheets, in which single-phase rutile is formed in the ph range below 1, and without brookite phase is formed in all ph range. 39 The formation of metastable brookite phase reveals the microwave hydrothermal process is suitable for synthesis of such unstable phase owing to its unique heating mechanism. 35, 36 In comparison with normal hydrothermal treatment of PA-HTO solutions, the microwave hydrothermal process can shorten the 87

93 significantly reaction time and complete fast crystallization in the short period of 35, 36 time. Figure 3.3. XRD patterns of products obtained by microwave hydrothermal treatment of PA-HTO nanosheet solutions with different ph values at 175 o C for 2 hrs. : K + -form HTO phase, : anatase phase, : brookite phase, and : rutile phase. 88

94 Figure 3.4. XRD patterns of PA-HTO nanosheet and products obtained by microwave hydrothermal treatment of PA-HTO nanosheet solution with dissimilar ph value at 95 o C-185 o C. (a) 0.5, (b) 1.5, (c) 3.5, (d) 5.5, (e) 7.5, (f) 9.5, (g) 11.5, and (h) : K-HTO, : anatase phase, : rutile phase, : brookite phase. Figure 3.6 shows FE-SEM images of the products obtained under different microwave hydrothermal conditions. MW is a mixture of anatase and layered phases (see Figure 3.7), and has a platelike particle morphology with the thickness about 50 nm (Figure 3.6(a)) that is much thicker than PA-HTO nanosheet of about 1 nm (see Figure 3.7(a)). This result suggests that a restacking reaction of HTO nanosheets into the platelike particles occurs under the low ph conditions. Namely, PA-H + ions adsorbed on HTO nanosheet surface are ion-exchanged with H + ions in the acidic solutions, which results the restacking the exfoliated HTO nanosheets. The platelike particle surface is covered with many small nanoparticles. The small nanoparticles correspond to the anatase phase. Platelike anatase particles with a thickness of about 30 nm were obtained at 165 o C-pH 1.5 (Figure 3.6(b)). The platelike particles are polycrystal particle constructed from small nanocrystals with a size of about 20 nm. The thickness of the platelike anatase particles is thicker than that of the HTO nanosheet, suggesting the restacking reaction of the HTO nanosheet during the conversion reaction from the HTO structure to the anatase structure under the low ph conditions. A mixture of platelike particles and tetragonal nanocrystals with a size of about 20 nm was obtained at 165 o C-pH 3.5 (Figure 3.6(c)). The XRD result indicates that this sample is single-phase anatase (see Figure 3.4(f)), therefore, the platelike and tetragonal nanocrystals are anatase phase. The formation of the platelike anatase particles can be explained by an in situ topochemical conversion reaction of HTO platelike particles to anatase platelike particles. The topochemical conversion reaction can retain particle morphology of the precursor after the structural 89

95 transformation. 14, 31 The formation of the small tetragonal anatase can be explained by splitting the platelike anatase particle to the small particles by dissolution-deposition reaction on the platelike particle surface. Figure 3.5. Phase diagram under microwave hydrothermal conditions, (A) mixture of rutile, brookite and anatase phases; (B) anatase phase; (C) mixture of layered and anatase phases; (D) layered phase. Figure 3.6. FE-SEM images of (a) MW , (b) MW , (c) MW , (d) MW , (e) MW , (f) MW , (g) MW , (h) MW , and (i) MW samples. 90

96 Three kinds of typical particle morphologies were observed in MW sample containing rutile, brookite, and anatase phases (Figure 3.6(d)). A TEM study reveals that the rodlike particles with a size of 300 nm in length and 50 nm in width correspond to the rutile phase, the blocking particles with a size of about 100 nm correspond to the brookite phase, and the nanocrystals with a size of about 20 nm correspond to anatase phase (see Figure 3.8). The rutile, brookite, and anatase phases are formed mainly by the dissolution-deposition reaction mechanism similar to the normal hydrothermal reaction under the acidic reaction conditions because their particle morphologies have not relationship with HTO nanosheet precursor At 175 o C-pH 1.5, the platelike anatase particle was almost split into the nanocrystals with a size about 20 nm (Figure 3.6(e)), owing to increase of the dissolution-deposition reaction with increasing the reaction temperature. At 175 o C-pH 3.5 (Figure 3.6(f)), tetragonal anatase nanocrystals with a size about 20 nm were observed, where without platelike anatase particles were observed. The average crystal size is similar to the small anatase crystals formed at 165 o C-pH 3.5, but the morphology is more uniform. The regular tetragonal anatase nanocrystals with size of about 80 nm were formed at 175 o C-pH 9.5 (Figure 3.6(g)). And the regular quadratic prism anatase nanocrystals with a size of about 150 nm in length and 30 nm in width were formed at 175 o C-pH 11.5 (Figure 3.6(h)). These results reveal that the morphology and size controllable anatase nanocrystals can be prepared by changing the reaction temperature and the ph value of solution, and the crystal size increases with increasing of the ph value in a ph range of 3.5 to MW sample with layered structure has platelike particle morphology (Figure 3.6(i)). The result reveals HTO nanosheets are restacked into a platelike particles of K + -HTO in the KOH solution and the layered structure is stable in the ph > 13 range under the microwave hydrothermal conditions. 14 Furthermore, compared to with normal 91

97 hydrothermal treatment of PA-HTO solutions (see Figure 3.9), microwave hydrothermal process can provide anatase nanocrystals with more uniform morphology and size, due to its uniform heating mechanisms. 39 Figure 3.7. (a) FE-SEM images, (b) TEM and (c) HR-TEM images of PA-HTO nanosheets. PA-HTO has a lepidocrocite-like layered structure (JCPDS File No , orthorhombic system). Lattice fringes with a d value of 0.37 nm corresponding to (100) plane of PA-HTO. Fourier transform (FTF) diffraction pattern shows diffraction spots of (200), and (001) planes, indicating that the [010]-direction is perpendicular to the basal plane of PA-HTO nanosheet. Figure 3.8. TEM and HR-TEM images of MW samples, the rodlike particles (a, b) correspond to rutile phase, which exhibit lattice fringes with a d-value of nm corresponding to (110) plane of rutile phase, the blocking particles (c, d) with a size of about 100 nm correspond to brookite phase, which exhibit lattice fringes with a d-value of nm corresponding to (120) plane of brookite phase. 92

98 Figure 3.9. FE-SEM images of anatase nanoparticles obtained using normal hydrothermal treatment of PA-HTO nanosheet solutions at (a) 135 o C-pH3.5, (b) 135 o C-pH9.5, and (c) 135 o C-pH Nanostructural Study on Conversion Reaction from HTO Nanosheets to TiO 2 Nanocrystals To understand the conversion reaction from HTO nanosheets to TiO 2 nanocrystals in detail, the synthesized nanocrystals were investigated using TEM and SAED (Figure 3.10). In MW , platelike particles constructed from nanocrystals with a size of about 20 nm are observed (Figure 3.10(a)). The platelike particle shows a Fourier transform (FTF) diffraction pattern of single crystal-like anatase phase with (101), (-101), and (004) planes. It indicates that all anatase nanocrystals in one platelike particle show same crystal orientation, and the [010]-direction is perpendicular to basal plane of the platelike particle, namely (010)-faceted anatase mesocrystal was obtained. The transformation reaction from HTO nanosheets to anatase is a topochemical reaction, in which there is a specific crystallographic topological correspondence between the HTO structure and the anatase structure, where the [010]-direction of HTO corresponds to the [010]-direction of anatase, which are perpendicular to basal planes of the anatase platelike particle and HTO nanosheet (see Figure 3.7)

99 Figure HR-TEM images, SAED and Fourier transform (FTF) diffraction patterns of (a) MW , (b) MW , (c, d) MW , and (e, f) MW samples. In MW , the tetragonal nanocrystals were observed mainly. The typical tetragonal nanocrystal exhibit lattice fringes with a d-value of 0.35 nm corresponding to (101) plane of anatase phase in its HR-TEM image, and diffraction spots of (101), (200), and (004) planes in its SAED pattern (Figure 3.10(b)). This result reveals that the [010]-direction is perpendicular to the basal plane of tetragonal nanocrystal, namely it is a (010)-faceted anatase nanocrystal exposing (010) facet on the basal plane. And two side faces correspond to (101) facet and other two side faces correspond to the facet vertical to (101) facets. The (101) lattice fringes and (002) lattice fringes are observed in the HR-TEM images of tetragonal nanocrystal of MW and quadratic prism nanocrystal of MW (Figure 3.10(c)-(f)), namely all these anatase nanocrystals prepared from the HTO nanosheets expose the (010) facet on their basal planes. In the tetragonal nanocrystal of MW and the quadratic prism nanocrystal of 94

100 MW , two side faces correspond to (100) facet and other two side faces to (001) facets, which are different to the tetragonal nanocrystal of MW Furthermore, axis-direction of the quadratic prism nanocrystal corresponds to [001]-direction. The above results are different to the results of normal hydrothermal reaction. Although all anatase nanocrystals prepared by treatment of HTO nanosheets under the normal hydrothermal conditions expose the (010) facet on their basal planes, similar to the microwave hydrothermal treatment, the main product is rhombic anatase nanocrystals at ph 3.5, and only one kind of the tetragonal nanocrystal with two (101)-faceted side faces and other two side faces of the facet vertical to (101) facet is formed under the normal hydrothermal conditions. 39 A reaction mechanism for the conversion reaction from the HTO nanosheets to TiO 2 nanocrystals can be summarized as in Figure 3.11 on the basis of the above results. In ph range of 1 to 3, firstly some PA-HTO nanosheets restack into platelike particle with a thickness of about 30 nm by ion-exchange of PA-H + ions adsorbed on HTO nanosheet surface with H + ions in the solution. And then it is transformed into {010}-faceted platelike anatase particle via in situ topochemical conversion mechanism. The platelike anatase particle is constructed from well-aligned anatase nanocrystals in {010}-orientation, namely it is a {010}-faceted anatase mesocrystal. Furthermore, the anatase platelike mesocrystal can be split into its unit nanocrystals at high temperature by the dissolution-decomposition reaction. 95

101 Figure Transformation reaction mechanism from PA-HTO nanosheets to anatase TiO 2 nanocrystals. In the ph range of 3 < ph < 13, firstly the PA-HTO nanosheet is transformed into {010}-faceted anatase nanosheet via in situ topochemical conversion reaction, and then the {010}-faceted anatase nanosheet can be split further into small anatase nanocrystals via the dissolution-decomposition reaction. Therefore, {010}-faceted anatase nanocrystals with tetragonal and quadratic prism morphologies can be formed. When the anatase nanosheets are split along (101) facet and facet vertical to (101) facet, the tetragonal nanocrystals with two (101)-faceted side faces and other two side faces of the facet vertical to (101) facets are formed (Figure 3.10(b)) at around ph 3.5. When the anatase nanosheets are split along (100) facet and (001) facet, the anatase nanosheet is split to the tetragonal nanocrystals with two (100)-faceted side faces and other two (001)-faceted side faces (Figure 3.10(d)) at around ph 9.5, and to the 96

102 quadratic prism nanocrystals with two (100)-faceted side faces and other two (001)-faceted side faces (Figure 3.10(f)) at around ph Furthermore, when the ph value is lower than 1, HTO can be transformed into rutile and brookite phases by the dissolution-decomposition reaction. Therefore, there are not correlations between the morphology of HTO nanosheet precursor and which of rutile and brookite nanocrystals similar to the normal hydrothermal reaction. When the ph value is higher than 13, the HTO nanosheet is stable and maintains primary morphology and structure under the microwave hydrothermal conditions. The above results reveal that two main types of reactions can occur simultaneously in the formation reaction processes of the anatase nanocrystals from PA-HTO nanosheets. One is the in situ topotactic transformation reaction, in which the structure of PA-HTO nanosheets is transformed to anatase structure but the morphology of the precursor is retained after the reaction. Another is the dissolution-deposition reaction on the surface of the PA-HTO nanosheets, which splits the anatase nanosheet and platelike mesocrystal into small nanocrystals. And the nanocrystal morphology and size can be controlled by the dissolution-deposition reaction. The microwave hydrothermal process is suitable for controlling dissolution-deposition reaction, owing to its uniform heating mechanism Electronic Band Structure and Photocatalytic Response of TiO 2 Nanocrystals Since the facet on TiO 2 crystal surface affects the photocatalytic activity, we investigated the photocatalytic activities of TiO 2 nanocrystals prepared from the PA-HTO nanosheets, and also other two samples of P25 and ST20 for the comparison. P25 is a benchmark TiO 2 nanocrystal sample commonly used in the photocatalytic 97

103 and DSSCs studies, owing to its high performances. P25 contains 80% anatase phase and 20% rutile phase. P25 contains about 80% of nanocrystals with a size of about 20 nm and about 20% of nanocrystals with a size about 80 nm (see Figure 3.12(a)). The nanocrystals with a size of about 20 nm are anatase phase, while which with a size of about 80 nm are rutile phase. There are two kinds of particle morphologies for the anatase nanocrystals. One is near-spherical morphology (content of about 40%) and other is tetragonal morphology (content of about 40%). The near-spherical nanocrystals have not specific facet on the crystal surface, while tetragonal nanocrystals have specific facet on the crystal surface. In the HR-TEM image of the tetragonal nanocrystal, the anatase lattice fringes of (101) and (011) facets (d=0.352 nm) are observed. The HR-TEM image reveals that the basal plane of the tetragonal particle correspond to a facet vertical to [111]-direction (we call it as [111]-facet), and other four planes correspond to the {101} facet as shown in Figure 3.12(d). The [111]-facet is different from {111} facet in the tetragonal system of anatase phase. The ST20 sample is anatase phase with spherical particle morphology without specific facet on the surface and has a crystal size of about 20 nm (see Figure 3.12(b)). Figure FE-SEM images of (a) P25 and (b) ST20 samples, (c) HR-TEM image and (d) model of crystal facets on surface of tetragonal anatase nanocrystal of P25 sample. 98

104 Table 3.1. Crystal phase, exposed facet, surface area, size, absorption edge, and bandgap for TiO 2 samples. Sample Crystal Exposed S BET Crystal Absorption E g Phase Facet (m 2 /g) Size Edge (ev) (nm) (nm) P25 rutile /anatase Vertical to [111] ST20 anatase MW anatase (010) MW anatase (010) MW anatase (010) MW anatase (010) Figure 3.13 shows the time-dependent photodegradation profiles of methylene blue (MB) dye over TiO 2 nanocrystals samples under ultraviolet irradiation. It can be seen that the photocatalytic activity increases in an order of ST20 < MW < MW < P25 < MW < MW It is well-known that the photocatalytic activity is strongly dependent on surface area of TiO 2, and enhances with increasing the surface area. The measured specific surface area results are shown in Table 3.1. Although P25, ST20, and MW have similar specific surface area values and similar crystal sizes of about 20 nm, MW exhibits highest photocatalytic activity in these three kinds of samples. This result implies that the {010} facet exhibits higher photocatalytic activity than other facets. Although MW has smaller specific surface area and larger crystal size than MW , but it exhibits higher photocatalytic activity than MW This result suggests that MW has higher surface activity than that of MW

105 The higher surface activity maybe results from the larger faction of {010} facet of MW that has tetragonal morphology with four {010} planes and two {001} planes on its surface (Figure 3.11). MW exhibits lower photocatalytic activity than that of MW , even it has similar facets on the particle surface. It is owing to the smaller specific surface area and larger crystal size for MW MW exhibits photocatalytic activity less than MW and MW , but higher than MW Although the platelike mesocrystal of MW has the smallest specific surface area, it exposes dominantly {010} facet on the surface. This result also implies that the mesocrystal structure maybe enhances photocatalytic activity. P25 exhibits higher activity than ST20, suggesting the activity of [111]-facet surface is higher than that of without specific facet. Therefore, the above results reveal that the surface photocatalytic activity is dependent on the facet exposing on the particle surface, ant it increases in an order of without specific facet < [111]-facet < (010) facet. Figure Photocatalytic degradation of methylene blue (MB) by MW , MW , MW , MW , P25, and ST20 samples. 100

106 To explain the surface photocatalytic activity order, we investigated the surface electronic band structures of the TiO 2 nanocrystals. The UV-visible absorption spectra of six kinds of TiO 2 samples are shown in Figure 3.14(a). The absorption edges can be evaluated from the spectra as 385, 388, 390, 391.5, 395, and nm for MW , MW , P25, MW , MW , and ST20, respectively. TiO 2 is an indirect semiconductor, and the relation between absorption coefficient (A) and incident photon energy (hv) can be represented as Kubelka-Munk function A = B(hv-Eg) 2 /hv, where B and Eg are the absorption constant and bandgap energy. 43 The bandgap energy was estimated from transformed Kubelka-Munk function versus the energy of light (Figure 3.14), 21, 22 and the results are shown in Table 3.1. In the samples prepared from PA-HTO nanosheets, the bandgap increases in an order of MW < MW < MW < MW , which corresponds to the average crystal size decrease order of MW (150 nm) < MW (90 nm) < MW (20 nm), except the mesocrystal sample of MW This can be explained by bandgap blue shit of nanocrystal with decreasing the crystal size. 20 Although MW , ST20, and P25 have almost same crystal size of about 20 nm, they show different bandgap values, which can be attributed to different facets on the particle surface. Namely, the {010} facet has larger bandgap than [111]-facet and without specific facet (spherical particle) on the surface. Although the {010}-faceted anatase mesocrystals of MW are constructed from nanocrystals with size about 20 nm, its bandgap (3.03 ev) is much smaller than that (3.13 ev) of the (010)-faceted 20 nm nanocrystals of MW This suggests that the blue shift effect decreases in the mesocrystal structure. 101

107 Figure (a) UV-visible absorption spectra; (b) corresponding plots of transformed Kubelka-Munk function versus the energy of photon; (c) schematic illustration of the electronic band alignments of (І) ST20, (ІІ) MW , (ІІІ) MW , (ІV) P25, (V) MW , and (VІ) MW samples. It has been reported that the energy levels of highest valence band of anatase nanocrystals are similar even they have different facets on the surface, 22 and the energy level of highest valence band of anatase is evaluated as -7.5 ev. 44 Although P25 is a mixed phase of anatase and rutile, the photocatalytic activity is contributed mainly from anatase phase because the principal component (80%) is anatase and anatase exhibits higher photocatalytic activity than that of rutile. 45 Furthermore, it has been reported that the energy levels of highest valence band of P25 is almost same as anatase. 46 Therefore, the energy levels of lowest conduction band and highest valence band for these six kinds of TiO 2 nanocrystals can be illustrated in Figure 102

108 3.14(c) by assuming the energy levels of their highest valence bands are same as -7.5 ev and using the bandgap results of Figure 3.14(b). The result reveals that the energy level of the conduction band increases in the order of ST20 < MW < MW < P25 < MW < MW This result suggests that energy level of lowest conduction band of anatase increases in an order of without specific facet < [111]-facet < {010} facet. In the photocatalytic reaction, the TiO 2 nanocrystals with a higher energy level of lowest conduction band can generate more strongly reductive electrons for the photocatalytic reaction, which will show superior photocatalytic activity. Hence, the surface electronic band structure of the {010}-faceted anatase can provide high potential electrons for the photo reduction reaction. We think this is the reason why the {010}-faceted anatase nanocrystals exhibit high surface photocatalytic activity. Figure Current-voltage characteristic curves of DSSCs fabricated using MW , MW , MW , P25, and ST20 samples DSSCs Performance of TiO 2 Nanocrystals We think the surface electronic band structure of TiO 2 nanocrystals can affect also the DSSCs performance, and therefore, we investigated the DSSCs performance of 103

109 the TiO 2 nanocrystals with different facets on the surface. The I-V characteristics of DSSCs cells fabricated using anatase nanocrystals prepared from PA-HTO nanosheets, P25, and ST20 are shown in Figure The DSSCs cell parameters are given in Table 3.2. In the TiO 2 nanocrystals with similar crystal sizes of about 20 nm and different facets on the surface, the J sc value increases in an order of ST20 (10.6 ma/cm 2 ) < P25 (13.3 ma/cm 2 ) < MW (13.7 ma/cm 2 ). The V oc value increases slightly in an order of MW (0.66 V) < P25 (0.67 V) < ST20 (0.68 V), and the ff value increases slightly also in an order of P25 (0.54) < MW (0.55) < ST20 (0.57). The η value increases in an order of ST20 (4.07%) < P25 (4.84%) < MW (4.97%) that is corresponds to the increasing order of J sc value. These results reveal that the η values are strongly dependent on the J sc values, and MW with the highest J sc value gives the highest η value. The result suggests that the DSSC performance (η) enhances in the order of without specific facet < [111]-facet < {010} facet, which corresponds to the increasing orders of the bandgap and the lowest energy level of conduction band (Figure 3.14(c)), namely the DSSC performance is affected by the facet on the nanocrystals surface. It has been reported the N719 dye adsorption constant K ad on {010}-faceted anatase particles is larger than which on [111]-faceted P25 nanocrystals and spherical nanocrystal surface without specific facet, and the larger K ad, namely, strong anchoring of the dye molecules onto the TiO 2 surface can enhance J sc. 24 In the nanocrystals with the {010} facet on the surface but the different crystal sizes and morphologies, the J sc value increases in an order of MW (9.54 ma/cm 2 ) < MW (11.5 ma/cm 2 ) < MW (13.7 ma/cm 2 ), which corresponds to the decreasing order of the crystal size. The V oc value decreases in an order of MW = MW (0.66 V) << MW (0.75 V), and the ff value increases slightly in an order of MW (0.55) < MW (0.58) < 104

110 MW (0.60), which corresponds to the increasing order of crystal size. The increasing order of η value corresponds to the increasing order of J sc value, MW (4.32%) < MW (4.44%) < MW (4.97%). The above DSSC results reveal that the DSSCs performance is affected not only by the facet on the particle surface, but also by the crystal size and morphology. And the {010}-faceted anatase nanocrystals with size of about 20 nm are suitable for high performance DSSCs. Table 3.2. Cell performance parameters of DSSCs fabricated using TiO 2 samples. Sample J sc V oc ff η (ma/cm 2 ) (V) (%) P ST MW MW MW Conclusion The {010}-faceted anatase nanocrystals with controllable crystal size and morphology and {010}-faceted anatase mesocrystal platelike particles can be synthesized by microwave hydrothermal treatment of PA-HTO nanosheet solutions. The sizes and morphologies of the TiO 2 nanocrystals strongly depend on the reaction temperature and ph value of the reaction solutions. There are two kinds of reactions in the formation process of the anatase nanocrystals. One is the in situ topochemical conversion reaction from PA-HTO nanosheet structure to anatase structure, and 105

111 another is the dissolution-deposition reaction on the particle surface, which splits formed TiO 2 particles into small nanocrystals. The microwave hydrothermal process is suitable to control the crystal size and morphology due to its uniform heating mechanism. The UV-visible spectrum results reveal that the bandgap of the anatase nanocrystals is dependent on the facet exposes on crystal surface and also the crystal size. The {010}-faceted anatase nanocrystal exhibits larger bandgap value than which of [111]-faceted nanocrystal and spherical nanocrystal without specific facet. The photocatalytic activities of the anatase nanocrystals enhance with increasing the bandgap. The DSSCs results for TiO 2 nanocrystals with different facets on surface suggest that the J sc and η values increase in the order of spherical nanocrystals without specific facet < [111]-faceted nanocrystals < {010}-faceted nanocrystals, namely the {010}-faceted nanocrystals are promising for the high performance DSSCs References [1]. Fujishima, A.; Honda, K. Nature 1972, 238, [2]. Fujishima, A.; Rao, T. N.; Tryk, D. A. Photobiol. C 2000, 1, [3]. Nakata, K.; Fujishima, A. Photobiol. C 2012, 13, [4]. O Regan, B.; Grätzel, M. Nature 1991, 353, [5]. Grätzel, M. Nature 2001, 414, [6]. Grätzel, M. J. Photochem. Photobiol. C 2003, 4, [7]. Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Science 2001, 293, [8]. Fujishima, A.; Zhang, X.; Tryk, D. A. Surf. Sci. Rep. 2008, 63, [9]. Zhang, S.; Yang, X.; Qin, C.; Numata, Y.; Han, L. J. Mater. Chem. A 2014, 2, 106

112 [10]. Linsebigler, A. L.; Lu, G.; Yates, Jr, J. T. Chem. Rev. 1995, 95, [11]. Sato, H.; Ono, K.; Sasaki, T.; Yamagishi, A. J. Phys. Chem. B. 2003, 107, [12]. Diebold, U. Appl. Phys. A 2003, 76, [13]. Wen, P.; Itoh, H.; Feng, Q. Chem. Lett. 2006, 35, [14]. Wen, P.; Itoh, H.; Tang, W.; Feng, Q. Lamgmuir 2007, 23, [15]. Yang, H. G.; Sun, C. H.; Qiao, S. Z.; Zou, J.; Liu, G.; Smith, S. C.; Cheng, H. M.; Lu, G. Q. Nature 2008, 453, [16]. Wu, B.; Guo, C.; Zheng, N.; Xie, Z.; Stucky, G. D. J. Am. Chem. Soc. 2008, 130, [17]. Han, X.; Kuang, Q.; Jin, M.; Xie, Z.; Zheng. L. J. Am. Chem. Soc. 2009, 131, [18]. Amano, F.; Yasumoto, T.; Prieto-Mahaney, O-O.; Uchida, S.; Shibayama, T.; Ohtani, B. Chem. Commun. 2009, 17, [19]. Pan, J.; Liu, G.; Lu, G.; Cheng, H. Angew. Chem. Int. Ed. 2011, 50, [20]. Pan, J.; Wu, X.; Wang, L.; Liu, G.; Lu, G.; Cheng, H. Chem. Commun. 2011, 47, [21]. Xu, H.; Ouyang, S.; Li, P.; Kako, T.; Ye, J. ACS Appl. Mater. Interfaces 2013, 5, [22]. Xu, H.; Reunchan, P.; Ouyang, S.; Tong, H.; Umezawa, N.; Kako, T.; Ye, J. Chem. Mater. 2013, 25, [23]. Wen, P.; Tao, Z.; Ishikawa, Y.; Itoh, H.; Feng, Q. Appl. Phys. Lett. 2010, 97, [24]. Wen, P.; Xue, M.; Ishikawa, Y.; Itoh, H.; Feng, Q. ACS Appl. Mater. Interfaces 2012, 4,

113 [25]. Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H. Chem. Rev. 2010, 110, [26]. Wu, J.; Hao, S.; Lin, J.; Huang, M.; Huang, Y.; Lan, Z.; Li, P. Cryst Growth Des. 2008, 1, [27]. Yang, W.; Wang, Y.; Shi W. CrystEngComm 2012, 14, [28]. Shiu, J. W.; Lan, C. M.; Chang, Y. C.; Wu, H. P.; Huang, W. K.; Diau, E. W. G. ACS Nano 2012, 6, [29]. Shen, P-S.; Tai, Y-C.; Chen, P.; Wu, Y. C. J. Power Sources 2014, 247, [30]. Wen, P.; Itoh, H.; Tang, W.; Feng, Q. Microporous Mesoporous Mat. 2008, 116, [31]. Wen, P.; Ishikawa, Y.; Itoh, H.; Feng, Q. J. Phys. Chem. C 2009, 113, [32]. Corradi, A. B.; Bondioli, F.; Focher, B. J. Am. Ceram. Soc. 2005, 88, [33]. Komarneni, S.; Roy, R.; Li, Q. H. Mater. Res. Bull. 1992, 27, [34]. Komarneni, S.; Rajha, R. K.; Katsuki, H. Mater. Chem. Phys. 1999, 61, [35]. Wilson, G. J.; Will, G. D.; Frost, R. L.; Montgomery, S. A. J. Mater. Chem. 2002, 12, [36]. Wilson, G. J.; Matijasevich, A. S.; Mitchell, D. R. G.; Schulz, J. C.; Will, G. D. Langmuir 2006, 22, [37]. Wang, H. E.; Zheng, L. X.; Liu, C. P.; Liu, Y. K.; Luan, C. Y.; Cheng, H.; Li, Y. Y.; Martinu, L.; Zapien, J. A.; Bello, I. J. Phys. Chem. C 2011, 115, [38]. Manseki, K. Kondo, Y.; Ban, T.; Sugiura, T.; Yoshida, T. Dalton Trans. 2013, 42, [39]. Chen, C.; Sewvandi, G.A.; Kusunose, T.; Tanaka, Y.; Nakanishi, S.; Feng, Q. CrystEngComm 2014, 16,

114 [40]. Zheng, Y.; Shi, E.; Chen, Z.; Li, W.; Hu, X. J. Mater. Chem. 2001, 11, [41]. Feng, X.; Zhai, J.; Jiang, L. Angew. Chem., Int. Ed. 2005, 44, [42]. Li, G.; Gray, K. A. Chem. Mater. 2007, 19, [43]. Butler, M. A. J. Appl. Phys. 1977, 48, [44]. Nakamura, R.; Tanaka, T.; Nakato, Y. J. Phys. Chem. B 2004, 108, [45]. Carp, O.; Huisman, C.L.; Reller, A. Progress in Solid State Chemistry 2004, 32, [46]. Martin, S. T.; Herrmann, H.; Hoffmann, M. R. J. Chem. Soc. Faraday Trans. 1994, 90, CHAPTER III (Synthesis of titanium dioxide nanocrystals from H 1.07 Ti 1.73 O 4 layered titanate nanosheets by microwave hydrothermal process for high performance photocatalyst and dye-sensitized solar cells) was reprinted (adapted) with permission from (Changdong Chen, Linfeng Xu, Galhenage Asha Sewvandi, Takafumi Kusunose, Yasuhiro Tanaka, Shunsuke Nakanishi, and Qi Feng, Microwave-Assisted Topochemical Conversion of Layered Titanate Nanosheets to {010}-Faceted Anatase Nanocrystals for High Performance Photocatalyst and Dye-Sensitized Solar Cells, Crystal Growth & Design, 2014, 14 (11), ) American Chemical Society. 109

115 CHAPTER IV Synthesis of [111]- and {010}-Faceted Anatase TiO 2 Nanocrystals from Tri-Titanate Nanosheets and Their Photocatalytic and DSSCs Performances 4.1. Introduction Titanium dioxide (TiO 2 ) is an environment-friendly functional materials because of its excellent photocatalytic 1, 2 and dye-sensitized solar cells (DSSCs) performances, 3-5 low cost, and nontoxicity. 6 The TiO 2 materials can be prepared by two kinds of main chemical processes. One is dry chemical process, such as vacuum arc deposition, metal-organic vacuum deposition, and simple thermal pyrolysis in the gas phase or in the solid phase. 7-9 Another is wet chemical process, such as sol-gel and hydrothermal methods. 10, 11 Compare with the dry chemical process, the wet chemical process can controls crystal phase, particle size, morphology and surface structure more easily by changing concentration and ph value of reaction solution, reaction time, temperature, 12, 13 and precursor. Recently the effects of crystal facet on the photocatalytic and photovoltaic performances have attached much attention. Some excellent experimental results and important theoretical calculation to expound the effects of crystal facet on the photocatalytic and photovoltaic performances have been reported Wen et al. in our group have reported the first study on photocatalytic performance of anatase TiO 2 nanocrystals with specific crystal facet of high energy on the surface, and expound that {010}-faceted anatase TiO 2 nanocrystals exhibit much higher photocatalytic 110

116 activity than the normal spherical nanocrystals without specific facet on the surface in The results stimulate other researchers to study the effect of crystal facet on TiO 2 photocatalytic properties. Yang et al. have reported synthesis of high percentage of {001}-faceted TiO 2 particles using hydrofluoric acid solution, and their results of first-principle quantum chemical calculation suggest that the {001} facet of anatase is one of reactive facets for the photocatalytic reactions. 18 Meanwhile, Wu et al. have prepared {010}-faceted and {001}-faceted anatase TiO 2 nanocrystals, and found that the {010}-faceted nanocrystals exhibit higher photocatalytic activity than the {001}-faceted nanocrystals and commercial P25 sample in degradation of methyl orange. 19 After that, Han et al. have reported that the {001}-faceted anatase TiO 2 nanosheets exhibit the higher photocatalytic activity than the P25 sample. 20 Moreover, Amano et al. have reported {101}-faceted perfect octahedral anatase crystallites exhibit higher photocatalytic activity than the P25 sample. 21 These results indicate that the crystal facet strongly affects the photocatalytic performance of TiO 2. The facet relates both surface atomic arrangement structure and surface electronic band structure, which affect the photocatalytic activity The hydrothermal process is an effective method to synthesize TiO 2 nanocrystals with controlled particle size and morphology, where the control of crystal morphology means the control of crystal facet on the crystal surface. The hydrothermal reaction can controls the crystal growth process easily; therefore it is an effective method for the synthesis of TiO 2 with specific facet on the surface Usually, anatase TiO 2 nanocrystals with specific facet on the surface are synthesized by hydrothermal treatments of the titanium precursors in solutions containing the hydrofluoric acid and organic compounds as crystal growth directing agents. 18, 20 In such solution reaction process, the TiO 2 particles are formed by dissolution-decomposition mechanism. Therefore, synthesis of small nanocrystals with specific facet on surface, such as 111

117 nanocrystals with size below 50 nm, is difficult. We have reported a unique and useful hydrothermal soft chemical process for the synthesis of anatase nanocrystals with specific facet on the surface from layered titanate nanosheets by topochemical conversion reaction. The {010}-faceted anatase nanocrystals can be obtained by hydrothermal treatment of H 1.07 Ti 1.73 O 4 layered titanate nanosheets with a 17, lepidocrocite-like structure. And [111]-faceted anatase nanocrystals can be obtained by topochemical conversion of H 2 Ti 4 O 9 layered titanate nanosheets to anatase under microwave-assisted hydrothermal conditions. 35 These results suggest that the anatase nanocrystals with different facet on the surface can be synthesized by using different layered titanate nanosheets as the precursor. Although the syntheses of TiO 2 from layered tri-titanates Na 2 Ti 3 O 7 and H 2 Ti 3 O 7 precursors have been reported, but the particle size of TiO 2 products are large. 36, 37 Furthermore, synthesis of TiO 2 form the exfoliated H 2 Ti 3 O 7 nanosheets has not been reported yet. In this chapter, we describe hydrothermal synthesis TiO 2 nanocrystals from the exfoliated H 2 Ti 3 O 7 layered titanate nanosheet precursor. We report the first study on the synthesis of anatase nanocrystals with {010} facet and [111]-facet on the surface from the exfoliated H 2 Ti 3 O 7 nanosheets, and control of their crystal size and morphology. This success gives us an opportunity to study the effects of crystal facet and size on the bandgap, surface electronic band structure, and photocatalytic activity of TiO 2 nanocrystals. The results reveal that the {010}-faceted anatase nanocrystals exhibit higher photocatalytic activity than [111]-faceted anatase nanocrystals and commercial P25 sample, due to the larger bandgap of {010} facet Experimental Section Preparations of Na 2 Ti 3 O 7 and H 2 Ti 3 O 7 Samples 112

118 The sodium tri-titanate (Na 2 Ti 3 O 7 ) was prepared by solid state reaction. A mixture of Na 2 CO 3 (Wako, 99.5%) and TiO 2 (Wako, anatase form, 98.5%) in a molar ratio of 1.1:3 were mixed by ball-milling for 24 hrs, and then the mixture were heated at 900 o C for 12 hrs. The 10 g of obtained Na 2 Ti 3 O 7 samples was acid-treated with a 1M HNO 3 solution (1 L) at 80 o C for 24 hrs under stirring conditions to exchange Na ions with H ions to obtain H + -form tri-titanate H 2 Ti 3 O 7 samples. The acid treatment was repeated again to complete the ion-exchange reaction. And then the tri-titanate H 2 Ti 3 O 7 sample was washed with distilled water several times and dried using a freeze drier Preparation of H 2 Ti 3 O 7 Nanosheet Colloidal Solution and TiO 2 Nanocrystals The 2 g of tri-titanate H 2 Ti 3 O 7 sample was hydrothermally treated in a 12.5% tetramethylammonium hydroxide (TMAOH, Wako) solution (20 ml) at 130 o C for 24 hrs under stirring conditions to intercalate TMA + ions into the interlayer of the tri-titanate H 2 Ti 3 O 7 and to obtain a TMA + -form tri-titanate H 2 Ti 3 O 7 sample. The obtained TMA + -form H 2 Ti 3 O 7 was dispersed in 150 ml of distilled water at room temperature by stirring for 48 hrs to exfoliate the TMA + -form tri-titanate H 2 Ti 3 O 7 sample to its nanosheets. The tri-titanate H 2 Ti 3 O 7 nanosheet colloidal solution was obtained after filtration to remove the unexfoliated particles and the tri-titanate H 2 Ti 3 O 7 nanosheet solution was named TMA-HTO nanosheet solution. The TiO 2 nanoparticles were prepared by hydrothermal treatment of the obtained TMA-HTO nanosheet solution after adjusting to a desired ph value with a 6 M HCl solution. The 40 ml of ph adjusted TMA-HTO nanosheet solution was sealed in a Teflon-lined stainless steel vessel with internal volume of 80 ml, and then 113

119 hydrothermally treated at a desired temperature for 24 hrs. After the hydrothermal treatment, the samples were washed with distilled water several times, and finally dried using a freeze drier. The obtained TiO 2 sample is named TMA-A-B, where A and B are the desired temperature of hydrothermal treatment, and desired ph value of the nanosheet solution, respectively Photocatalytic Characterization The 20 mg of obtained TiO 2 sample was added into a 100 ml of 10 ppm methylene blue (MB) aqueous solution and stirred for 2 hrs to disperse the TiO 2 sample well in the MB solution and to reach MB adsorption equilibrium on TiO 2 nanocrystals surface under dark conditions without UV-irradiation. And then the suspension was illuminated by a 100 W ultraviolet lamp with nm wavelength (UVA, Asahi Spectra, LAX-Cute) at room temperature, and the ultraviolet lamp was located at 20 cm away from the suspension under continuously stirring conditions. At every 20 min interval, 3 ml solution was drawn from the suspension and immediately centrifuged to separate the TiO 2 samples from the solution under rpm for 10 min. The MB concentration in the solution was determined by using a Shimadzu UV-2450 spectrophotometer. The degradation efficiency of MB solution by the photocatalytic reaction was calculated from the variation of the MB concentration in the solutions before and after UV light irradiation. The P25 nanocrystals sample was purchased from Degussa. The ST20 nanocrystals sample was synthesized by hydrothermal treatment of a commercial anatase nanoparticle sample (ST01, Ishihara Sangyo) with crystal size of 7 nm at 200 o C for 12 hrs Fabrication and Characterization of Dye-Sensitized Solar Cells 114

120 A TiO 2 nanocrystal paste was prepared by mixing TiO 2 nanocrystal sample (0.5 g), ethanol (2.5 g), α-terpineol (2.0 g), 10 wt% solution of ethyl-cellulose 10 (8-14mPas, 1.4 g), and 10 wt% solution of ethyl-cellulose 45 (45-65mPas, 1.1 g). The mixture was dispersed by ultrasonication for 30 min, and then ball-milling for 72 hrs. After ball-milling, the ethanol was removed from the mixture using a rotary-evaporator. The TiO 2 photoelectrode was prepared as follows. The fluorine-doped tin oxide (FTO) conducting glass plate was cleaned in distilled water and acetone by ultrasonication for 10 min, orderly. Then the FTO glass plate was dipped in 0.1M titanium tetraisopropoxide (TTIP) solutions for 1 min and washed with distilled water and ethanol, dried at room temperature and calcined at 480 o C for 1 h to coat the FTO glass surface with a dense TiO 2 thin film. The prepared TiO 2 paste was coated (10 10 mm) on the TTIP-treated FTO glass plates by screen printing technique and kept in an ethanol box until the TiO 2 film surface smoothly and then dried at 120 o C for 10 min. This process was repeated to obtain a desired thickness of TiO 2 film. After the TiO 2 paste coating, the TiO 2 film was calcined at 450 o C for 30 min to obtain a TiO 2 porous electrode. The TiO 2 porous electrodes were dipped in the 0.1M TTIP solution for 1 min, and washed with distilled water and ethanol, dried at room temperature and calcined at 480 o C for 1 h again. After cooling to 80 o C, the TiO 2 porous electrodes were soaked in a M N719 dye (di-tetrabutylammonium cisbis(isothiocyanato) bis(2,2'-bipyridyl-4,4' -dicarboxylato)ruthenium(ii)) solution for 24 hrs at room temperature, and then washed with a t-butyl alcohol and acetonitrile (v/v=50%:50%) mixed solvent. The DSSCs were comprised of a dye-adsorbed TiO 2 electrode as an anode, a Pt-coated FTO glass as a cathode, and an electrolyte solution between the anode and the cathode. The electrolyte solution contains 0.1M LiI, 0.01M I 2, 0.6M of 1-butyl-3-n-propylimidazolium iodide (BMII), 0.4M 4-tert-butylpyridine (TBP) and 115

121 0.1M guanidine thiocyanate (GT) in acetonitrile and valeronitrile (v/v=85%:15%). The photocurrent voltage characteristic curves for the DSSCs were measured using a Hokuto-Denko BAS100B electrochemical analyzer under irradiation with simulated sunlight of AM 1.5 (100mW/cm 2 ), using a sunlight simulator (YSS-E40, Yamashita Denso). A light-passing mask was fixed on the surface of FTO glass of the anode to set the effectively irradiating area on the cell as 0.25 cm Physical Analysis The crystal structure of the sample was investigated using a powder X-ray diffractometer (Shimadzu, Model XRD-6100) with Cu Kα (λ= nm) radiation. The size and morphology of the samples were characterized by field emission scanning electron microscopy (FE-SEM) (Hitachi, Model S-900). Transmission electron microscopy (TEM) observation and selected-area electron diffraction (SAED) were performed on a JEOL Model JEM-3010 system at 300 kv. The specific surface area of TiO 2 nanoparticles was calculated from the adsorption data using the Brunauer-Emmett-Teller (BET) method. The thickness of the TiO 2 porous film was measure using surface texture measuring instrument (SURFCOM 480A) Results and Discussion Preparations of Tri-Titanate H 2 Ti 3 O 7 Nanosheet Solution The sodium tri-titanate (Na 2 Ti 3 O 7 ) as the precursor was prepared by heating a mixture of Na 2 CO 3 and TiO 2 at 900 o C for 12 hrs. And a Na + /H + ion-exchange treatment of Na 2 Ti 3 O 7 was carried out to prepare H 2 Ti 3 O 7 sample. XRD patterns of Na 2 Ti 3 O 7 and H 2 Ti 3 O 7 samples were shown in Figure 4.1. The Na 2 Ti 3 O 7 sample shows an XRD pattern corresponding to JCPDS File No (monoclinic system, 116

122 space group P21/m, a= , b=0.3803, c= nm, β=101.6 o ), indicating that Na 2 Ti 3 O 7 phase with the layered structure is obtained. The basal spacing of the Na 2 Ti 3 O 7 layered structure is 0.83 nm. After the Na + /H + ion-exchange reaction, the H 2 Ti 3 O 7 sample maintains the essential layered structure but the basal spacing changes to 0.79 nm, and XRD pattern corresponds to JCPDS File No (monoclinic system, space group C2/m, a=1.6023, b=0.3749, c= nm, β= o ). Figure 4.1. XRD patterns of (a) Na 2 Ti 3 O 7, (b) H 2 Ti 3 O 7, and (c) TMA + -form H 2 Ti 3 O 7 samples. To exfoliate H 2 Ti 3 O 7 layered structure into its nanosheets, tetramethylammonium ions (TMA + ) are intercalated into the H 2 Ti 3 O 7 layered structure by hydrothermal treatment of H 2 Ti 3 O 7 sample in a TMAOH solution at 130 o C. After the hydrothermal treatment, the TMA + -form H 2 Ti 3 O 7 sample maintains the layered structure, which shows main diffraction peaks of (200), (400), and (600) with d-values of 1.79, 0.895, and nm, respectively (Figure 4.1(c)). The increase of basal spacing from 0.79 nm of H 2 Ti 3 O 7 to 1.79 nm reveals that TMA + ions are intercalated into the interlayer space of H 2 Ti 3 O 7 sample. The higher hydrothermal treatment temperature of 130 o C is necessary to intercalate TMA + ions into the H 2 Ti 3 O 7 structure than those for 117

123 H 1.07 Ti 1.73 O 4 with lepidocrocite-like layered structure and H 2 Ti 4 O 9 layered titanates. TMA + ions can be intercalated into H 1.07 Ti 1.73 O 4 layered titanate at room temperature, 33 and into H 2 Ti 4 O 9 layered titanate at 90 o C. 35 This fact can be explained by the different charge density of the host layer of the layered titanates, which increases in an order of H 1.07 Ti 1.73 O 4 < H 2 Ti 4 O 9 < H 2 Ti 3 O 7. The layered structures are held by attracting between the negatively charged host layers and the positively charged H + or H 3 O + ions in the interlayer spaces. The attracting between the host layers and interlayer ions enhance with increasing charge density. Therefore, the higher charge density, it is difficult to be interlayered and exfoliated into the nanosheets. When TMA-HTO sample is dispersed in distilled water, the layered structure is exfoliated into nanosheets and the TMA-HTO nanosheet colloidal solution is obtained. Figure 4.2. SEM images of (a) Na 2 Ti 3 O 7 and (b) H 2 Ti 3 O 7 samples, (c) crystal structure of H 2 Ti 3 O 7, (d) TEM image and SAED pattern of TMA-HTO nanosheet sample, and (e, f) HR-TEM images of TMA-HTO nanosheet sample. 118

124 Figures 4.2(a) and (b) show the typical SEM images of the Na 2 Ti 3 O 7 and H 2 Ti 3 O 7 samples. The Na 2 Ti 3 O 7 sample exhibits rodlike particle morphology with a size of about 10 µm in length and 2 µm in width. After ion-exchange reaction, the H 2 Ti 3 O 7 sample maintains the rodlike particle morphology. The axis-direction of the rodlike H 2 Ti 3 O 7 particles corresponds to b-axis direction (Figure 4.2(c)). 38, 39 Figure 4.2(d-f) shows the HR-TEM images and SAED pattern of TMA-HTO nanosheet sample. After the exfoliation treatment, the particle morphology changes to nanosheetlike morphology. In the SAED pattern of the nanosheet sample, (020) and (003) diffractions are observed, the HR-TEM images also accord with the result of SAED patterns, which is indicated the basal plane of the HTO nanosheet is parallel to b- and c-axis, namely the exfoliation occurs along a-axis direction. 119

125 Figure 4.3. XRD patterns of the samples obtained by hydrothermal treatment of TMA-HTO nanosheet solutions at 140 o C-200 o C. : H 2 Ti 3 O 7, : anatase, : rutile, : brookite Synthesis of TiO 2 Nanocrystals from TMA-HTO Nanosheet Colloidal Solution Figure 4.4. Dependence of the products on the reaction temperature and the ph value, : anatase, : rutile, : anatase+rutile, : layered titanate+anatase, : anatase+rutile+brookite. The prepared TMA-HTO nanosheet colloidal solutions were hydrothermally treated to synthesize TiO 2 nanocrystals. The XRD patterns of the products prepared under various temperature and ph conditions are shown in Figure 4.3. The dependence of the products on the hydrothermal reaction conditions is summarized in Figure 4.4. In a range of ph < 1, rutile, anatase, brookite phases can be formed, and the percentages of brookite and anatase phases decrease gradually with increasing the hydrothermal reaction temperature. The brookite phase disappears above 170 o C, and anatase phase disappears above 200 o C. This result reveals that the stability increases in an order of brookite < anatase < rutile under the acidic conditions of ph < 1. Furthermore, the single anatase phase can be obtained in a wide ph range of 1 < ph < 12. The 120

126 crystallinity of anatase phase increases with increasing of the temperature and ph value (see Figure 4.5). The percentage of the TMA-HTO layered phase decreases with increasing of the temperature at the same ph value (see Figure 4.3), and the TMA-HTO nanosheets are transformed completely to anatase phase above 180 o C in this ph range (Figure 4.4). Furthermore, the EDS analysis results reveal that without sodium residual in the synthesized TiO 2 samples (see Figure 4.6). Figure 4.5. Dependences of the (101) peak intensity of anatase on the ph values of nanosheet solution. Figure 4.6. EDS analysis results of synthesized TiO 2 samples. Without sodium peak was observed, indicating without sodium residual in the synthesized TiO 2 samples. 121

127 Figure 4.7 shows the FE-SEM images of the products obtained by hydrothermal treatment of the TMA-HTO nanosheets. At 160 o C-pH 0.65, the main particles are nanorod-like particles with a size of about 50 nm in length and 20 nm in width, and some large rod-like particles with a size of about 500 nm in length and 50 nm in width, are observed also (Figure 4.7(a)). The nanorod-like particles and the large rod-like particles can be assigned to anatase and rutile phases, respectively, because the XRD result indicates that the sample contains anatase main phase and small amount of rutile (see Figure 4.3). At 170 o C-pH 0.56, the main rod-like particles with size of about 500 nm in length and 100 nm in width and small amount of small nanoparticles with size of about 20 nm are observed (Figure 4.7(b)). The rod-like particles correspond to rutile phase because the XRD result indicates that the main phase is rutile (see Figure 4.3). The size of rutile particles increases with increasing the reaction temperature, and become about 1 µm in length at 200 o C-pH 0.65 (Figure 4.7(c)). The result suggests that the rutile phase is formed by a dissolution-deposition reaction under the low ph conditions Around ph 3, TMA sample is a mixture of anatase and layered phases, and shows nanorod-like particles with size of about 50 nm in length and 20 nm in width and nanosheet-like particles (Figure 4.7(d)). The nanorod-like and nanosheet-like particles can be assigned to anatase and layered phases, respectively. The nanosheet-like particles disappear, and nanorod-like particles of anatase are formed in TMA sample (Figure 4.7(e)). The nanorod-like particles of anatase tend to change to tetragonal nanoparticles with increasing the reaction temperature to 200 o C around ph 3 (Figure 4.7(f)). 122

128 Figure 4.7. FE-SEM images of (a) TMA , (b) TMA , (c) TMA , (d) TMA , (e) TMA , (f) TMA , (g) TMA , (h) TMA , (i) TMA , (j) TMA , (k) TMA , and (l) TMA samples. Around ph 7, TMA sample containing the layered and anatase phases shows nanosheet-like particles, tetragonal particles and nanorod-like particles (Figure 4.7(g)). Tetragonal nanoparticles of anatase with size about 100 nm are formed in TMA sample (Figure 4.7(h)). At 180 o C and 200 o C, nanorod-like anatase particles with a size about 100 nm in length and 50 nm in width are formed (Figures 4.7(i) and (j)). In TMA sample, nanosheet-like particles of the layered phase and nanoleaf-like particles of anatase phase with a size about 500 nm in length and 50 nm in width are observed (Figure 4.7(k)), and the nanoleaf-like particles of anatase phase 123

129 are also observed in TMA sample (Figure 4.7(l)). The above results reveal that the particle morphology of anatase is strongly dependent on the hydrothermal reaction temperature and ph value of the reaction solution, which is similar to the cases of hydrothermal reactions of the lepidocrocite-like H 1.07 Ti 1.73 O 4 and H 2 Ti 4 O 9 layered titanate nanosheets Nanostructural Study of TiO 2 Nanocrystals and Transformation Reaction Mechanism from TMA-HTO Nanosheets to TiO 2 Nanocrystals Figure 4.8. (a, b) TEM images and (c) FFT diffraction pattern of TMA sample, and (d) structural transformation from layered structure to anatase structure. The nanostructures of TiO 2 nanocrystals and their formation mechanism from TMA-HTO nanosheets were studied using TEM. Figure 4.8 shows the TEM images and fast Fourier Transform (FFT) diffraction pattern of the TMA sample containing the HTO layered and anatase phases. A nanosheet-like particle shows the (003) and (010) diffraction spots of HTO layered phase and the (101) and (011) diffraction spots of anatase phase in the FFT diffraction. This nanosheet-like particle is an intermediate product of topochemical transformation reaction from the layered 124

130 phase to the anatase phase. There is specific crystallographic relationship between the layered structure and anatase structure in the topochemical structural transformation reaction as shown in Figure 4.8(d). The (100) facet of the layered phase corresponds to a facet vertical to [111]-direction of anatase, and these facets are parallel to the basal plane of the nanosheet-like particle. The [010] orientation of the layered phase is transformed to the [101] orientation of anatase phase by rotating an angle of 20 o on the basal plane of the nanosheet-like particle (Figure 4.8(c)). This topochemical reaction causes formation anatase nanosheet with a crystal facet vertical to [111]-direction on the nanosheet-like surface. We call it [111]-faceted anatase nanocrystal. The [111]-facet is different from {111}-facet in the tetragonal system of 34, 35, 43 anatase phase. The TEM images and FFT diffraction patterns of TiO 2 nanocrystals synthesized from TMA-HTO are shown in Figure 4.9. In the TMA sample, rod-like crystal shows lattice fringes with d-values of 0.32 and 0.25 nm in TEM image and their diffraction spots in FFT pattern, which correspond to (110) and (101) planes of rutile phase, respectively (Figure 4.9(a)). This result indicates that the rod-like crystals correspond to rutile phase, the two side faces of the rod-like crystal correspond to {110} facet. In the TMA sample, rod-like crystal shows rutile lattice fringes of (110) plane with d-values of 0.32 and (111) plane with d-value of 0.22 nm in TEM image and the corresponding diffraction spots in FFT pattern (Figures 4.9(b, c)), indicating that the two side faces also correspond to (110) facet of rutile. 125

131 Figure 4.9. TEM images and FFT diffraction patterns of (a) TMA , (b, c) TMA , (d) TMA , (e) TMA , (f) TMA , (g) TMA , (h) TMA , and (i) TMA samples. In TMA sample, main particles are nanorod-like anatase particles. These nanocrystals exhibit the lattice fringes of (101) and (011) planes with a d-value of 0.35 nm (Figure 4.9(d)). This result reveals that the four side faces of nanorod correspond to {101} facet and the basal plane to [111]-facet. The nanorod-like anatase particles in TMA sample also show the lattice fringes and the diffraction spots of (101) and (011) planes with a d-value of 0.35 nm (Figure 4.9(e)). The above results suggest that [111]-faceted anatase nanocrystals are formed mainly from the TMA-HTO nanosheets around ph 3 under hydrothermal conditions. A few of rhombic and hexagonal anatase TiO 2 nanocrystals with {010} facet on the surfaces were 126

132 observed also around ph 3 (see Figure 4.10). Figure TEM images SAED diffraction pattern of (a, b, c) TMA samples. In TMA sample, the main particles are anatase tetragonal nanocrystals which have [111]-facet on the surface (Figure 4.9(f)). In TMA sample, the nanorod-like anatase nanocrystals exhibit lattice fringes with d-values of 0.35 nm and 0.46 nm, respectively, corresponding to (101) and (002) planes in its HR-TEM image and its FFT diffraction pattern (Figure 4.9(g)). The result reveals that the basal plane of the nanorod-like anatase TiO 2 nanocrystals correspond to a (010) facet, and other two side faces correspond to (100) facet. The long-axis direction of the nanorod corresponds to [001]-direction. These results suggest that at around ph7, the [111]-faceted anatase nanocrystals are formed preferentially at low temperature, and {010}-faceted anatase nanocrystals are formed preferentially at high temperature. At around ph 12, the typical nanoleaf-like nanocrystals are obtained. Which exhibit lattice fringes with a d-value of 0.35 nm corresponding to (101) plane and (-101) plane of anatase phase (Figure 4.9(h)), and d-values of 0.35 nm and 0.48 nm corresponding to (101) plane and (002) plane of anatase phase (Figure 4.9(i)) in HR-TEM images. The result reveals the nanoleaf-like anatase nanocrystals expose {010}-facet on the basal plane. Furthermore, the long-axis and width-axis of the nanoleaf-like nanocrystal along with [001]-orientation and [100]-orientation, respectively. 127

133 Figure Transformation reaction mechanism from TMA-HTO nanosheets to TiO 2 nanocrystals. On the basis of the results described above, we give a reaction mechanism for the transformation of TMA-HTO nanosheets to TiO 2 under the hydrothermal conditions as shown in Figure The structure and particle morphology of TiO 2 are strongly dependent on ph value of the reaction solution and reaction temperature. The rutile phase with rod-like morphology is formed under low ph conditions of ph < 1 by dissolution-decomposition reaction, due to higher solubility of TiO 2 under the low ph conditions. 17 Around ph 3, the TMA-HTO nanosheets are transformed firstly to [111]-faceted nanosheet-like anatase particles by the topochemical reaction (Figure 128

134 4.9). And then the nanosheet-like anatase particle is split into nanorod-like anatase particles by dissolution reaction along {101} facet. The formed nanorod-like anatase particles have basal plane corresponding to [111]-facet and four sides faces corresponding to {101} facets. At around ph 7 and low temperature, similar reactions at ph 3 occur, and then tetragonal anatase nanoparticles with basal plane corresponding to [111]-facet and four side faces to {101} facets. Around ph 12, we think that the TMA-HTO nanosheets are transformed firstly to {010}-faceted nanosheet-like anatase particles by the topochemical reaction. 34, 35 And then the nanosheet-like anatase particle is split into nanoleaf-like anatase particles by dissolution reaction along (100) facet. The formed nanosheet-like anatase particles have basal plane corresponding to {010} facet and long-axis corresponding to [001]-direction. At around ph 7 and high temperature, similar reactions at ph 12 occur. Namely the TMA-HTO nanosheets are transformed firstly to {010}-faceted nanosheet-like anatase particles, and then the {010}-faceted nanosheet-like anatase particles is split into nanorod-like anatase nanoparticles with basal plane and two sides faces corresponding to {010} facets, and long-axis corresponding to [001]-direction. The results described above reveal that the anatase nanocrystals with different facets on their surfaces can be obtained under different ph and temperature conditions, and two kinds of typical facets expose on anatase nanocrystal surface can be obtained. One is {010}-faceted, and another is [111]-faceted anatase nanocrystals. The results are similar to formation of anatase nanocrystals form tetra-titanate nanosheets under hydrothermal conditions in our previous study. 35 The formation of these two kinds of typical facets expose on anatase nanocrystal surface can be explained by splitting the layered titanate nanosheets along different facets under different ph conditions. The splitting along {001} facet of HTO nanosheet causes formation of [111]-faceted 129

135 anatase nanoparticles, while that along {010} facet of HTO nanosheet causes formation of {010}-faceted anatase nanoparticles. 35 The above results also reveal that the TiO 2 nanocrystals can be formed from TMA-HTO nanosheets by two kinds of simultaneous reactions. 17, 32, 33 One is the in situ topochemical conversion reaction, which transforms the layered structure of TMA-HTO nanosheets to the anatase TiO 2 structure. Another is the dissolution-deposition reaction, which transforms the nanosheet-like particle morphology to other nanoparticle morphology. Therefore, controlling the dissolution-deposition reaction is significant to control the particle morphology and size Electronic Band Structure and Photocatalytic Activity of TiO 2 Nanocrystals To study crystal facet effect on the photocatalytic activity of TiO 2 nanocrystals, the photocatalytic degradations of methylene blue (MB) dye by TiO 2 nanocrystals were measured under UV-light irradiating conditions. In the photocatalytic study, two anatase nanocrystal samples with uniform crystal morphology, TMA ([111]-faceted anatase nanorods, Fig. 4(f)) and TMA ({010}-faceted anatase nanorods, Fig. 4(g)), and a rutile nanocrystal sample of TMA (nanorods, Fig. 4(c)) are chosen to compare with commercial P25 TiO 2 nanocrystal sample as standard sample. The P25 sample is a well-known high active photocatalytic TiO 2 sample, which contains 80% anatase phase with particle size of about 20 nm and 20% rutile phase with particle size of about 80 nm, where half of anatase is [111]-faceted tetragonal nanocrystal and other half is sphere nanocrystal without specific on the surface. 34, 43 Furthermore, an anatase nanocrystal sample of ST20 with particle size of 130

136 about 20 nm and spherical particle morphology was used also as a standard sample without specific facet on the surface. The BET specific surface area, particle size, and other parameters of these TiO 2 samples are shown in Table 4.1. To understand the facet effect on the photocatalytic activity easily, the facet proportions exposed on the surface for the samples with specific facet were estimated based on their crystal morphologies (Fig. 4) and results of the nanostructural studies (Fig. 6). The result suggests that TMA , TMA , and P25 contain 85% of {010}-facet, 43% of [111]-facet, and 22% of [111]-facet, respectively. Table 4.1. Crystal phase, exposed facet, surface area, crystal size, and bandgap energy for TMA , TMA , TMA , ST20 and P25 samples. Sample Crystal phase Exposed S BET Crystal size Eg facet (m 2 /g) (nm) (ev) (Anatase) P25 Anatase (80%) [111] (40%) (Anatase) 3.04 Rutile (20%) None (40%) 80 (Rutile) ST20 Anatase None TMA Rutile (width length) TMA Anatase [111] TMA Anatase {010} (width length) The results of photocatalytic degradation of MB reveal that photocatalytic activity (mg/g) increases in an order of TMA < ST20 < TMA < TMA < P25 in Figure 4.12(a). The P25 and TMA exhibit highest and lowest photocatalytic activities, respectively in these samples. It is well-known 131

137 that the photocatalytic activity is strongly dependent on the surface area of TiO 2 because the photocatalytic reaction occurs on the TiO 2 surface, and also anatase phase 44, 45 exhibits higher photocatalytic activity than that of rutile phase. The low photocatalytic activity of TMA is due to its small surface area and rutile phase. The surface area of the anatase samples increase in an order of TMA < TMA < P25 < ST20, which does not corresponds to the increasing order of the photocatalytic activity. We think this difference can be explained by different photocatalytic activities of the different crystal facets exposing on the TiO 2 surface. To understand the surface photocatalytic activity, we evaluated the MB degradation amount by per surface area of TiO 2 sample (mg(mb)/m 2 (TiO 2 surface area)), and the result is shown in Figure 4.12(b). Therefore, the area surface photocatalytic activity (mg/m 2 ) increases in an order of TMA < ST20 < P25 < TMA < TMA By considering the facet proportions exposed on the surface, this result suggest that the area surface photocatalytic activity increases in an order of rutile < anatase (non-facet) < anatase ([111]-facet) < anatase ({010} facet). Figure Photocatalytic degradations of methylene blue (MB) by TMA (rutile), TMA ([111]-faceted anatase), TMA ({010}-faceted anatase), P25 (partially [111]-faceted anatase), and ST20 (non-faceted anatase) samples presented by (a) mg(mb)/g(tio 2 ) and (b) mg(mb)/m 2 (TiO 2 ), respectively. 132

138 Furthermore, we study the surface electronic band structures of the TiO 2 nanocrystals in order to explain the surface photocatalytic activity. The UV-visible absorption spectra of TiO 2 samples are shown in Figure 4.13(a). Because of the TiO 2 is an indirect semiconductor, the relation between absorption coefficient (A) and incident photon energy (hv) can be represented as Kubelka-Munk function A = B(hv-E g ) 2 /hv, where B and E g are the absorption constant and bandgap energy. 46 The bandgap energy can be estimated from transformed Kubelka-Munk function versus the energy of light (Figure 4.13(b)), 26, 47 and the results are shown in Table 4.1. The bandgap increases in an order of TMA < ST20 < P25 < TMA < TMA , which corresponds to the increasing order of surface photocatalytic activity. The different bandgap values can be attributed to their different crystal phases and exposed facets on the particle surface Namely, the rutile phase sample has smaller bandgap than anatase phase samples. 48 In the anatase samples, the bandgap increases in an order of ST20 (non-facet) < P25 (22%-[111]-facet) < TMA (43%-[111]-facet) < TMA (85%-{010}-facet), namely non-facet < [111]-facet < {010}-facet. The result reveals that the bandgap is dependent on the facet on the surface

139 Figure (a) UV-visible absorption spectra; (b) the corresponding plots of transformed Kubelka-Munk function versus the energy of photon; (c) schematic illustration of the electronic band alignments of (I) TMA (rutile), (II) ST20 (non-faceted anatase), (III) P25 (partially [111]-faceted anatase), (IV) TMA ([111]-faceted anatase) and (V) TMA ({010}-faceted anatase) samples. Basis on the bandgap results described above and the energy level of highest valence band reported, we illustrate the band structures of these TiO 2 nanoparticle sample in Figure 4.13(c). The literature values of -7.1 ev and -7.5 ev are used as highest valence band level for rutile 49 and anatase 50 phases, respectively. It has been reported that the energy levels of highest valence band of anatase nanocrystals are almost independent on their crystal size and facet on the surface Therefore, the same value (-7.5 ev) of highest valence band is used for all anatase nanocrystal samples. Although P25 is a mixed phase of anatase and rutile, -7.1 ev is used as the highest valence band value for P25, because its principal component (80%) is anatase 134

140 and anatase exhibits higher photocatalytic activity than that of rutile, furthermore it would be reasonable to use the energy level of highest valence band of anatase for P25 photocatalytic study. The result reveals that the lowest energy level of the conduction band increases in an order of ST20 (non-facet) < P25 (22%-[111]-facet) < TMA (43%-[111]-facet) < TMA (85%-{010}-facet) for the anatase samples. This result suggests that energy level of lowest conduction band of anatase increases in an order of non-facet < [111]-facet < {010}-facet. In the photocatalytic reaction, the TiO 2 nanocrystals with a higher energy level of lowest conduction band can generate more strongly reductive electrons for the photocatalytic reaction, meaning stronger reduction ability and superior photocatalytic activity. Hence, the surface electronic band structure of the {010}-faceted anatase can provide high potential electrons for the reduction reaction. Although the rutile sample (TMA ) has a higher energy level of lowest conduction band than the anatase samples, but its energy level of highest valence band is also higher than that of the anatase samples. The higher energy level of highest valence band will show weaker oxidation ability in the photocatalytic reaction DSSC Performance of [111]-Faceted TiO 2 Nanocrystals In our previous studies, it has found that the DSSC performance is dependent the facet of anatase nanocrystals, which increases in an order of non-facet (ST20) < [111]-facet (P25) < {010} facet of anatase nanocrystals. 34, 43 But P25 is mixed phase of anatase and rutile, and only 40% of anatase is [111]-faceted anatase nanocrystal. To confirm the DSSC performance of the [111]-faceted anatase nanocrystal, we investigate the DSSC performance of [111]-faceted anatase sample of TMA

141 and compare with that of P25 because they have similar particle size and the particle size of about 30 nm is suitable for DSSCs. The cell parameters and I-V characteristics of DSSCs cells fabricated using TMA and P25 with the similar thickness of porous TiO 2 film are given in Table 4.2 and in Figure TMA exhibits larger J sc and V oc values (11.9 ma/cm 2 and 0.67 V) than which (11.5 ma/cm 2 and 0.61 V) of P25, and almost same ff value (0.6), resulting larger η value of TMA than that of P25. The result suggests that the DSSC performance (η) enhances in the order of non-facet < [111]-facet < {010} facet. The increasing order corresponds to the increasing orders of the bandgap and the lowest energy level of conduction band (Figure 4.13(c)). We also confirmed that crystal morphology and facet of the anatase nanocrystals did not change after calcination treatment in the fabrication process of TiO 2 electrode by FE-SEM and TEM analyses (see Figure 4.14 and 4.15). Figure Current-voltage characteristic curves of DSSCs fabricated using TMA and P25 samples. 136

142 Figure TEM images of TMA anatase nanocrystal samples after the calcination at 480 o C for 1h. The anatase nanocrystals keep their crystal morphologyies and facets exposed on the crystal surfaces after the calcination. The different DSSCs performances may be due to different dye adsorption behavior on the different facets. It has been reported the N719 dye adsorption constant K ad on anatase nanocrystal surface increases in an order of non-facet < [111]-facet < {010} facet, 35 namely the strong adsorption of the dye molecules onto the TiO 2 surface can enhance η value The dye molecules can be anchored more strongly on [111]-faceted anatase nanocrystal surface of TMA than that on P25 nanocrystal surface, and the strong anchoring can improve the injection rate of photogenerated electrons from the dye molecules into the conduction band of TiO 2, 34, 51 which can enhance J sc. Table 4.2. Cell parameters of DSSCs fabricated using TMA and P25 samples. Sample Thickness J sc V oc ff (μm) (ma/cm 2 ) (V) TMA P