Anatase (101)-like Structural Model Revealed for Metastable Rutile

Feb 23, 2017 - ABSTRACT: Titanium dioxide has been widely used as an efficient transition metal oxide photocatalyst. However, its photocatalytic activ...
0 downloads 0 Views 928KB Size
Subscriber access provided by University of Newcastle, Australia

Letter

Anatase (101)-like Structural Model Revealed for Metastable Rutile TiO2(011) Surface Meiling Xu, Sen Shao, Bo Gao, Jian Lv, Quan Li, Yanchao Wang, Hui Wang, Lijun Zhang, and Yanming Ma ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b16449 • Publication Date (Web): 23 Feb 2017 Downloaded from http://pubs.acs.org on February 25, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 13

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Anatase (101)-like Structural Model Revealed for Metastable Rutile TiO2(011) Surface Meiling Xu†,‡, Sen Shao†, Bo Gao†, Jian Lv#, Quan Li#,†, Yanchao Wang*,†, Hui Wang*,†, Lijun Zhang#,†, and Yanming Ma†, †

State Key Lab of Superhard Materials, Department of Physics, Jilin University, Changchun 130012, China

‡ #

Beijing Computational Science Research Center, Beijing 100084, China College of Materials Science and Engineering and Key Laboratory of Automobile Materials of MOE, Jilin

University, Changchun 130012, China International Center of Future Science, Jilin University, Changchun 130012, China

Corresponding Authors: * E-mail: [email protected] * E-mail: [email protected] *E-mail: [email protected]

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT: Titanium dioxide has been widely used as an efficient transition metal oxide photocatalyst. However, its photocatalytic activity is limited to the ultraviolet spectrum range due to the large bandgap beyond 3 eV. Efforts to reduce the bandgap to achieve a broader spectrum range of light absorption have been successfully attempted via the experimental synthesis of dopant-free metastable surface structures of rutile-type TiO2 (011) 2×1. This new surface phase possesses a reduced bandgap of ~2.1 eV, showing great potential for an excellent photocatalyst covering a wide range of visible light. There is a need to establish the atomistic structure of this metastable surface to understand the physical cause for the bandgap reduction and to improve the future design of photocatalysts. Here, we report computational investigations in an effort to unravel this surface structure via swarm structure-searching simulations. The established structure adopts the anatase (101)-like structure model, where the topmost two-fold O atoms form a quasi-hexagonal surface pattern and bond with the unsaturated five-fold and four-fold Ti atoms in the next layer. The predicted anatase (101)-like surface model can naturally explain the experimental observation of the STM images, the electronic bandgap, and the oxidation state of Ti4+. Dangling bonds on the anatase (101)-like surface are abundant making it a superior photocatalyst. First-principles molecular dynamics simulations have supported the high photocatalytic activity by showing that water and formic acid molecules dissociate spontaneously on the anatase (101)-like surface. KEYWORDS: titanium dioxide, surface reconstruction, anatase, reduced bandgap, dopant-free, CALYPSO

ACS Paragon Plus Environment

Page 2 of 13

Page 3 of 13

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

 INTRODUCTION Since the discovery of the Honda-Fujishima effect on n-type TiO2 electrodes under ultraviolet light in 19721, TiO2 has attracted a great deal of research interest to make use of its remarkable photocatalytic properties. TiO2 has been widely applied in the splitting and purifying of water, removal of organic pollutants, and air self-cleaning/disinfecting of coatings, as well as in photoelectrochemical cells2–8. However, the actual application of TiO2 has been limited to the ultraviolet spectrum range because of its large bandgap beyond 3 eV9. Narrowing the bandgap to achieve the adsorption of light in a broader spectrum range has been a persistent challenge. Many attempts, such as noble metal loading10, composite semiconducting11,12, metal ion doping13, nonmetal elemental (N, C, and S) doping14–16, and metal and nonmetal elemental codoping17–19, have been made to tune the bandgap from a variety of aspects to extend the spectral response to the solar spectrum. Among these works, doping efforts have been demonstrated to be the most efficient method20, where impurity states are introduced into the band gap, enabling the absorption of visible solar energy. However, the use of dopants would inevitably introduce charge traps that act as recombination centers for the photogenerated charge carriers. It is therefore highly desirable to establish an alternative strategy to reduce the bandgap without the introduction of any dopants. Surface atoms are usually not saturated and have many dangling bonds, truncated from the chemically stoichiometric bulk21. Dangling bonds can cause high surface activities to stabilize the reconstructed surface phases2,22, deviating largely from the bulk structures. These surface phases can introduce electronic states into the bandgap, causing significantly reduced bandgaps for long-tail optical absorption and

red-shifted optical emission2,21. For example, the reconstructed anatase (001) and (101) surfaces have reduced bandgaps of ~2.323 and 2.024 eV, respectively, showing visible light activity in the photooxidation of formic acid23, while the reconstructed rutile (011) surface possesses a band gap of ∼2.1 eV25. These experimental efforts demonstrated that the electronic and optical properties of TiO2 can be efficiently improved by self-structured modification via surface reconstruction. Due to the scientific and technical importance, reconstructed surfaces of TiO2 have been intensively investigated in recent decades26–36. The studies on rutile (011)-2×1 surface reconstruction via the examination of titanyl37, microfaceting missing-row38, and brookite (001)-like structural models39,40 have found that the brookite (001)-like structure is the energetically most stable one, whose bandgap, however, is similar to that of the bulk rutile phase. Recently, a metastable two-dimensional (2D) phase formed at the rutile (011) surface with a markedly reduced bandgap of ~ 2.1 eV has been synthesized25, showing great potential as an excellent photocatalyst covering a wide range of visible light. The 2D phase can be ascribed as a (2×1) surface unit cell, and the bright spots observed in the scanning tunneling microscope (STM) images are arranged in a (distorted) hexa gonal arrangement. The atomic structure of this 2D phase has attracted much attention to understand the physical cause for the bandgap reduction and to better design the optimal photocatalyst41,42. The titanyl model was first considered as the possible candidate for this reconstructed 2D surface structure41. However, its simulated STM image cannot account for the experimental observation. There is a recent report on the prediction of an interesting Ti4O4-2×1 surface structure model42 that is able to explain the

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

experimental STM data. This structure model breaks the ideal stoichiometry of TiO2 and gives rise to the charge state of Ti2+. Experimentally, the charge states of the Ti atoms in the metastable 2D (011) surface structure have been measured to be Ti4+ by the X-ray photoemission spectroscopy technique25. In this work, we performed extensive structure searching studies for the metastable rutile (011) surface reconstruction via the swarm-intelligence based CALYPSO code in combination with first-principles total energy calculations. In our structure searches, a novel anatase (101)-like structure is established, with which the experimental observations from STM images, the electronic bandgap, and the oxidation state of Ti4+ are well explained. We further demonstrated, by performing first principles molecular dynamics simulations, that the predicted anatase (101)-like surface has superior photocatalytic reactivity by efficiently dissociating water and formic acid molecules once adsorbed.  COMPUTATIONAL METHODS

We perform a systematic surface structure searching study on the rutile (011) surface of TiO2 via the swarm-intelligence based CALYPSO (Crystal structure AnaLYsis by Particle Swarm Optimization) method and its same-name code, which has been widely applied to predict stable and metastable structures of extended crystals, low-dimensional materials, reconstructed surfaces, and 43–47 nanoclusters . Our surface structure slab model typically consists of three regions: the bulk region and the unreconstructed and reconstructed surface regions (see Ref. 45 for details), where only the surface regions are structurally optimized. We used an asymmetric slab of 5 TiO2 layers and a vacuum layer of ~10 Å for performing large-scale simulations of

Page 4 of 13

reconstructed surface structures searches using CALYPSO. This choice relies on a series of tests on the number of TiO2 layers used and the height of the vacuum layer chosen, giving a good converged surface energies (~1.7 meV/atom). Equally important, this set of parameters is computationally feasible for large-scale simulations of CALYPSO structure searches to ensure exploration of large configuration space that is necessary for a good structure searching simulation. During the structure search, to remove the detrimental effects of the surface dangling bonds, 4/3H and 2/3 H are needed to passivate the Ti dangling bonds and the O dangling bonds, respectively48. For adatoms, four multiplications of the unit cell are considered (namely, TiO2, Ti2O4, Ti3O6, and Ti4O8). The 2,000 structures we searched are ranked according to the calculated enthalpy. Once we have the low-energy structures at hand, finer parameters are used for calculations of surface energies and electronic properties. To obtain accurate relative surface energies, a symmetric slab containing 11 TiO2 layers and the vacuum layer of 18 Å was used for structural relaxation without any constraints and for electronic structural calculations. The surface energy that determines the stability of a surface is defined as γ =

1 ( Etot − Eref − ∑ ni µi ) N

(1)

where Etot and Eref are the total energy of the surface under consideration and of the reference cleaved surface, respectively; ni and µi are the number of excess atoms of the cleaved surface and the chemical potential for each species, respectively; and N is equal to (m×n) for an (m×n) surface cell and serves as a normalization factor. The chemical potentials must satisfy the following constraints: (1) µTi ≤ µTibulk , (ii) µO ≤ 12 µO2 , and (3) µTi + 2 µO = ETiO , where ETiO2 is the internal 2

ACS Paragon Plus Environment

Page 5 of 13

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

energy of the bulk rutile phase. Eq. (1) can thus be written as 1  Etot − Eref − nTi ETiO2 − µO ( nO − 2nTi )  (2) N According to the constraints of the chemical potential, the range of µO is set as

γ=

1 2

µ O ≥ µO ≥ 2

1 2

(E

TiO2

− µTi ) .

The underlying structure relaxations, surface energy and electronic calculations are performed with the plane-wave pseudopotential method as implemented in the VASP code49. The Perdew-Burke-Ernzerhof generalized gradient approximation50 is chosen for the exchange-correlation functional, and a Hubbard U term (U = 4.2 eV) is considered to address the self-interaction error of the generalized gradient approximation51. The electron-ion interaction is described by the projector augmented-wave potentials52, with 3p63d34s1 and 2s22p4 configurations treated as valence electrons for Ti and O, respectively. A kinetic cutoff energy of 500 eV and Monkhorst-Pack k meshes with a grid spacing of 2π × 0.04 Å−1 are then adopted to give well converged total energies (~1meV/atom). The ionic positions are fully relaxed until the residual force acting on each ion is less than 0.02 eV/Å. In the electronic structure computation, the GGA + U (U = 4.2 eV) method51 is used to take the good account on the energy band gap. We also perform first principle Born-Oppenheimer molecular dynamics (MD) simulations as implemented in the CP2KQUICKSTEP program53,54. The electronic structures are calculated using density functional theory based on the mixed Gaussian plane wave approach. The local-density approximation to the exact exchange and correlation functional and 55 Goedecker-Teter-Hutter pseudopotentials are employed. The auxiliary plane wave basis cutoff for the electron density is set to 600 Ry. The MD simulations are carried out with a time

step of 0.5 fs, and the total simulation time of a single MD run is 20 ps. The temperature is controlled at 300 K with the Nosé-Hoover chain thermostat. The anatase (101)-like surface structure is modeled as a periodic slab with four TiO2 layers (~9.5 Å thick), where the vacuum separation between slabs is ~16 Å, and contains 32 TiO2 units in a 2×1 supercell with a surface area of 9.2×10.9 Å2. The adsorption energies are calculated according to the equation Eads = Eslab + Emol - Etotal, where Eslab, Emol, and Etotal represent the energies of the anatase (101)-like slab, the free molecule and the whole adsorption system, respectively.  RESULTS AND DISCUSSION Our structure searches independently identified the energetically stable brookite (001)-like structure39,40 and titanyl structure37 previously proposed. In addition, we predict a novel metastable structure unseen in other works, which is depicted in Figure 1. The structure is slightly corrugated,

Figure 1. The predicted anatase (101)-like reconstructed surface structure in the side view (a) and in the top view (b). (c) Simulated (left panel) and experimental (right panel) STM images. The Ti and O atoms on the surface and subsurface oxide layer are shown in the ball and frame modes, respectively. Atoms in the gray color represent Ti atoms. Atoms in red, yellow, and green colors represent O atoms in the topmost, middle, and downmost positions of the surface oxide layer, respectively.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

with a characteristic sawtooth profile along the [100] direction [see Figure 1 (a)]. In this structure, the topmost two-fold O atoms (O2c, in the red color) of the surface oxide layer form a quasi-hexagonal surface pattern, as depicted in Figure 1 (b) (solid line), and they are bonded to two types of unsaturated Ti atoms, having five-fold and four-fold [denoted Ti5c and Ti4c in Figures 1 (a) and (b), respectively] coordination. The middle and downmost O atoms of the surface oxide layer are two-fold (in the yellow color) and three-fold (O3c, in the green color), respectively. The surface oxide layer has a perfect stoichiometric form of TiO2 containing 4 Ti and 8 O atoms in a 2×1 supercell. The coordinates of the Ti and O atoms in the surface oxide layer are listed in Table 1. Table 1. Coordinates of the Ti and O atoms on the surface oxide layer of the anatase (001)-like model.

Atom

corrugated along the [100] direction than its unreconstructed (011) surface [Figure 2 (b)]. In the top view, the anatase (101)-like model is characterized by “rings” of 13 under-coordinated surface atoms, as depicted in Figure 1(b) (dashed line). The structure parameters of the anatase (101)-like model are calculated. The average interatomic O2c-Ti [O2c is the topmost O atoms in Figure 1 (a)] and O3c-Ti [O3c is the downmost O atoms in Figure 1 (a)] distances are 1.85 Å and 1.97 Å, respectively, which are the same as those of the ideal anatase (101) surface. However, the average interatomic distance between Ti and O, colored by yellow in Figure 1 (a) in the middle region of the surface oxide layer, is 1.84 Å, smaller than that (1.99 Å) of the anatase (101) surface.

Coordinates (x, y, z) (Å)

Anatase (101)-like Ti-1

7.124

4.986

33.586

Ti-2

0.552

0.790

33.027

Ti-3

5.145

2.192

33.202

Ti-4

2.470

3.519

33.240

O-1

5.604

0.525

33.880

O-2

3.804

2.883

34.360

O-3

2.158

5.357

33.353

O-4

3.715

2.947

31.914

O-5

0.923

2.519

33.553

O-6

6.657

3.184

33.578

O-7

8.010

0.219

32.028

O-8

8.659

0.067

34.431

We find a large similarity of the newly predicted surface structure in the surface oxide layer to the ideal anatase (101) [Fig. 2 (a)] from the side view, which is the lowest-energy surface of anatase56. We therefore name it as the anatase (101)-like model. In the atomic scale, the predicted anatase (101)-like model is less

Figure 2. The structures of the anatase (101) surface (a), the unconstructed rutile (011) surface (b), and the brookite (001)-like surface (c). The Ti and O atoms in the surface layer and in the sublayers are shown in the ball and frame modes, respectively. Atoms in gray and red colors represent Ti and O atoms, respectively.

ACS Paragon Plus Environment

Page 6 of 13

Page 7 of 13

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

The surface energies of the unreconstructed, titanyl37, Ti4O4-2×142, and brookite (001)-like structures as a function of the O chemical potential have been calculated, as shown in Figure 3 (a). The brookite (001)-like structure [see Figure 2 (c)] possesses the lowest surface energy, while the anatase (101)-like structure is metastable with a surface energy of 1.11 eV per 1×1 unit cell, higher than that of the brookite (001)-like structure. The physical origin for the metastability of the anatase (101)-like structure stems from the larger number of dangling bonds compared to that in the brookite (001)-like surface, arising from the unsaturated Ti atoms. It should be noted that the unsaturated Ti atoms on the surface of the brookite (001)-like structure are five-fold, while there are two types of unsaturated Ti atoms on the anatase (101)-like surface, five-fold Ti5c and four-fold Ti4c, shown in Figures 1 (a) and (b), respectively. In Figure 3 (a), it is seen that the anatase (101)-like structure is energetically more favorable than the unreconstructed surface by 0.23 eV, the earlier proposed titanyl model37 by 0.31 eV and more stable than the Ti4O4-2×1 model42 for 0 ≤ ∆µO ≤ -4.28eV, respectively. The STM images of the anatase (101)-like structure are simulated using the Tersoff method57 under positive sample biasing conditions with four energy bands and a fixed distance of 1 Å above the topmost O atoms. As shown in Figure 1(c), the calculated STM images show clear pseudo-hexagonal patterns, in excellent agreement with the experimental observations25. We further examined the calculated STM image by extending distances up to 3 Å, within which the pseudo-hexagonal STM pattern remains unaltered. The computed projected densities of states (PDOSs) of the predicted anatase (101)-like

surface model are shown in Figure 3 (b). It is seen that the valence and conduction states are derived mainly from the O 2p orbitals and the Ti 3d orbitals, respectively. No mid-gap states are seen within the band gap region since there are no dopants introduced. The resultant band gaps for the anatase (101)-like surface structure phase is about 2.6 eV using the GGA+U method.

Figure 3. (a) Computed surface energies of various structures as a function of the chemical potential of O. (b) Calculated projected density of states (PDOS) for the anatase (101)-like structure. As an example, the label “Ti-d” represents the PDOS for the d orbital of Ti atoms.

It is noted that the band gap of the experimentally metastable TiO2 (011)-(2×1) reconstruction is approximately ~2.1 eV, which agrees with our theoretical results. This supports our suggestion that the anatase (101)-like surface structure might be the best candidate

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

structure for the metastable TiO2(011)-2×1 reconstructed surface.

rutile

The photoreactivity of a crystal facet is intimately related to both its surface atomic structure and surface electronic band structure58. For the anatase (101)-like model, on the one hand, the number of dangling bonds is large; on the other hand, the upper edge of the valence band is actually formed by the states from the surface O atoms. Therefore, the anatase (101)-like model might have excellent photocatalytic properties

Figure 4. Structures of water (a) and (b) formic acid molecules on the anatase (101)-like surface. The left and right panels are the structures before and after dissociation, respectively. The anatase (101)-like structure and adsorption molecules (water and formic acid) are shown in the frame and ball modes, respectively. Atoms in dark gray, light gray and red colors represent the C, H and O atoms of water and formic acid molecules, respectively.

for applications. Below, we perform MD simulations to examine the photocatalytic activity of the anatase (101)-like structure via the adsorption of typical water and formic acid molecules on the surface. Our MD simulations show that both water

and formic acid molecules dissociate spontaneously on the anatase (101)-like surface. The dissociation is closely related to the computed high adsorption energies of 1.61 eV and 2.06 eV for water and formic acid molecules, respectively. The resultant dissociation structures are illustrated in Figure 4. From the right panel of Figure 4 (a), we can see that the OH moiety (a hydroxyl group) of the dissociated H2O exhibits the bidentate adsorption configuration, in which the hydroxyl group binds to the surface through two O-Ti4c and O-Ti5c bonds with bond lengths of 2.03 and 1.95 Å, respectively, and H binds to the nearby topmost O2c atoms. This bidentate chelating configuration of water adsorption on the Ti atoms at the sides of each O2c has not been previously seen. After accepting the hydroxyl group, the O2c atom moves slightly downward. In the same manner, the formic acid molecule spontaneously dissociates into anionic formate (HCOO-) and cationic H+ on the anatase (101)-like surface, as shown in the right panel of Figure 4 (b). The formate bonds to the surface Ti5c and Ti4c atoms in a bidentate adsorption configuration with bond lengths of 1.93 Å and 2.10 Å, respectively, and the dissociated H+ is adsorbed on the adjacent O2c atoms with a bond length of 0.93 Å. This adsorption behavior is commonly formed when a formic acid molecule dissociates at the surface of TiO22,59 Unbiased swarm structure searches and density functional total energy calculations are performed to explore the atomistic structures of the experimentally reconstructed metastable rutile (011) surface with a reduced bandgap of 2.1 eV. There is excellent mutual agreement between experiments and theory in terms of the STM image, the electronic bandgap, and the charge state of the Ti atoms, supporting the

ACS Paragon Plus Environment

Page 8 of 13

Page 9 of 13

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

anatase (101)-like model as the likely candidate structure of the experimental metastable surface. Our MD simulations certified that the predicted anatase (101)-like surface has a superior photocatalytic property by efficiently dissociating water and formic acid molecules.

 ACKNOWLEDGEMENTS The authors acknowledge the funding support from the National Natural Science Foundation of China (under Grants Nos. 11534003 and 11274136), the National Key Research and Development Program of China (under Grant No. 2016YFB0201200), the Science Challenge Project (JCKY2016212A501 ), and the 2012 Changjiang Scholar of Ministry of Education. Part of the calculations were performed in the high performance computing center of Jilin University and Tianhe2-JK in the Beijing Computational Science Research Center.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

 REFERENCES (1) (2) (3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

Fujishima, A. Electrochemical Photolysis of Water at a Semiconductor Electrode. Nature 1972, 238, 37-38. Diebold, U. The Surface Science of Titanium Dioxide. Surf. Sci. Rep. 2003, 48 (5-8), 53–229. O’Regan, B.; Grätzel, M. A Low-Cost, High-Efficiency Solar Cell Based on Dye-Sensitized Colloidal TiO2 Films. Nature 1991, 353 (6346), 737–740. Bard, A. J.; Fox, M. A. Artificial Photosynthesis: Solar Splitting of Water to Hydrogen and Oxygen. Acc. Chem. Res. 1995, 28 (3), 141–145. Liu, L.; Zhao, H.; Andino, J. M.; Li, Y. Photocatalytic CO2 Reduction with H2O on TiO2 Nanocrystals: Comparison of Anatase, Rutile, and Brookite Polymorphs and Exploration of Surface Chemistry. ACS Catal. 2012, 2 (8), 1817–1828. Yaghoubi, H.; Taghavinia, N.; Alamdari, E. K.; Volinsky, A. A. Nanomechanical Properties of TiO2 Granular Thin Films. ACS Appl. Mater. Interfaces 2010, 2 (9), 2629–2636. Ismail, A. a.; Bahnemann, D. W. Mesoporous Titania Photocatalysts: Preparation, Characterization and Reaction Mechanisms. J. Mater. Chem. 2011, 21 (32), 11686–11707. Fujishima, A.; Zhang, X.; Tryk, D. A. TiO2 Photocatalysis and Related Surface Phenomena. Surf. Sci. Rep. 2008, 63 (12), 515–582. Kavan, L.; Grätzel, M.; Gilbert, S. E.; Klemenz, C.; Scheel, H. J. Electrochemical and Photoelectrochemical Investigation of Single-Crystal Anatase. J. Am. Chem. Soc. 1996, 118 (28), 6716–6723. Anpo, M. The Design and Development of Highly Reactive Titanium Oxide Photocatalysts Operating under Visible Light Irradiation. J. Catal. 2003, 216 (1-2), 505–516. Chen, X.; Shen, S.; Guo, L.; Mao, S. S. Semiconductor-Based Photocatalytic Hydrogen Generation. Chem. Rev. 2010, 110 (11), 6503–6570. Kumar, S. G.; Devi, L. G. Review on Modified TiO2 Photocatalysis under UV/Visible Light: Selected Results and Related Mechanisms on Interfacial Charge Carrier Transfer Dynamics. J. Phys. Chem. A 2011, 115 (46), 13211–

(14) (15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

13241. Choi, W.; Termin, A.; Hoffmann, M. R. The Role of Metal Ion Dopants in Quantum-Sized TiO2: Correlation between Photoreactivity and Charge Carrier Recombination Dynamics. J. Phys. Chem. 1994, 98 (51), 13669– 13679. Asahi, R. Visible-Light Photocatalysis in Nitrogen-Doped Titanium Oxides. Science 2001, 293 (5528), 269–271. Wu, G.; Wang, J.; Thomas, D. F.; Chen, A. Synthesis of F-Doped Flower-like TiO2 Nanostructures with High Photoelectrochemical Activity. Langmuir 2008, 24 (7), 3503–3509. Cho, I. S.; Lee, C. H.; Feng, Y.; Logar, M.; Rao, P. M.; Cai, L.; Kim, D. R.; Sinclair, R.; Zheng, X. Codoping Titanium Dioxide Nanowires with Tungsten and Carbon for Enhanced Photoelectrochemical Performance. Nat. Commun. 2013, 4, 1723. Gai, Y.; Li, J.; Li, S. S.; Xia, J. B.; Wei, S. H. Design of Narrow-Gap TiO2: A Passivated Codoping Approach for Enhanced Photoelectrochemical Activity. Phys. Rev. Lett. 2009, 102 (3), 23–26. Zhang, J.; Pan, C.; Fang, P.; Wei, J.; Xiong, R. Mo + C Codoped TiO2 Using Thermal Oxidation for Enhancing Photocatalytic Activity. ACS Appl. Mater. Interfaces 2010, 2 (4), 1173–1176. Liu, Y.; Shu, W.; Chen, K.; Peng, Z.; Chen, W. Enhanced Photothermocatalytic Synergetic Activity toward Gaseous Benzene for Mo+C-Codoped Titanate Nanobelts. ACS Catal. 2012, 2 (12), 2557–2565. Khan, S. U. M. Efficient Photochemical Water Splitting by a Chemically Modified N-TiO2. Science 2002, 297 (5590), 2243–2245. Liu, L.; Chen, X. Titanium Dioxide Nanomaterials: Self-Structural Modifications. Chem. Rev. 2014, 114 (19), 9890–9918. Martra, G. Lewis Acid and Base Sites at the Surface of Microcrystalline Anatase. Appl. Catal. A Gen. 2000, 200 (1), 275– 285. Ariga, H.; Taniike, T.; Morikawa, H.; Tada, M.; Min, B. K.; Watanabe, K.; Matsumoto, Y.; Ikeda, S.; Saiki, K.; Iwasawa, Y. Surface-Mediated Visible-Light Photo-Oxidation on Pure TiO2 (001). J. Am. Chem. Soc. 2009, 131 (41), 14670–14672.

ACS Paragon Plus Environment

Page 10 of 13

Page 11 of 13

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

(33)

(34)

Dette, C.; Pérez-Osorio, M. a.; Kley, C. S.; Punke, P.; Patrick, C. E.; Jacobson, P.; Giustino, F.; Jung, S. J.; Kern, K. TiO2 Anatase with a Bandgap in the Visible Region. Nano Lett. 2014, 14 (11), 6533– 6538. Tao, J.; Luttrell, T.; Batzill, M. A Two-Dimensional Phase of TiO2 with a Reduced Bandgap. Nat. Chem. 2011, 3 (4), 296–300. Bennett, R. A.; Stone, P.; Price, N. J.; Bowker, M. Two (1 × 2) Reconstructions of TiO2(110): Surface Rearrangement and Reactivity Studied Using Elevated Temperature Scanning Tunneling Microscopy R. Phys. Rev. Lett. 1999, 82 (19), 3831–3834. Blanco-Rey, M.; Abad, J.; Rogero, C.; Mendez, J.; Lopez, M. F.; Martin-Gago, J. A.; de Andres, P. L. Structure of Rutile TiO2(110)-(1 × 2): Formation of Ti2O3 Quasi-1D Metallic Chains. Phys. Rev. Lett. 2006, 96 (5), 055502. Herman, G. S.; Sievers, M. R.; Gao, Y. Structure Determination of the Two-Domain (1 × 4) Anatase TiO2(001) Surface. Phys. Rev. Lett. 2000, 84 (15), 3354–3357. Lazzeri, M.; Selloni, A. Stress-Driven Reconstruction of an Oxide Surface: The Anatase TiO2(001)-(1 × 4) Surface. Phys. Rev. Lett. 2001, 87 (26), 266105. Li, M.; Hebenstreit, W.; Diebold, U. Morphology Change of Oxygen-restructured TiO2(110) Surfaces by UHV Annealing: Formation of a Low-temperature (1 × 2) Structure. Phys. Rev. B 2000, 61 (7), 4926–4933. Liang, Y.; Gan, S.; Chambers, S. A.; Altman, E. I. Surface Structure of Anatase TiO2(001) Reconstruction, Atomic Steps, and Domains. Phys. Rev. B 2001, 63 (23), 235402. Onishi, H.; Iwasawa, Y. Dynamic Visualization of a Metal-Oxide-Surface/Gas-Phase Reaction: Time-Resolved Observation by Scanning Tunneling Microscopy at 800 K. Phys. Rev. Lett. 1996, 76 (5), 791– 794. Park, K. T.; Pan, M. H.; Meunier, V.; Plummer, E. W. Surface Reconstructions of TiO2(110) Driven by Suboxides. Phys. Rev. Lett. 2006, 96 (22), 226105. Park, K. T.; Pan, M.; Meunier, V.; Plummer, E. W. Reoxidation of TiO2(110) via Ti Interstitials and Line Defects. Phys. Rev. B - Condens. Matter Mater. Phys.

(35)

(36)

(37)

(38) (39)

(40)

(41)

(42)

(43)

(44) (45)

2007, 75 (24), 245415. Wang, Q.; Oganov, A. R.; Zhu, Q.; Zhou, X.-F. New Reconstructions of the (110) Surface of Rutile TiO2 Predicted by an Evolutionary Method. Phys. Rev. Lett. 2014, 113 (26), 266101. Pieper, H. H.; Venkataramani, K.; Torbrugge, S.; Bahr, S.; Lauritsen, J. V; Besenbacher, F.; Kuhnle, A.; Reichling, M. Unravelling the Atomic Structure of Cross-Linked (1 × 2) TiO2(110). Phys. Chem. Chem. Phys. 2010, 12 (39), 12436–12441. Beck, T.; Klust, A.; Batzill, M.; Diebold, U.; Di Valentin, C.; Selloni, A. Surface Structure of TiO2(011)-(2 × 1). Phys. Rev. Lett. 2004, 93 (3), 036104. Kubo, T.; Orita, H.; Nozoye, H. Surface Structures of Rutile TiO2(011). J. Phys. Chem. B 2007, 2 (011), 1–5. Gong, X.-Q.; Khorshidi, N.; Stierle, A.; Vonk, V.; Ellinger, C.; Dosch, H.; Cheng, H.; Selloni, A.; He, Y.; Dulub, O.; Diebold, U. The 2 × 1 Reconstruction of the Rutile TiO2(011) Surface: A Combined Density Functional Theory, X-Ray Diffraction, and Scanning Tunneling Microscopy Study. Surf. Sci. 2009, 603 (1), 138–144. Torrelles, X.; Cabailh, G.; Lindsay, R.; Bikondoa, O.; Roy, J.; Zegenhagen, J.; Teobaldi, G.; Hofer, W. A.; Thornton, G. Geometric Structure of TiO2(011)(2×1). Phys. Rev. Lett. 2008, 101 (18), 185501. Wang, Q.; Oganov, A. R.; Feya, O. D.; Zhu, Q.; Ma, D. The Unexpectedly Rich Reconstructions of Rutile TiO2(011)-(2 × 1) Surface and the Driving Forces behind Their Formation: An Ab Initio Evolutionary Study. Phys. Chem. Chem. Phys. 2016, 18 (29), 19549–19556. Zhou, R.; Li, D.; Qu, B.; Sun, X.; Zhang, B.; Zeng, X. C. Rutile TiO2(011)-2 × 1 Reconstructed Surfaces with Optical Absorption over the Visible Light Spectrum. ACS Appl. Mater. Interfaces 2016, 8 (40), 27403–27410. Li, Q.; Liu, H.; Zhou, D.; Zheng, W.; Wu, Z.; Ma, Y. A Novel Low Compressible and Superhard Carbon Nitride: Body-Centered Tetragonal CN2. Phys. Chem. Chem. Phys. 2012, 14 (37), 13081–13087. Lv, J.; Wang, Y.; Zhu, L.; Ma, Y. Predicted Novel High-Pressure Phases of Lithium. Phys. Rev. Lett. 2011, 106 (1). Lu, S.; Wang, Y.; Liu, H.; Miao, M.-S.; Ma, Y. Self-Assembled Ultrathin

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(46)

(47)

(48)

(49)

(50)

(51)

(52)

(53)

(54) (55)

(56)

(57)

Nanotubes on Diamond (100) Surface. Nat. Commun. 2014, 5 (100), 3666. Gao, B.; Shao, X.; Lv, J.; Wang, Y.; Ma, Y. Structure Prediction of Atoms Adsorbed on Two-Dimensional Layer Materials: Method and Applications. J. Phys. Chem. C 2015, 119 (34), 20111– 20118. Li, Q.; Ma, Y.; Oganov, A. R.; Wang, H.; Wang, H.; Xu, Y.; Cui, T.; Mao, H.-K.; Zou, G. Superhard Monoclinic Polymorph of Carbon. Phys. Rev. Lett. 2009, 102 (17), 175506. Deng, H.-X.; Li, S.-S.; Li, J.; Wei, S.-H. Effect of Hydrogen Passivation on the Electronic Structure of Ionic Semiconductor Nanostructures. Phys. Rev. B 2012, 85 (19), 195328. Kresse, G.; Furthmüller, J. Efficient Iterative Schemes for Ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set. Phys. Rev. B 1996, 54 (16), 11169–11186. Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77 (18), 3865–3868. Dudarev, S. L.; Savrasov, S. Y.; Humphreys, C. J.; Sutton, A. P. Electron-Energy-Loss Spectra and the Structural Stability of Nickel Oxide: An LSDA+U Study. Phys. Rev. B 1998, 57 (3), 1505–1509. Kresse, G.; Joubert, D. From Ultrasoft Pseudopotentials to the Projector Augmented-Wave Method. Phys. Rev. B 1999, 59 (3), 1758–1775. Vandevondele, J.; Krack, M.; Mohamed, F.; Parrinello, M.; Chassaing, T.; Hutter, J. Quickstep: Fast and Accurate Density Functional Calculations Using a Mixed Gaussian and Plane Waves Approach. Comput. Phys. Commun. 2005, 167 (2), 103–128. Developers, CP2K. "Open Source Molecular Dynamics." Development Version, http://www. cp2k. org. Vandevondele, J.; Hutter, J. Gaussian Basis Sets for Accurate Calculations on Molecular Systems in Gas and Condensed Phases. J. Chem. Phys. 2007, 127 (11), 114105. Lazzeri, M.; Vittadini, A.; Selloni, A. Structure and Energetics of Stoichiometric TiO2 Anatase Surfaces. Phys. Rev. B 2001, 63 (15), 155409. Tersoff, J.; Hamann, D. R. Theory of the Scanning Tunneling Microscope. Phys.

Rev. B 1985, 31 (2), 805–813. (58) Pan, J.; Liu, G.; Lu, G. Q. M.; Cheng, H.-M. On the True Photoreactivity Order of {001}, {010}, and {101} Facets of Anatase TiO2 Crystals. Angew. Chemie Int. Ed. 2011, 50 (9), 2133–2137. (59) Gong, X.; Selloni, A.; Vittadini, A. Density Functional Theory Study of Formic Acid Adsorption on Anatase TiO2 (001): Geometries, Energetics, and Effects of Coverage, Hydration, and Reconstruction. J. Phys. Chem. B 2006, 110 (6), 2804–2811.

ACS Paragon Plus Environment

Page 12 of 13

Page 13 of 13

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Paragon Plus Environment