Unfolding the Anatase-to-Rutile Phase Transition in TiO2 Nanotubes

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article 2

Unfolding the Anatase-to-Rutile Phase Transition in TiO Nanotube Using X-Ray Spectroscopy and Spectromicroscopy Jun Li, Zhiqiang Wang, Jian Wang, and Tsun-Kong Sham J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b07613 • Publication Date (Web): 31 Aug 2016 Downloaded from http://pubs.acs.org on September 2, 2016

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.

The Journal of Physical Chemistry C 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 30

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

The Journal of Physical Chemistry

Unfolding the Anatase-to-Rutile Phase Transition in TiO2 Nanotube Using X-Ray Spectroscopy and Spectromicroscopy Jun Li,† Zhiqiang Wang,† Jian Wang,‡ and Tsun-Kong Sham†,¶,* †

Department of Chemistry, University of Western Ontario, Chemistry Building, 1151 Richmond Street, London, Ontario, Canada N6A 5B7 ‡

Canadian Light Source Inc., University of Saskatchewan, Saskatoon, Canada S7N 2V3

¶

Soochow University-Western University Centre for Synchrotron Radiation Research, the University of Western Ontario, London, Canada N6A 5B7

1 ACS Paragon Plus Environment


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

Page 2 of 30

ABSTRACT

This work reports a study of the anatase-to-rutile phase transition (ART) in a highly ordered TiO2 nanotube (NT) specimen fabricated using an electrochemical process followed by thermal annealing at 750 ℃ (NT750). Two-dimensional X-ray absorption near-edge structure–X-ray excited optical luminescence spectroscopy reveals the hierarchically two-layered structure of NT750 by resolving the surface anatase luminescence and bulk rutile optical emission. Scanning transmission X-ray microscopy analysis of a sliced NT750 lamella spatially differentiates the top nanotubular anatase structure from the dense rutile bottom layer with a gradual ART interface layer. Based on these results together with the known behavior of size and anisotropy dependence of ART in TiO2 nanocrystal, we propose the “bottom-up” mechanism for ART in anodic TiO2 NT. This result is particularly relevant to the fundamental understanding of phase transition in nanostructures as well as the fabrication of desired TiO2 NT mixed phase composite with an excellent control of anatase/rutile phase ratio.


Page 3 of 30

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


INTRODUCTION In nature, TiO2 polymorphs include anatase, rutile and brookite, of which anatase and rutile are most common and they are often the subjects of investigation for technological applications due to their photoactivity among others.1-4 Between the two, anatase TiO2 has been widely accepted as the desired phase in photocatalysis due to its slower charge recombination kinetics,5 higher charge carrier mobility6 and many other merits over rutile,7-10 while the latter shows its superior activity in photocatalytic reduction reactions.8,

11

In fact, more often than not, anatase-rutile

mixed phase with a proper ratio can deliver the synergistic efficiency in photocatalysis; the notion of mixed phase has led to a commercialized product (Degussa P25) and drives further studies on mixed-phase TiO2 toward photo-performance optimization.12, 13 To engineer the desired anatase/rutile (mixed) phase for a specific application, the anatase-torutile solid phase transition of TiO2 has been extensively investigated due to the fact that irreversible anatase-to-rutile transition (ART) can be initiated by annealing at elevated temperature.2, 14, 15 To date, research on solid phase ART in TiO2 primarily focus on its size, exposed surfaces (anisotropy), and morphology dependence.2,

14-20

First of all, the general

consensus of the size effect on ART is the proportional relationship between the grain size and annealing temperature.21, 22 In line with the general view, the size dependency of ART in TiO2 shows the same trend.16, 19, 20 Thermodynamically, rutile phase is more stable than anatase in the bulk due to its lower Gibbs free energy.14, 23 Accordingly, during high temperature annealing, the smaller size of anatase TiO2 nanocrystal (NC) possesses higher surface energy and exposes more anatase grain interfaces for rutile nucleation, which would trigger the ART for bigger rutile crystal growth by the coalescence of neighboring anatase NCs.2 In addition, recent reports show that the ART of TiO2 NC exhibits an anisotropic behavior,15, 18 in which the exposure of the



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

Page 4 of 30

thermodynamically unstable {001} facets of anatase TiO2 NC would decrease the energy barrier for rutile nucleation, thus favoring the ART process. Particularly, Zhao et al.18 pointed out that the ART temperature can be reduced by ~100 ℃ when the {001} facet percentage in anatase TiO2 NC increases from 32 % to 63 %. Nevertheless, unlike the size and crystal orientation dependence, research on the morphology-dependent ART study of anatase TiO2 has only received sporadic attention. In particular, the mechanism of ART in the one-dimensional (1D) TiO2 nanostructures like nanotubes (NTs) remains elusive, although these 1D structures are usually regarded as the more preferable morphologies for the application of TiO2 due to their unique architecture and relevant properties.24,

25

As a typical case, highly ordered TiO2 NT

created by electrochemical anodization possesses a particularly high surface area by having the ordered porous/ tubular structure. Hence, great efforts have been made to reveal the NT growth mechanism in order to achieve suitable NT morphology for various platforms,26, 27 yet, only a few reports involving the ART of anodic TiO2 NT are available.28-30 In fact, literature results are inconclusive due to the lack of proper techniques which allow to reveal the spatial locations of co-existing anatase and rutile phases. Therefore, ART of anodic TiO2 NTs certainly needs further work with the use of more advanced techniques, especially those with high spatial resolution. In this work, we report the use of two-dimensional X-ray absorption near edge structure–X-ray excited optical luminescence (2D XANES-XEOL) spectroscopy in combination with scanning transmission X-ray microscopy (STXM) to investigate the ART behavior of highly ordered TiO2 NT, hence revealing the ART mechanism. 2D XANES-XEOL is an advanced spectroscopic technique; it tracks the optical response of a specimen upon site specific X-ray excitation and it can be used to study the interplay between structural and optical properties of luminescent nanomaterials.31,

32

Note that XANES arises from the excitation of core electron into the


Page 5 of 30

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


unoccupied states set up by the potential and is sensitive to the chemical surrounding of the absorbing atom, such as local symmetry, oxidation states and coordination; XEOL, on the other hand, records the de-excitation process by collecting the optical photons from various decay channels, where the X-ray absorbed at each photon energy is partially converted to optical photons via radiative decay involving energy transfer and electron–hole recombination after a cascade of a thermalization process of the energetic electrons and holes. Thus 2D XANESXEOL spectroscopy is useful for differentiating anatase and rutile TiO2 due to their very different characteristic luminescence,2,

33

i.e., the fact that anatase (defect) shows a green

emission band while rutile (defect) exhibits a near-infrared (NIR) luminescence. Besides, the deconvolution of 2D XANES-XEOL spectroscopy provides the luminescent site sensitive optical XANES (also known as photoluminescence yield: PLY), in which the shape of PLY spectrum associated with XEOL intensity variation can essentially reveal the corresponding luminescent site. Details of 2D XANES-XEOL spectroscopy are reported elsewhere.2, 31, 32 In addition, the combined use of the XANES and X-ray microscopy with a nanosize focused beam in STXM delivers a spectral resolution of 0.05 eV and a spatial resolution of 30 nm, respectively.34, 35 Recent works focusing on individual nanostructure characterization using STXM exemplify its versatility.34-37 Therefore, we here use 2D XANES-XEOL spectroscopy in combination with STXM analysis to resolve the respective luminescent sites and further the spatial locations of coexisting anatase and rutile components in a TiO2 NT specimen which holds an anatase-rutile mixed phase. Correspondingly, we propose a bottom-up ART mechanism for anodic TiO2 NTs. MATERIALS and METHODS Sample Preparation. Highly ordered TiO2 NTs were synthesized by a two-step electrochemical anodization process using a two-electrode cell. Specifically, a pure Ti foil



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

Page 6 of 30

(Goodfellow, 0.1 mm thick) with a size of ~1 cm × 2 cm was used as the anode after ultrasonically rinsed with de-ionized (DI) water and ethanol, the cathode was a platinum wire at a distance of ~2 cm from the anode. The electrolyte was composed by 0.3 wt % NH4F (Alfa Aesar, 98.0% min), 2 vol % DI water and ethylene glycol. First, Ti was anodized under 50 V (Hewlett-Packard 6209B DC power supply) at room temperature (~20 ℃) for 4 h. Then the firstly anodized Ti foil was ultrasonically rinsed in 1 M HCl to remove the top NT layer. Next, the refreshed Ti foil was cleaned by DI water and ethanol for several times and used as the anode for the following anodization. At the second step, the refreshed (first-layer-removed) Ti foil was anodized for 30 min under 50 V using the same but fresh electrolyte. Afterwards, the as-grown NTs (AGNT) on the Ti substrate was rinsed with DI water and then ethanol several times to remove the remaining electrolyte on the NT surface. The AGNT was cut into 6 pieces. One of which was kept as was (AGNT), the other five were annealed at 500 ℃, 700 ℃, 750 ℃, 800 ℃ and 900 ℃ for 2 h under ambient air to induce crystallization, henceforth denoted as NT500, NT700, NT750, NT800 and NT900, respectively. Characterization. Scanning electron microscopy (SEM, LEO 1540XB) and X-ray diffraction (XRD, Rigaku RU-200BVH, Co Kα radiation with λ = 1.7892 Å) were employed to track the morphology and crystal structure of NTs. Synchrotron experiments were carried out at the Canadian Light Source (CLS) located in Saskatoon, SK, Canada. XANES measurements together with the 2D XANES-XEOL maps were collected at the Spherical Grating Monochromator (SGM) beamline (E/∆E > 5000).38 During the run, the NTs attached to the Ti substrate were mounted on a sample holder with a normal angle of incidence toward the photon beam. For the Ti L3,2-edge and O K-edge XANES and XEOL, regions of excitation photon of 450-475 eV and 520-570 eV were scanned, respectively. Both the total electron yield (TEY) and


Page 7 of 30

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


partial fluorescence yield (PFY) detection modes were used. Of which the former was recorded by measuring the specimen current, the latter collected the element specific fluorescence X-rays using four silicon drift detectors with their locations set to ~45 deg relative to the incident beam and corresponding fluorescence energy windows for Ti L and O K. Meantime, XEOL signal for the 2D XANES-XEOL map was collected by an energy dispersive spectrometer (QE65000, Ocean Optics) with its geometry at ~30 deg with respect to the incident light. Specifically, when the photon energy was tuned across the edge of interest (e.g., Ti L3,2-edge or O K-edge), a diagram of excitation photon energy versus luminescence wavelength (200-960 nm) with the optical intensity color-coded was generated accordingly, i.e., a 2D XANES-XEOL map. Hereafter, the wavelength selected photoluminescence yield (PLY) XANES and the excitation energy specific XEOL spectrum were obtained by the deconvolution of the 2D XANES-XEOL map; the former was achieved by vertically cutting the 2D map with an optical window (wavelength selected) whereas the latter was acquired by horizontal cuts (typically at a given excitation energy). A detailed information of 2D XANES-XEOL spectroscopy has been described elsewhere.2 All XANES and XEOL spectra were normalized to the incident photon flux, and the associated geometries of XANES and XEOL measurements are indicated in Figure S1. STXM characterization of NT750 lamella was conducted at the soft X-ray spectromicroscopy (SM) beamline. The design and working principles of STXM have been reported elsewhere.39 Briefly, a 25 nm outermost-zone zone plate (CXRO, Berkeley Lab) with a 30 nm diffractionlimited spatial resolution was applied to focus the monochromatic X-ray beam to a 30 nm spot on the region of interest (ROI) of sliced NT750 lamella. Then the selected ROI was rasterscanned to generate the absorption image a pixel at a time via recording the incident and



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

Page 8 of 30

transmitted X-rays simultaneously. As the photon energy was tuned across the Ti L3,2-edge (450480 eV) and O K-edge (526-553 eV) energy ranges, image sequence scans at these two edges of interest were compiled accordingly. Details of the NT750 lamella preparation and STXM characterizations and analysis are included in the Supporting Information. RESULTS and DISCUSSION The SEM images of as-grown NTs (AGNT) and NTs annealed at 750 ℃ (NT750) are shown in Figure 1a~c and Figure 1d~f, respectively. Apparently, AGNT shows a homogeneous porous


Page 9 of 30

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


Figure 1. SEM (top, side and bottom) views of as-grown NTs (a~c) and NT750 (d~f) together with their XRD patterns (g). structure on the top (inner tube diameter is ~80 nm; tube wall thickness is ~30 nm) and ordered NTs underneath together with closed-end caps at the bottom. After annealing, NT750 suffers from sintering effect and exhibits the shrinkage of entire NT structure. Although it maintains the nanotubular architecture, the top surface of NT750 becomes more coarsened with rich cracks growing along the NT wall, and more interestingly, the bottom caps completely collapsed and transformed to nanoparticles (NPs) with a size of ~50 nm. X-ray diffraction (XRD) analysis in Figure 1g shows that AGNT is amorphous, in agreement with previous results,2, 40 whereas NT750 presents an anatase-rutile mixed phase with its dominant rutile character. In order to conduct a comprehensive study on the highly ordered TiO2 NTs toward the ART, we have employed the site-specific spectroscopy and spatially resolved spectro-microscopy on NT750. These results lead to the understanding of the ART in anodic TiO2 NTs. 2D XANES-XEOL maps of NT750 at the Ti L3,2-edge and O K-edge are shown in Figure 2a and 2b. The maps are constructed by displaying the optical luminescence wavelength (x-axis) as a function of excitation photon energy (y axis) with the XEOL intensity color-coded. The XEOL intensity difference between the Ti L3,2-edge and O K-edge is worth noting: at the Ti L3,2-edge, the 2p3/2,1/2 excitations are turned on whereas at the O K-edge, not only the O 1s excitation is turned on but the excitation cross-section of the Ti L3,2-edge is still significant. Let us take a look at these two 2D XANES-XEOL maps, we see two broad emission bands with the dominant green luminescence at excitation energy above the Ti L3,2-edge (~455 eV). The green emission centered at ~500 nm is attributed to the characteristic anatase defects emission whereas the luminescent band at ~825 nm in the near-infrared (NIR) region originates from the rutile



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

Page 10 of 30

defects,2 corroborating the anatase-rutile mixed phase structure of NT750. In addition, a closer observation of the XEOL intensity variation on both of these two maps shows different XEOL behaviors of the two emission bands as the photon energy changes across the edge; that is, whereas the anatase luminescence increases, the rutile emission decreases at the Ti L3,2-edge, albeit the variation at the O K-edge is less dramatic. It suggests that the two emission bands originate from different sites (e.g., surface or bulk).2

Figure 2. 2D XANES-XEOL maps of NT750 recorded with photon energies across (a) the Ti L3,2-edge and (b), the O K-edge with their color-coded emission intensity bars at the bottom. (c) Ti L3,2-edge and (d) O K-edge alignments among the TEY, PFY, PLY-V and PLY-NIR spectra of NT750 in comparison with the TEY XANES of anatase and rutile standards shown at the top panels, PLY-V and PLY-NIR are integrated with vertically cutting ranges of 400-650 nm and 750-900 nm, respectively, from both (a) and (b). Energy-selective XEOL spectra of NT750 across (e) the Ti L3,2-edge and (f) the O K-edge collected by horizontal cuts from (a) and (b),


Page 11 of 30

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


respectively, with their relevant excitation energies indicated as (+) in (c) and (d). All TEY, PFY, PLY and XEOL spectra are normalized to the incident photon flux. To deconvolute the 2D XANES-XEOL map, a series of vertical cuts are performed at first, which records the element and site specific photoluminescence yield (PLY), sometimes known as optical XANES with selected optical emission channel. It should be mentioned that for the convenience of discussion, PLY achieved by an integration of vertical cuts of the map across the visible region (400 nm ~ 650 nm) is denoted as PLY-V, and similar process performed at the NIR region (750 nm ~ 900 nm) is recorded as PLY-NIR. As shown in Figure 2c and 2d, the PLY-V and PLY-NIR spectra obtained by vertical cuts of the 2D XANES-XEOL maps at the Ti L3,2-edge and O K-edge are aligned with their corresponding total electron yield (TEY) and partial fluorescence yield (PFY) XANES spectra. The Ti L3,2-edge and O K-edge TEY XANES spectra of anatase and rutile standards are shown for comparison. It is worth noting that the differences among TEY, PFY and PLY are their different penetration depth and chemical environment sensitivity.2 First, the surface-sensitive TEY follows the universal curve and usually has an electron escape depth at several nanometers; PFY on nanoporous structure can penetrate much deeper to hundreds of nanometers and even microns due to the large attenuation length of X-rays (Figure S2); the extra thermalization process2 involved in PLY allows its further deeper detection depth on the basis of PFY. Second, both TEY and PFY are the average depiction of surface and bulk local environments of NT750. In contrast, PLY offers better sensitivity of the chemical environment because of the very different luminescence from anatase and rutile.2 Therefore, compared to TEY and PFY XANES, PLY XANES holds larger detection depth and higher chemical environment sensitivity.



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

Page 12 of 30

The interpretations of both the Ti L3,2-edge and O K-edge XANES of TiO2 are well documented.2,

40-43

On one hand, the fine structures at the Ti L3,2-edge (Figure 2c) can be

separated into the Ti L3-edge (458~463 eV) and the Ti L2-edge (463~470 eV). Whereas the former results from electronic transitions from the Ti 2p3/2 to 3d5/2,3/2 states, electronic transitions from the Ti 2p1/2 to 3d3/2 states are involved in the latter.2, 40 Moreover, a splitting within each edge into two peaks is induced by the crystal field splitting of the Ti 3d orbitals into the t2g and eg states located at the comparatively lower and higher energy regions, respectively. More importantly, the Ti L3-eg peak further splits into peak A and B due to the alignment between the Ti eg orbitals and the 2p orbitals of O ligands in the presence of the crystal field, contributing to the high sensitivity of the eg band to the local structure.41, 42 In the case of anatase and rutile TiO2, the much higher intensity of peak A over B indicates its anatase character with a D 2d local symmetry, whereas the rutile structure (D2h) delivers the inverse with a more intense peak B over A. On the other hand, O K-edge XANES originates from electronic transitions from O 1s to unoccupied O 2p states.2, 43 Specifically, for anatase and rutile TiO2 as shown in Figure 2d, peaks a and b result from the O 2p covalently hybridized with the Ti 3d-t2g and 3d-eg states, respectively; peaks c, d and e arise from covalent hybridization between O 2p with Ti 4sp states; peak f is usually treated as a sign of the long-range order of TiO2.43 The difference between anatase (D2d) and rutile (D2h) at the O K-edge is mainly from the energy range of 538~560 eV, where the O K-edge XANES of rutile has the extra feature d as well as the energy blue shifts of peaks c and f comparing to those of anatase. For the sample of interest, NT750, both its TEY and PFY XANES at both edges share the similar feature with the anatase TiO2 standard although the PFY spectra get broadened due to self-absorption.40 Meanwhile, the PLY-V and PLY-NIR spectra at both edges show the respective positive and negative XANES patterns with their


Page 13 of 30

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


broadened spectral features (i.e., loss of fine structures) due to the saturation effect involved in optically thick luminescent materials.2 As for the spectrum pattern of PLY XANES, an abrupt increase in absorption coefficient coupled with a drop in X-ray penetration depth occurs when the excitation energy goes above the edge jump (Figure S2). If the luminescence sites are in the bulk, a drop in penetration depth will reduce the thermalization path of the Auger electrons which can escape the surface without contributing to the energy transfer to the optical channel, leading to a drop in quantum efficiency and an inverted XANES in PLY. In contrast, if the luminescence sites are in the near surface region, the optical luminescence yield will increase with decreasing penetration depth, thus generating a positive PLY XANES.2 As shown in Figure 2c and 2d, the inverse PLY spectra at the NIR region, henceforth, suggests the rutile luminescent sites are more likely buried in the bulk while the anatase phase resides on the surface (by showing the positive PLY-V spectra), consistent with the anatase-like TEY and PFY XANES from the top region of NT750. To better understand the positive PLY-V spectrum from anatase and the negative PLY-NIR spectrum from rutile, a direct demonstration of the two XEOL bands at the two edges is illustrated in Figure 2e and 2f by horizontally cutting the relevant 2D XANES-XEOL maps (Figure 2a and 2b) with excitation energies indicated in Figure 2c and 2d (color-coded “+”). From the perspective of the Ti L3,2-edge, interestingly, the rutile emission is the superior component when the photon energy is below the edge, whereas it quickly switches to the dominated anatase luminescence with sharply reduced rutile emission once the photon energy goes above the edge. It further justifies the bulk instead of the surface rutile location, which accords well with the penetration depth behavior of X-rays (Figure S2). At photon energies below the Ti L-edge, incident X-rays can reach a depth with a magnitude of microns in porous



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

Page 14 of 30

NT750, since at these energies, only the valence and inner valence (Ti 3s, 3p) electrons are excited and their cross-section are small relative to the Ti 2p when the latter is turned on at the edge. Thus both anatase on the top and rutile in the bulk can be excited to generate the respective green and NIR luminescence (Figure S2 and S3a). Nevertheless, energy absorbed goes through a competing process with energy transfer to anatase and rutile optical decay channels. 2 Since the electron–hole recombination rate of rutile is reported to be much greater than that of anatase,7, 8 thus the generation of rutile emission is more efficient. On the other hand, as photon energies go above the Ti L-edge, the sharp absorption at the top region of NT750, due to the instant increase of the Ti 2p absorption cross section, reduces the X-ray penetration depth immediately to a few hundreds of nanometers (Figure S2 and S3b). It results in a dramatic drop of rutile emission although a small amount of X-rays still can reach the bulk rutile owing to the porous structure of NT750. The same reasoning can be applied to the XEOL behaviors of the two emission bands at the O K-edge (Figure 2f). Nevertheless, two differences are noted. First, the change of rutile luminescence from below to above the O K-edge is not as significant as the Ti L3,2-edge, which is mainly attributed to the slighter variation of X-ray absorption length in the former than in the latter (Figure S2). Second, the intensity ratio of green over NIR band above the O K-edge becomes smaller than that above the Ti L3,2-edge (Figure 2e). The reasons are due to the larger X-ray penetration depth with comparatively higher photon energy at the O K-edge, and the coexcitation of both Ti and O sites at the O K-edge.2 Therefore, the phase distribution of NT750 can be pictured as a two-layered architecture with stacked minor anatase phase on the top of a major rutile phase structure (Figure 1g). STXM is employed to disclose the spatial locations of these co-existing anatase and rutile components in NT750. Figure 3a shows the SEM view at the cross section of a NT750 lamella


Page 15 of 30

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


prepared by FIB milling and further thinning (Supporting Information). Note that the top region of the NT750 is facing up and covered by a layer of deposited platinum (darker band at the top) for the FIB performance. Clearly, the cross section view of NT750 depicts its two-layered structure with a layer of homogeneous NTs (with a length of ~3 μm) stacked upon a dense structure. In addition, the thickness of the lamella is controlled to be thinner than 100 nm, which is smaller than the attenuation length of X-rays (based on one absorption length of sample thickness, Supporting Information) with photon energies at the Ti L3,2-edge and O K-edge (Figure S2). Hence, thickness effect on the XANES is negligible.34 At first, STXM scans across the Ti L3,2-edge and O K-edge have been taken separately at the same region of interest (ROI) as indicated with a green rectangle in Figure 3a. By performing the chemical imaging using the Ti L3,2-edge and O K-edge XANES spectra of anatase and rutile standards as references (Supporting Information), the chemical maps of the selected ROI at the Ti L3,2-edge and O Kedge are shown in Figure 3b and 3c, respectively. Two maps manifest the persistent results, which evidently demonstrate that the pure anatase phase structure (red) resides at the top nanotubular region, whereas the bottom dense layer contains pure rutile structure (green). Notably, a slim yellow band due to color mixing locates between those two layers, indicating the possible anatase-rutile mixed phase at this interface region. To corroborate the speculation, the Ti L3,2-edge (Figure 3d) and O K-edge (Figure 3e) XANES spectra of those three regions (color-coded with the STXM maps) are extracted directly from their respective STXM optical density image stack (Supporting information). It is clear to see that clean anatase and rutile XANES profiles at both edges are attained from the homologous red and green ROIs (Figure 3b and 3c). Interestingly, the yellowish spectra at both edges extracted from the interface region deliver the XANES patterns resembling to both anatase and rutile. Particularly, the intensity ratio



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

Page 16 of 30

value of peak A over B in Figure 3d sets between that of anatase (red) and rutile (green), which unambiguously illustrates the anatase-rutile mixed nature of this interface layer.2

Figure 3. (a) SEM views of the as-made NT750 lamella. STXM chemical map at the (b) Ti L3,2edge and (c) O K-edge. (d) Ti L3,2-edge XANES extracted from color-coded ROIs in (b). (e) O K-edge XANES extracted from color-coded ROIs in (c). The selected ROI for STXM analysis in (b) and (c) is indicated by a green rectangle in (a). The scale (white) bars in (b) and (c) are 1 μm. To further reveal the structure of the interface region, a finer STXM scan with high spatial resolution on a narrower ROI marked by a blue rectangle (Figure 3a) is performed at the Ti L3,2edge. Figure 4a shows the corresponding optical density (supporting information) image. To 16 ACS Paragon Plus Environment

Page 17 of 30

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


pinpoint the variation across the interface layer, we selected seven ROIs stacking as mini slabs from the dense (ROI-1) to the tubular (ROI-7) area. The corresponding XANES (color-coded) extracted from the respective ROI are displayed in Figure 4b. Figure 4c depicts the relevant A/B intensity ratio across the region and a closer observation of the XANES spectra at the Ti L3eg region (peaks A and B, inset of Figure 4c). Clearly, the XANES evolve from plain rutile (ROI-1) to anatase-rutile mixtures with different ratio (ROI-2 to ROI-6), and eventually to plain anatase (ROI-7) together with the linear trend of peak A/B intensity ratio, suggesting that the interface region involves a gradual phase transition from bottom rutile to top anatase. This finding holds the greatest significance. It reveals that the ATR phase transition in ordered NTs is triggered at the bottom layer, followed by rutile nucleation upward and continuing rutile growth by making contact with the upper NT layer, resulting in the breakdown of NTs and transformation from nanotubular anatase (~3.8 g cm-3) to a dense rutile (~4.2 g cm-3) layer at the bottom via structure reconstruction. The pure anatase phase on the top NT layer is further confirmed here from two perspectives. First, examinations on seven parallel ROIs (Figure S4) show their unmistakable anatase characteristics with the same peak A/B intensity ratio in their Ti L3,2-edge XANES. Second, another seven ROIs selected with vertically across the top NT layer (Figure S5) also exhibit the homogeneous anatase phase for the whole nanotubular region.



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

Page 18 of 30

Figure 4. (a) A STXM optical density image at 462 eV of ROI marked by the blue rectangle in Figure 3a, the optical density bar is included underneath. The scale bar is 300 nm. (b) Ti L3,2edge XANES extracted from color-coded ROIs indicated by mini slabs in (a) with a size of ~ 30 nm × 130 nm. (c) peak intensity ratio of A over B from ROI-1 to ROI-7 shown in (b) together with a magified view of the Ti L3-eg XANES region (inset). Thereby, we propose a “bottom-up” ART model of anodic TiO2 NT based on this work. As shown in Figure 5a, five different stages are used to illustrate how phase transition occurs during thermal annealing at elevated temperatures. From stage 1 to 2, anatase crystallization from asgrown amorphous NTs occurs. Morphologically, Figure 5b and 5c show the side and bottom SEM views of anatase NT500 (Figure S6, NT500: as-grown NTs annealed at 500 ℃). The sintering effect is noticeable, in which NT500 exposes a large amount of cracks along the NT wall compared to the crack-free surface of as-grown NTs (Figure 1b). The presence of those cracks results from diffusion of bulk defects to the NT surface during structural ordering (i.e., crystallization) and the strain-relief inside the NT structure during annealing. In contrast, the bottom view (Figure 5c) of NT500, rather than showing cracks, shows the formation of tens of 18 ACS Paragon Plus Environment

Page 19 of 30

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


small NPs with the size of ~20 nm on each NT cap. Thus anatase NCs formed in the bottom caps deliver a smaller grain size than that formed in the upper NT walls. This can be attributed to the slightly different chemical nature between the bottom caps and the upper tube walls. Typically, a fluorine rich layer is formed at the bottom of anodic TiO2 NTs due to the accumulation of F- ions (from electrolyte) at the metal-oxide (Ti-TiO2) interface (NT caps) for further NT growth.26, 44 Our previous study shows that F- ions in the as-grown NTs are strongly correlated with the titania lattice and replace some of the oxygen substitutionally to form both surface and bulk Ti-F bonds.40 In turn, these F impurities can exert a morphology-tailoring effect during anatase crystallization via annealing, which exposes the thermodynamically unstable {001} facet of anatase NC but prevents its stable {101} facet from forming.45, 46 In fact, F-containing reagents have been widely applied to engineer the {001} faceted anatase NCs.47,

48

A high F- ion

concentration usually facilitates the fromation of smaller anatase NC with a high percentage of {001} facets because of the strong interaction of F 2p electron with both Ti 3d and O 2p electrons, resulting in the stabilization of Ti and O atoms on the {001} facet.47, 48 Interestingly, Zhao et al.18 found that the {001} facet of anatase NC can easily accumulate defects from the bulk via annealing and thus it appears much more defective than the {101} facet, which is due to the high surface energy of {001} facet with its distorted Ti-O-Ti bond angle and a high percentage of 5-coordination Ti.48 Therefore, anatase NCs in the bottom caps presumably would have a smaller size with a greater percentage of defective {001} facets than NCs in the upper tube walls where the bulk defects diffuse to the tube wall surface as cracks bewteen NCs other than the detention of defects on NC surface. Consistent evidence is provided by Chen et al. 41 who suggested the formation of electronically defective (Ti3+ states) anatase phase at the bottom cap layer after anatase crystallization by showing a yellow color, ranther than the natural white



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

Page 20 of 30

color of anatase TiO2 presented on the top layer. Besides, Yang et al.30 claimed that the anatase {101} only appears on the tube wall of anodic TiO2 NTs.

Figure 5. (a) Schematic view of the “bottom-up” ART model. (b) Side and (c) bottom SEM views of NT500. (d) Side and (e) bottom SEM views of NT700. (f ~ g) Side SEM views of NT800 and NT900, respectively. From stage 2 to 3, the rutile nucleation and growth from pure anatase NTs take place from the bottom caps. This can be rationalized in the following. First, the size effect plays an important role. Note that the size-dependency in ART of TiO2 NPs has been well documented,14, 16, 17, 20 of which the ART temperature is proportional to the size of anatase NPs. Because the smaller size


Page 21 of 30

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


of NP coupled with its accompanying higher surface area and energy not only offers more interfaces for rutile nucleation, but also provides the thermodynamic driving force for ART. 2, 19 Consequently, the closely packed small NPs in the bottom caps would trigger the ART. Second, ART in anatase NC exhibits an anisotropic behavior. Zhu et al.15 recently conducted a comprehensive analysis on the rutile nucleation site of anatase NC at the beginning of ART using the density functional theory (DFT) calculation. They claimed that the exposed {001} facet on anatase NC would significantly reduce the ART energy barrier. Hence, it energetically favors the rutile nucleation. Also, the higher percentage of {001} facet can further lower the ART energy barrier. Zhao et al.18 echoed that the ART temperature can drop by ~100 ℃ once the {001} facet concentration increases from 32 % to 63 %. By contrast, rutile nucleation on thermodynamically stable {101} facet would be much more difficult. It is caused by the endothermic nature of ART process and a much greater energy barrier for the rutile nucleation from the {101} facet.15 In our case, as shown in Figure 5d and 5e, NT700 (as-grown NTs annealed at 700 ℃) with its major anatase and minor rutile phase (Figure S6) structure exhibits its heavily cracked tube walls as well as the dense nanoparticulate bottom. On one hand, the defective anatase NCs in the bottom caps proposed above with their comparatively higher percentage of {001} facets would trigger the rutile nucleation much easier than anatase NCs in the upper tube walls. Thus the formation of rutile nucleus and further rutile crystal growth proceed favorably at the capped bottom region by aggregation and coalescence (i.e., Ostwald Ripening).15 Meantime, the significant local compressive strain induced by structure reconstruction during ART leads to the severely external deformation of the bottom caps and finally results in the total collapse of the cap layer. On the other hand, the {101} anatase in the tube walls are energetically less favorable for rutile nucleation.15, 30 Furthermore, the geometry



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

constraints imposed by the tube wall would impede the ART process.49,

Page 22 of 30

50

Thus the tubular

morphology containing the anatase phase would retain at the upper tube wall region although cracks develop into larger ones due to further defect exposure and strain-relief at elevated temperature annealing. In addition, previous studies also show the involvement of Ti substrate in thermal oxidation during annealing, enhancing the rutile nucleation at the NT bottom.28, 29, 49 Transformation from stage 3 to 4 depicts further rutile growth from the bottom toward the top at the expense of the tubular anatase structure, where ART in the upper anatase NT layer takes place by making contact at the interface region with the bottom dense rutile crystals. At the interface, further rutile crystal growth occurs by either at the rutile-anatase interface or merging with another rutile crystals.29 Thus there will not be an abrupt interface. Evidence is provided in Figure 5f where NT800 (as-grown NTs annealed at 800 ℃), together with its predominant rutile and anatase-rutile mixed phase, manifests a two-layered structure with anatase NT bundles on the top while a dense rutile layer underneath (Figure S6: rutile dominant character of NT800 from bulk XRD; Figure S7: anatase-like TEY and PFY XANES collected at the top region of NT800). Finally, from stages 4 to 5, the full rutile growth prevails at the expense of total annihilation of top nanotubular anatase (Figure 5g). As expected, NT900 (as-grown NTs annealed at 900 ℃) yields the pure rutile XRD and XANES profiles at both the Ti L3,2-edge and O K-edge (Figure S6 and S7). It is worth mentioning that, beyond the bottom-up ART model given in Figure 5a, the rutile nucleation directly in the tube wall between stage 3 and stage 5 is also possible once the energy provided by elevated annealing temperature is high enough. Nevertheless, this possibility is ruled out in this study, because the portion of the as-formed rutile nucleus is too small to be competitive with the bottom-up mechanism.


Page 23 of 30

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


CONCLUSION This work reveals in details the anatase-to-rutile phase transition (ART) behavior of highly ordered TiO2 NTs using the luminescent site sensitive 2D XANES-XEOL spectroscopy in combination with spatially resolved STXM. The as-grown amorphous NTs system annealed at 750 ℃ with its anatase-rutile mixed phase structure is taken as the representative sample of interest for characterizations. In comparison to the surface sensitive anatase-like TEY XANES of NT750, the more bulk-sensitive PLY XANES from the 2D XANES-XEOL maps at both the Ti L3,2-edge and O K-edge, exhibit the respective positive and negative XANES profiles from the anatase and rutile luminescent channels. It clearly suggests NT750 holds a two-layered structure with its anatase phase (surface) stacked upon the rutile base (bulk). STXM examination on the sliced NT750 lamella further corroborates the double-layered heterostructure. Of which the nanotubular anatase structure resides on the top layer whereas a dense rutile layer forms underneath, and a thin interface layer with a gradual ART is located in-between. Therefore, based on the rationalization of size and the anisotropy dependence of the ART process for TiO 2 NC, a “bottom-up” ART model for the ordered TiO2 NT is proposed: rutile nucleation initiates from the NT bottom caps and rutile structure growth proceeds further from bottom to the top at the expense of the nanotubular anatase phase via structural reconstruction. In addition, we also want to note that, powerful advanced spectroscopy and microscopy, such as the 2D XANESXEOL and STXM used in this work using synchrotron radiation are becoming more readily available for the research community and they will be able to play an ever increasing role in tracking fundamental processes in materials research. ACKNOWLEDGMENT



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

Page 24 of 30

Research at the University of Western Ontario is supported by the Discovery grant of the Natural Science and Engineering Research Council of Canada (NSERC), the Canada Research Chair (CRC) Program, the Canada Foundation for Innovation (CFI), and the Interdisciplinary Initiative (IDI) grant of the University of Western Ontario (UWO). The work at the Canadian Light Source (CLS) is supported by CFI, NSERC, National Research council (NRC), Canadian Institute for Health Research (CIHR), and the University of Saskatchewan. We would like to thank Dr. T. Regier for technical support at the SGM beamline at CLS, and Mr. T. Simpson for the assistance of nanotube lamella preparation at UWO. J. L. acknowledges the receipt of support from the CLS Graduate Student Travel Support Program. The authors wish to gratefully acknowledge the Nanofabrication Facility at Western University for SEM imaging and nanotube lamella preparation. ASSOCIATED CONTENT Supporting Information. This includes the detailed STXM sample preparation, characterization and the associated data analysis, the setup for XANES and XEOL measurements (Figure S1), calculated attenuation length of X-rays for bulk anatase and rutile TiO2 (Figure S2), schematic views of X-ray penetration depth in porous NT750 (Figure S3), STXM examinations on the top nanotubular region (Figure S4 and S5), XRD analysis of various NT samples (Figure S6), and XANES analysis of NT800 and NT900 (Figure S7). This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION Corresponding Author *[email protected]; 519-661-2111 ext. 86341 24 ACS Paragon Plus Environment

Page 25 of 30

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


Notes The authors declare no competing financial interests.



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

Page 26 of 30

REFERENCES (1) Chen, X. B.; Liu, L.; Yu, P. Y.; Mao, S. S. Increasing Solar Absorption for Photocatalysis with Black Hydrogenated Titanium Dioxide Nanocrystals. Science 2011, 331, 746–750. (2) Li, J.; Liu, L. J.; Sham, T. K. 2D XANES-XEOL Spectroscopy Studies of MorphologyDependent Phase Transformation and Corresponding Luminescence from Hierarchical TiO2 Nanostructures. Chem. Mater. 2015, 27, 3021–3029. (3) Li, L. D.; Yan, J. Q.; Wang, T.; Zhao, Z. J.; Zhang, J.; Gong, J. L.; Guan, N. J. Sub-10 nm Rutile Titanium Dioxide Nanoparticles for Efficient Visible-Light-Driven Photocatalytic Hydrogen Production. Nat. Commun. 2015, 6, 5881. (4) Satoh, N.; Nakashima, T.; Kamikura, K.; Yamamoto, K. Quantum Size Effect in TiO2 Nanoparticles Prepared by Finely Controlled Metal Assembly on Dendrimer Templates. Nat. Nanotechnol. 2008, 3, 106–111. (5) Xu, M. C.; Gao, Y. K.; Moreno, E. M.; Kunst, M.; Muhler, M.; Wang, Y. M.; Idriss, H.; Woll, C. Photocatalytic Activity of Bulk TiO2 Anatase and Rutile Single Crystals Using Infrared Absorption Spectroscopy. Phys. Rev. Lett. 2011, 106, 138302. (6) Tang, H.; Prasad, K.; Sanjines, R.; Schmid, P. E.; Levy, F. Electrical and Optical-Properties of TiO2 Anatase Thin-Films. J. Appl. Phys. 1994, 75, 2042–2047. (7) Murdoch, M.; Waterhouse, G. I. N.; Nadeem, M. A.; Metson, J. B.; Keane, M. A.; Howe, R. F.; Llorca, J.; Idriss, H. The Effect of Gold Loading and Particle Size on Photocatalytic Hydrogen Production from Ethanol over Au/TiO2 Nanoparticles. Nat. Chem. 2011, 3, 489–492. (8) Kim, W.; Tachikawa, T.; Moon, G. H.; Majima, T.; Choi, W. Molecular-Level Understanding of the Photocatalytic Activity Difference between Anatase and Rutile Nanoparticles. Angew. Chem., Int. Ed. 2014, 53, 14036–14041. (9) Luttrell, T.; Halpegamage, S.; Tao, J. G.; Kramer, A.; Sutter, E.; Batzill, M. Why Is Anatase A Better Photocatalyst than Rutile? - Model Studies on Epitaxial TiO2 Films. Sci. Rep. 2014, 4, 4043. (10) Liu, L. J.; Zhao, H. L.; 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, 1817–1828. (11) Ryu, J.; Choi, W. Substrate-Specific Photocatalytic Activities of TiO2 and Multiactivity Test for Water Treatment Application. Environ. Sci. Technol. 2008, 42, 294–300. (12) Luo, Z.; Poyraz, A. S.; Kuo, C. H.; Miao, R.; Meng, Y. T.; Chen, S. Y.; Jiang, T.; Wenos, C.; Suib, S. L. Crystalline Mixed Phase (Anatase/Rutile) Mesoporous Titanium Dioxides for Visible Light Photocatalytic Activity. Chem. Mater. 2015, 27, 6–17. (13) Tsukamoto, D.; Shiraishi, Y.; Sugano, Y.; Ichikawa, S.; Tanaka, S.; Hirai, T. Gold Nanoparticles Located at the Interface of Anatase/Rutile TiO2 Particles as Active Plasmonic Photocatalysts for Aerobic Oxidation. J. Am. Chem. Soc. 2012, 134, 6309–6315. (14) Koparde, V. N.; Cummings, P. T. Phase Transformations during Sintering of Titania Nanoparticles. ACS Nano 2008, 2, 1620–1624. (15) Zhu, S. C.; Xie, S. H.; Liu, Z. P. Nature of Rutile Nuclei in Anatase-to-Rutile Phase Transition. J. Am. Chem. Soc. 2015, 137, 11532–11539. (16) Barnard, A. S.; Xu, H. F. An Environmentally Sensitive Phase Map of Titania Nanocrystals. ACS Nano 2008, 2, 2237–2242. (17) Sabyrov, K.; Burrows, N. D.; Penn, R. L. Size-Dependent Anatase to Rutile Phase Transformation and Particle Growth. Chem. Mater. 2013, 25, 1408–1415.


Page 27 of 30

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


(18) Zhao, Y. B.; Zhang, Y. F.; Liu, H. W.; Ji, H. W.; Ma, W. H.; Chen, C. C.; Zhu, H. Y.; Zhao, J. C. Control of Exposed Facet and Morphology of Anatase Crystals through TiOxFy Precursor Synthesis and Impact of the Facet on Crystal Phase Transition. Chem. Mater. 2014, 26, 1014–1018. (19) Zhang, J.; Xu, Q.; Li, M. J.; Feng, Z. C.; Li, C. UV Raman Spectroscopic Study on TiO 2. II. Effect of Nanoparticle Size on the Outer/Inner Phase Transformations. J. Phys. Chem. C 2009, 113, 1698–1704. (20) Satoh, N.; Nakashima, T.; Yamamoto, K. Metastability of Anatase: Size Dependent and Irreversible Anatase-Rutile Phase Transition in Atomic-Level Precise Titania. Sci. Rep. 2013, 3, 1959. (21) Dick, K.; Dhanasekaran, T.; Zhang, Z. Y.; Meisel, D. Size-Dependent Melting of SilicaEncapsulated Gold Nanoparticles. J. Am. Chem. Soc. 2002, 124, 2312–2317. (22) Goldstein, A. N.; Echer, C. M.; Alivisatos, A. P. Melting in Semiconductor Nanocrystals. Science 1992, 256, 1425–1427. (23) Zhang, H.; Banfield, J. F. Structural Characteristics and Mechanical and Thermodynamic Properties of Nanocrystalline TiO2. Chem. Rev. 2014, 114, 9613–9644. (24) Shin, K.; Il Seok, S.; Im, S. H.; Park, J. H. CdS or CdSe Decorated TiO2 Nanotube Arrays from Spray Pyrolysis Deposition: Use in Photoelectrochemical Cells. Chem. Commun. 2010, 46, 2385–2387. (25) Hensel, J.; Wang, G. M.; Li, Y.; Zhang, J. Z. Synergistic Effect of CdSe Quantum Dot Sensitization and Nitrogen Doping of TiO2 Nanostructures for Photoelectrochemical Solar Hydrogen Generation. Nano Lett. 2010, 10, 478–483. (26) Ghicov, A.; Schmuki, P. Self-Ordering Electrochemistry: A Review on Growth and Functionality of TiO2 Nanotubes and Other Self-Aligned MOx Structures. Chem. Commun. 2009, 2791–2808. (27) Regonini, D.; Bowen, C. R.; Jaroenworaluck, A.; Stevens, R. A Review of Growth Mechanism, Structure and Crystallinity of Anodized TiO2 Nanotubes. Mater. Sci. Eng., R 2013, 74, 377–406. (28) Jarosz, M.; Syrek, K.; Kapusta-Kolodziej, J.; Mech, J.; Malek, K.; Hnida, K.; Lojewski, T.; Jaskula, M.; Sulka, G. D. Heat Treatment Effect on Crystalline Structure and Photoelectrochemical Properties of Anodic TiO2 Nanotube Arrays Formed in Ethylene Glycol and Glycerol Based Electrolytes. J. Phys. Chem. C 2015, 119, 24182–24191. (29) Varghese, O. K.; Gong, D. W.; Paulose, M.; Grimes, C. A.; Dickey, E. C. Crystallization and High-Temperature Structural Stability of Titanium Oxide Nanotube Arrays. J. Mater. Res. 2003, 18, 156–165. (30) Yang, B.; Ng, C. K.; Fung, M. K.; Ling, C. C.; Djurisic, A. B.; Fung, S. Annealing Study of Titanium Oxide Nanotube Arrays. Mater. Chem. Phys. 2011, 130, 1227–1231. (31) Wang, Z. Q.; Guo, X. X.; Sham, T. K. 2D XANES-XEOL Mapping: Observation of Enhanced Band Gap Emission from ZnO Nanowire Arrays. Nanoscale 2014, 6, 6531–6536. (32) Ward, M. J.; Han, W. Q.; Sham, T. K. 2D XAFS-XEOL Mapping of Ga1-xZnxN1-xOx Nanostructured Solid Solutions. J. Phys. Chem. C 2011, 115, 20507–20514. (33) Murphy, M. W.; Zhou, X. T.; Ko, J. Y. P.; Zhou, J. G.; Heigl, F.; Sham, T. K. Optical Emission of Biaxial ZnO-ZnS Nanoribbon Heterostructures. J. Chem. Phys. 2009, 130, 084707. (34) Li, J.; Wang, Z. Q.; Zhao, A. K.; Wang, J.; Song, Y.; Sham, T. K. Nanoscale Clarification of the Electronic Structure and Optical Properties of TiO2 Nanowire with An Impurity Phase upon Sodium Intercalation. J. Phys. Chem. C 2015, 119, 17848–17856. 27 ACS Paragon Plus Environment


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

Page 28 of 30

(35) Wang, Z. Q.; Wang, W.; Sham, T. K.; Yang, S. G. Origin of Luminescence from ZnO/CdS Core/Shell Nanowire Arrays. Nanoscale 2014, 6, 9783–9790. (36) Zhou, J. G.; Wang, J.; Sun, C. L.; Maley, J. M.; Sammynaiken, R.; Sham, T. K.; Pong, W. F. Nano-Scale Chemical Imaging of A Single Sheet of Reduced Graphene Oxide. J. Mater. Chem. 2011, 21, 14622–14630. (37) Zhou, J. G.; Wang, J.; Fang, H. T.; Wu, C. X.; Cutler, J. N.; Sham, T. K. Nanoscale Chemical Imaging and Spectroscopy of Individual RuO2 Coated Carbon Nanotubes. Chem. Commun. 2010, 46, 2778–2780. (38) Regier, T.; Paulsen, J.; Wright, G.; Coulthard, I.; Tan, K.; Sham, T. K.; Blyth, R. I. R. Commissioning of the Spherical Grating Monochromator Soft X-Ray Spectroscopy Beamline at the Canadian Light Source. AIP Conf. Proc. 2007, 879, 473–476. (39) Kilcoyne, A. L. D.; Tyliszczak, T.; Steele, W. F.; Fakra, S.; Hitchcock, P.; Franck, K.; Anderson, E.; Harteneck, B.; Rightor, E. G.; Mitchell, G. E. et al. Interferometer-Controlled Scanning Transmission X-Ray Microscopes at the Advanced Light Source. J. Synchrotron Radiat. 2003, 10, 125–136. (40) Li, J.; Liu, C. H.; Ye, Y. F.; Zhu, J. F.; Wang, S. D.; Guo, J. H.; Sham, T. K. Tracking the Local Effect of Fluorine Self-Doping in Anodic TiO2 Nanotubes. J. Phys. Chem. C 2016, 120, 4623–4628. (41) Chen, C. L.; Dong, C. L.; Chen, C. H.; Wu, J. W.; Lu, Y. R.; Lin, C. J.; Liou, S. Y. H.; Tseng, C. M.; Kumar, K.; Wei, D. H. et al. Electronic Properties of Free-Standing TiO2 Nanotube Arrays Fabricated by Electrochemical Anodization. Phys. Chem. Chem. Phys. 2015, 17, 22064–22071. (42) Degroot, F. M. F.; Fuggle, J. C.; Thole, B. T.; Sawatzky, G. A. 2P X-Ray Absorption of 3D Transition-Metal Compounds – An Atomic Multiplet Description Including the CrystalField. Phys. Rev. B 1990, 42, 5459–5468. (43) Zhou, J. G.; Fang, H. T.; Maley, J. M.; Murphy, M. W.; Ko, J. Y. P.; Cutler, J. N.; Sammynaiken, R.; Sham, T. K.; Liu, M.; Li, F. Electronic Structure of TiO2 Nanotube Arrays from X-Ray Absorption Near Edge Structure Studies. J. Mater. Chem. 2009, 19, 6804–6809. (44) Albu, S. P.; Ghicov, A.; Aldabergenova, S.; Drechsel, P.; LeClere, D.; Thompson, G. E.; Macak, J. M.; Schmuki, P. Formation of Double-Walled TiO2 Nanotubes and Robust Anatase Membranes. Adv. Mater. 2008, 20, 4135–4139. (45) Alivov, Y.; Fan, Z. Y. A TiO2 Nanostructure Transformation: from Ordered Nanotubes to Nanoparticles. Nanotechnology 2009, 20, 405610. (46) Naduvath, J.; Bhargava, P.; Mallick, S. Mechanism of Formation of Faceted Titania Nanoparticles from Anodized Titania Nanotubes. J. Phys. Chem. C 2015, 119, 9574–9579. (47) Gordon, T. R.; Cargnello, M.; Paik, T.; Mangolini, F.; Weber, R. T.; Fornasiero, P.; Murray, C. B. Nonaqueous Synthesis of TiO2 Nanocrystals Using TiF4 to Engineer Morphology, Oxygen Vacancy Concentration, and Photocatalytic Activity. J. Am. Chem. Soc. 2012, 134, 6751–6761. (48) Yang, H. G.; Sun, C. H.; Qiao, S. Z.; Zou, J.; Liu, G.; Smith, S. C.; Cheng, H. M.; Lu, G. Q. Anatase TiO2 Single Crystals with A Large Percentage of Reactive Facets. Nature 2008, 453, 638–641. (49) Varghese, O. K.; Paulose, M.; Shankar, K.; Mor, G. K.; Grimes, C. A. Water-Photolysis Properties of Micron-Length Highly-Ordered Titania Nanotube-Arrays. J. Nanosci. Nanotechnol. 2005, 5, 1158–1165.


Page 29 of 30

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


(50) Kumar, K. P.; Keizer, K.; Burggraaf, A. J.; Okubo, T.; Nagamoto, H. Textural Evolution and Phase-Transformation in Titania Membranes: Part 2 – Supported Membranes. J. Mater. Chem. 1993, 3, 1151–1159.



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

Page 30 of 30

TOC Graphic

(8.5 cm × 4.27 cm)


Unfolding the Anatase-to-Rutile Phase Transition in TiO2 Nanotubes

Recommend Documents