and Perovskite Interfaces - American Chemical Society

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Surfaces, Interfaces, and Applications 3-#

Creation and Ordering of Oxygen Vacancies at WO and Perovskite Interfaces Kelvin H L Zhang, Guoqiang Li, Steven R. Spurgeon, Le Wang, Pengfei Yan, Zhaoying Wang, Meng Gu, Tamas Varga, Mark E. Bowden, Zihua Zhu, Chongmin Wang, and Yingge Du ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b03278 • Publication Date (Web): 25 Apr 2018 Downloaded from http://pubs.acs.org on April 26, 2018

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 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 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.

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 22 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

Creation and Ordering of Oxygen Vacancies at WO3-δδ and Perovskite Interfaces Kelvin H.L. Zhang†,‡, Guoqiang Li†§, Steven R. Spurgeon†, Le Wang†, Pengfei Yan# ∥ , Zhaoying Wang#⊥, Meng Gu#&, Tamas Varga#, Mark E. Bowden#, Zihua Zhu#, Chongmin Wang#, Yingge Du†* †

Fundamental and Computational Sciences Directorate, Pacific Northwest National

Laboratory, Richland, WA, USA ‡

College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, P.R.

China §

Key Laboratory of Photovoltaic Materials of Henan Province, School of Physics &

Electronics, Henan University, Kaifeng 475004, P.R. China ∥

Institute of Microstructure and Properties of Advanced Materials, Beijing University of

Technolgoy, Beijing, 100124, P. R. China ⊥

Department of Chemistry, Beijing Key Laboratory of Microanalytical Methods and

Instrumentation, Tsinghua University, Beijing, 100084, P.R. China &

Department of Materials Science and Engineering, Southern Unviersity of Science and

Technology, Shenzhen, Guangdong 518055, P.R. China #

Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory,

Richland, WA, USA KEYWORDS: WO3, epitaxy, oxygen vacancy, defect ordering ABSTRACT: Changes in structure and composition resulting from oxygen deficiency can strongly impact the physical and chemical properties of transition metal oxides, which may lead to new functionalities for novel electronic devices. Oxygen vacancies (Vo) can be readily formed to accomodate the lattice mismatch during epitixial thin film growth. In this paper, the effects of substrate strain and oxidizing power on the creation and distribution of Vo in WO3−δ thin films are investigated in detail. An 18O2 isotope labeled time-of-flight secondary ion mass spectrometry study reveals that WO3-δ films grown on SrTiO3 substrates display a significantly larger oxygen vacancy gradient along the growth direction compared to those 1 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

grown on LaAlO3 substrates. This result is corroborated by scanning transmission electron microscopy imaging which reveals a large number of defects close to the interface to accommodate interfacial tensile strain, leading to the ordering of Vo and the formation of semi-aligned Magnéli phases. The strain is gradually released and tetragonal phase with much better crystallinity is observed at the film/vacuum interface. The changes in structure resulting from oxygen defect creation are shown to have a direct impact on the electronic and optical properties of the films.

2 ACS Paragon Plus Environment

Page 2 of 22

Page 3 of 22 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 Transition metal oxides (TMO) form a series of compounds with a wide range of physical and chemical properties, and are highly useful in technologies associated with heterogeneous catalysis, electronics, spintronics, and high temperature superconductivity.1-2 The fabrication of high-quality heterointerfaces with precise control over crystal structure and oxygen stoichiometry is a key challenge hindering our current understanding of the intrinsic properties of TMOs, as well as their broader utilization in next-generation devices. For example, the existence of oxygen vacancies in as-grown SrCrO3-δ and SrCoO3-δ films is found to significantly change the structure and physical properties from those of their stoichiometric counterparts.3-4 On the other hand, in many cases, the catalytic properties of TMOs depend on the presence of defects such as oxygen vacancies, dislocations, and grain boundaries, which locally perturb metal-oxygen bonds.5 Thus, the thin-film growth community has shown significant interest in tuning the oxygen stoichiometry and defect concentrations by adjusting controllable variables such as growth temperature, oxygen partial pressure, and substrateinduced strain.1, 6-8 Stoichiometric tungsten trioxide (WO3) is a semiconductor (~2.8-3.0 eV bandgap) which can be easily reduced under electron beam, UV light, or heat treatment.9-11 To better understand and control the oxygen vacancy concentration in WO3 is not only of interest for fundamental studies, but can have direct impact on several important technologies, including electrochromic devices, smart windows, photocatalysis, and Li-ion batteries.12-13 WO3 crystals generally consist of corner-sharing WO6 octahedra. The tilting and distortion of WO6 octahedra lead to several structures include monoclinic I (γ-WO3), monoclinic II (ε-WO3), triclinic (δ-WO3), orthorhombic (β-WO3), tetragonal (α-WO3), and cubic WO3.14 Epitaxial WO3 has been deposited on various lattice mismatched substrates,11, 15-18 but the effect of interfacial strain on film structure and properties has not been systematically studied. Different from the mechanism of strain relaxation in traditional semiconductors, it has been 3 ACS Paragon Plus Environment


Page 4 of 22

shown that other relaxation mechanisms, such as rotation/tilt of oxygen octahedral, creation of planar defects, and structural transition may also play roles for WO3 epitaxy. Specifically, sub-stoichiometric WO3 (WO3-δ) can crystallize in many different forms known as Magnéli phases depending on the extent of oxygen deficiency. These include W25O73 (WO2.92), W5O14 (WO2.8), W17O47 (WO2.77), and W18O49 (WO2.72),19-20 and are characterized by the presence of crystallographic shear (CS) planes to accommodate ordered oxygen vacancies through edge sharing among WO6 octahedra. The relatively simple crystal structure of WO3, its small lattice mismatch to commercially available perovskite substrates, and the existing knowledge base for the Magnéli phases make tungsten oxide an ideal model system for studying of oxygen vacancy defects. Our study entails varying the substrate and molecular beam epitaxy (MBE) deposition conditions for epitaxial WO3−δ films and following the effect(s) that these process variables have on the extent of reduction. Fully stoichiometric WO3 films are readily deposited on both LaAlO3(001) (LAO) and SrTiO3(001) (STO) substrates in an ambient of activated oxygen.11, 16

In contrast, when deposited in an ambient of 18O2 at a comparable oxygen partial pressure,

the films grown on STO are more oxygen deficient than those deposited on LAO. Time-offlight secondary ion mass spectrometry (ToF-SIMS) studies show no measurable increase in 18

O concentration in the film, indicating no further oxidation or oxygen exchange during

deposition of WO3-δ by congruent evaporation of WO3 powder in

18

O2. ToF-SIMS also

reveals a measurable O concentration gradient within the films, with the film/substrate interface being significant off-stoichiometry (reduced). In combination with high-resolution scanning transmission electron microscopy (STEM), we ascribe this result to the influence of strain. A large number of semi-ordered Magnéli phases are observed near the interface of WO3−δ and STO. As strain relaxation occurs at larger thicknesses, the oxygen vacancy concentration drops. The incorporation of oxygen defects is shown to lead to increases in 4 ACS Paragon Plus Environment

Page 5 of 22 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


conductivity and coloration of the resultant films. The ability to tune the oxygen vacancy concentration and distribution by interfacial strain and oxidation potential during deposition allows us to better control the electronic and electrochromic properties of WO3 thin films.

EXPERIMENTAL SECTION The WO3 thin films were grown in a custom MBE system which has been described elsewhere.4 High-purity WO3 powders (Sigma-Aldrich, >99.99%) were evaporated from a high-temperature effusion cell at a film growth rate of ~0.2 Å/s as calibrated by a quartz crystal microbalance (QCM). The (001)-oriented LAO and STO (10×10×0.5 mm, MTI Corporation) substrates were mounted on a 2” sample holder, and cleaned at 600 oC for 20 min in an oxygen partial pressure of 6.0 × 10-6 Torr prior to film growth. The substrates were heated to 500 oC and were rotated at 1 rpm during the deposition to ensure composition and thickness uniformity. An activated oxygen plasma beam (with O2 partial pressure in the chamber set at ~3×10-6 Torr) or an 18O isotope-labeled (Cambridge Isotope Laboratories, Inc., 97%) molecular oxygen (18O2) beam was incident on the sample during deposition to study the effect of the oxidizing agent on the sample stoichiometry. In situ reflection high-energy electron diffraction (RHEED) was used to monitor the overall morphology and surface structure. After deposition, the substrate temperature was reduced at a rate of 10 °C min-1 under the same oxygen environment. The thicknesses of the films range from 50 to 150 nm. In-situ x-ray photoelectron spectroscopy (XPS) using monochromatic Al Kα x-rays (hν = 1486.6 eV) was carried out at normal emission (electron take-off angle = 90o relative to the surface plane) with a VG/Scienta SES 3000 electron energy analyzer in an appended chamber. The total energy resolution was 0.5 eV. Spectroscopic ellipsometry measurements have been performed using a commercially available rotating analyzer instrument with compensator (V-VASE; J.A. Woollam Co., Inc.) within the spectral range from 0.4 to 4.6 eV. Data has been collected at three incidence angles (50°, 60° and 70°). The optical bandgap Eg 5 ACS Paragon Plus Environment


can be determined from the absorption coefficient α, calculated using Tauc plot following αE =A(E–Eg)n, where α is the absorption coefficient, E is the photon energy, A is a constant, and n is equal to 2 for our indirect-gap WO3 films, respectively. The electrical resistivity measurements were performed by Van der Pauw method using Ag contacts in the temperature range of 150-330 K.

The high-resolution X-ray diffraction (XRD) patterns were recorded using a Philips X'Pert Materials Research Diffractometer equipped with a fixed Cu anode operating at 45 kV and 40 mA, a hybrid monochromator consisting of four-bounce double crystal Ge(220), and a Cu X-ray mirror. ToF-SIMS measurement was performed using a TOF.SIMS5 spectrometer (IONTOF GmbH, Germany). The experimental details are described in a previous paper.21 In brief, a 2.0 keV Cs+ beam was used for sputtering with a sputter area of 300×300 µm2, while a 25 keV Bi+ beam was used for data collection, which was scanned on a 100×100 µm2 area at the center of the Cs+ sputter crater. TEM studies were carried out on an FEI Titan 80-300 transmission electron microscope under scanning mode at 300 kV. The TEM is equipped with probe corrector and high-angle annular-dark-field (HAADF) detector which enables subAngstrom resolution under STEM-HAADF imaging.22 The STEM-HAADF images were collected along the [010] axis by the annular detector in the range of 55-220 mrad with electron beam convergence angle as 17.8 mrad. Geometric phase analysis (GPA) was conducted using the FRWRTools plugin for Digital Micrograph.23 An adaptive Wiener filter was applied to the measured STEM-HAADF images and the (001)-type reflections were selected using a mask size of 2.5 nm with a 0.5 smoothing. All strains were calculated relative to the reference region in the substrate; some fluctuation in the strain profile is present at the substrate-film interface because of the abrupt change in contrast.

RESULTS AND DISCUSSION 6 ACS Paragon Plus Environment

Page 6 of 22

Page 7 of 22 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


In our earlier work, we demonstrated that WO3 films deposited in activated oxygen (O plasma) on both STO(001) and LAO(001) are epitaxial and fully stoichiometric, as revealed by in-situ RHEED and XPS.11, 16 The lattice mismatches for cubic WO3 (aWO3=3.73 Å) are 1.7% for LAO (aLAO=3.796 Å) and 4.5% for STO (aLAO=3.905 Å). As a result, a difference in surface morphology is observed at the initial nucleation stage for films grown on the two substrates, regardless of whether activated molecular oxygen is used. Figure 1 (a-f) displays RHEED patterns for two WO3 film thicknesses grown on LAO and STO in O2. The RHEED beam was blanked during the majority of deposition time and the patterns in the figure were taken with minimal beam exposure to avoid beam-induced sample reduction.11 The 6 nm film grown on STO is characterized by a sharp, spotty pattern as shown in Figure 1e, indicating island growth. The diffraction spots for the film on LAO are elongated and the RHEED pattern is between spotty and streaky, indicating a smoother surface in comparison with that grown on STO. This difference is most likey due to the smaller lattice mismatch between WO3 and LAO. The islands coalesce to form larger crystals as the film becomes thicker and the RHEED patterns for both 50 nm films become streaky due to less beam transmission and more reflection from the crystallite surfaces, as seen in Figure 1c and f. The faint half-order diffraction spots in WO3 films have also been observed by several other groups during MBE. It is most likely due to surface reconstruction at low oxygen partial pressure during growth, which has been shown by ex situ scanning tunnelling microscopy and low energy electron diffraction studies.24-27 XPS was performed on all samples to examine the oxidation states of W. The W 4f core-level photoelectron spectra for the films grown in O2 on LAO and STO are shown in Figure 1g and 1h, respectively. The spectrum for the film on LAO is composed of a wellresolved W 4f5/2 and 4f7/2 doublet, which are the same as those grown in activated oxygen on either LAO or STO, consistent with that of W6+ from stoichiometric phase.11 On the other hand, fitting the spectrum for the film on STO requires a second doublet at 1.2 eV lower 7 ACS Paragon Plus Environment


binding energies, revealing the presence of W5+ as a result of oxygen vacancies.11, 28-29 From the fitting, the W5+ peak area is ~2% of the total. As each oxygen vacancy donates two electrons, a formula of WO2.99 is extracted from these data. Moreover, the slight reduction also has an impact on the electronic and optical properties of the films. As seen from the XPS valence band (VB) spectrum for the film grown on STO (inset to Figure 1h), a pronounced feature is observed close to the Fermi level. This feature is attributed to the presence of W5+ with a 5d1 configuration resulting from electron donation from oxygen vacancies.11, 28-31 The creation of oxygen defects also affects the crystallographic properties of the films. Figure 2a shows XRD out-of-plane θ-2θ scans around the (200) substrate and film reflections for four films: WO3 on STO and LAO with and without the use of activated oxygen. The four samples are denoted as O_STO, O_LAO, O2_STO, and O2_LAO, respectively, in the following discussion. All samples exhibit a single Bragg peak near the substrate (002) reflection, confirming epitaxy. The O plasma treated WO3 films (O_STO and O_LAO) both show an asymmetric, broad peak, due to the competition among the three different lattice parameters of the developing monoclinic WO3 phase. As we reported earlier, WO3 films grown under O plasma exhibit a thickness dependent structural transition from a strained tetragonal phase to its relaxed, bulk monoclinic phase.16 For the films grown in O2 (O2_STO and O2_LAO), the increase in out-of-plane lattice parameter predominantly results from the “chemical expansivity” effect due to the introduction of oxygen vacancies.4, 32-33 The O2_LAO and O2_STO films also show very similar in-plane lattice parameters (3.87 Å), but their (113) in-plane peaks (shown in Figure 2b) reveal the differences in crystallinity. For the same film thickness, the Bragg peak shape for O2_LAO is much more defined and the full width at half maximum (FWHM) values,

1.3o for O2_LAO and 2.4o for O2_STO, suggests that the

crystallite size is roughly twice as large in the film on LAO. This difference clearly reveals that the O2_LAO film is of higher quality, which agrees with the RHEED observations. The slight reduction of W from 6+ to 5+ revealed in O2_STO films (Figure 1h) lowers the 8 ACS Paragon Plus Environment

Page 8 of 22

Page 9 of 22 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


resistivity of the WO3-δ films by two orders of magnitude compared to that of O2_LAO,16 and air anneal (550oC for 1 hour) cannot fully reoxidize the O2_STO films (Figure 2c). As seen in Figure 2b, the film grown on LAO still keeps yellowish color, while the film on STO is light blue. The yellowish colored film on LAO is similar to that of stoichiometric bulk WO3 due to photo-excited inter-band transition with an absorption edge at ~420 nm.34 To further study the optical properties, we performed ellipsometry measurement on the film grown on STO, as shown in Figure 2d. One extra absorption peak at 0.5 – 1.2 eV (marked by one blue arrow) is clearly identified in the bluish O2_STO film due to sample reduction, which can be annhilated by air annealing. According to recent hybrid DFT calculation by Wang et al.,35 this absorption peak can be ascribed to the oxygen vacancy induced states, which has been shown to reside in the band gap about 0.5-1.0 eV below the conduction band minimum at low defect concentrations. This is also consistent with the in-gap state observed at the bottom of conduction band by our VB measurement shown in Fig. 1h inset. Also in the tauc plot results (Figure 2d), no obvious change in the band gap (~2.9 eV) is detected for before and after air annealing films.

We note that although the XPS results point to complete oxidation for O2_LAO film, XRD data indicate that it is oxygen deficient compared to O_LAO film, based on lattice expansion in the c direction. This discrepancy suggests that either the oxygen deficiency level is below the XPS detection limit, or the sub-surface layers are more reduced than the nearsurface region (2 - 4 nm) probed by XPS. ToF-SIMS and STEM studies reveal that the latter is the correct explanation. Because the O2_STO and O2_LAO samples were grown in

18

O2,

ToF-SIMS was employed to investigate the 18O/16O exchange process and the overall oxygen stoichiometry. Figure 3a shows SIMS depth profile for

18

O/O, TiO- and WO3- for a 150 nm

thick O2_STO sample. The cross over for the TiO- and WO3- signals determine the position of the interface. The sharpness of the interface was determined by the drop in WO3- signal 9 ACS Paragon Plus Environment


Page 10 of 22

from 90% to 10% of that measured in the center of the film, which leads to ~4 nm, a relative low value for oxide interfaces by SIMS depth profiling.36 A significant rise in

18

O

concentration is observed as the TiO- (WO3-) signal starts to increase (drop), revealing that the STO surface region is 18O rich compared to the natural background level (0.205%). This sharp increase is due to 18O/16O exchange at the STO surface resulting from the healing the oxygen vacancies created during high-temperature annealing just before deposition.37 The

18

O

concentration is very close to natural background level in the WO3-δ film, as it is within the STO substrate, indicating that there is neither oxidation of the evolving WO3 film from the ambient gas, nor 16O/18O exchange. Figure 3b plots the normalized O to W ratio (O-/WO3-) from ToF-SIMS data for films on both substrates. The normalization was performed by setting the top surface to 3 (2.99) for O2_LAO (O2_STO) according to XPS results. Values lower than 3 indicate oxygen deficiency. While both films show oxygen deficiency towards the film/substrate interface as indicated by the red dotted lines, it is clear that the O2_STO film displays a smaller O/W ratio, indicating that it is more reduced. At the buried interface, the normalized stoichiometry is close to WO2.9 (WO2.8) for O2_LAO (O2_STO). Both substrates thus induce oxygen deficiency close to the interface but not at the surface, which is consistent with O vacancy creation being drive by tensile strain.16,

38

Based on the larger lattice mismatch with STO, it is expected that the

WO3/STO interface would be more oxygen deficient than the WO3/LAO interface, as observed. In order to better understand the impact of oxygen stoichiometry on film structure, the 150 nm O2_STO sample was examined by STEM. The O2_STO film has a flat surface as shown in Figure 4a. The high-resolution high-angle annular dark field (STEM-HAADF) image of the WO3-δ/vacuum interface shown in Figure 4b is very similar to that of stoichiometric WO3,16 demonstrating the excellent crystallinity of the film. The black amorphous region is a carbon coating applied after film growth to protect the film during 10 ACS Paragon Plus Environment

Page 11 of 22 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


STEM sample preparation. Near the film vacuum surface, no defects are observed in the STEM image. However, in the vicinity of the film/substrate interface, many kinds of defects can be clearly seen. A misfit dislocation is present in Figure 4c, ~8 nm from the interface. In addition, several W atom rows are displaced from their ideal pseudo-cubic positions and the local structures are greatly distorted. Similar defects are more common at the film/substrate interface, as seen in Figure 4d and 4e and marked by the arrow and square. In our previous study, interfacial defects were found to be caused by the tilting and deformation of WO6 octahedra for stoichiometric O_STO.16 For the O2_STO film, the corner sharing network of WO6 octahedra is interrupted by extensive O vacancy defects. In Figure 4d, each bright spot represents a column of W atoms aligned along the beam direction [010]. It is clear that many of the W columns are shifted away from their pseudo-cubic positions, indicating the ordering of oxygen defects. Such structures are commonly seen in more reduced Magnéli phases, such as W17O47 (WO2.76), and W18O49 (WO2.72) crystal structure,19-20 in which corner and edge sharing octahedra coexist. Motifs of W18O49 including W rings with pentagonal and hexagonal shapes viewed along [010] direction are identified in Figure 4c and 4d. Because many of the W columns shift from their ideal pseudo-cubic positions, it is expected that the W rows will blur along [001] when imaging along [100] direction. Indeed, because the cubic STO substrate is four-fold symmetric, we have observed other crystal domains where W columns appear as continuous lines instead of sharp spots, as shown in Figure 4e. A larger scale STEM-HAADF image shown in Figure 4f reveals that most of the defects (marked by arrows) are located within ~15 nm near the film/substrate interface. It is known that these highly defected structures display distinctly different electronic and chemical properties relative to those for stoichiometric WO3.19-20, 39 To better understand this phenomenon, we have conducted geometric phase analysis (GPA) to visualize local strain from our STEM-HAADF images down to the nanometer scale with approximately 0.1% strain resolution.40-41 Using this method we can visualize the strain fields around the defects in 11 ACS Paragon Plus Environment


Page 12 of 22

the films and explore the possibility of strain relaxation away from the substrate interface. Figure 5a and 5d show STEM-HAADF micrographs of the films used to calculate the strain maps for the O2_STO (top row) and O_STO (bottom row) films. Figure 5b and 5e show the in-plane strain component (εεxx) calculated relative to the substrate region marked by the rectangle. These maps clearly indicate the presence of multiple defects extending from the substrate (indicated by the arrows), which are associated with strain fields that extend several nanometers laterally. Figure 5c and 5f show maps of the out-of-plane strain component (εεyy), overlaid with the average profile integrated parallel to the interface. The films grown in the two oxidizing conditions exhibit very different behavior. The O2_STO film displays a highly compressively strained interface region of εyy ≈ -5.5% which gradually relaxes toward a smaller value (εεyy ≈ -3.5%) ~40 nm away from the interface. Accompanying this strain release is the transition from highly defected Magnéli structures at the interface region to pseudo cubic WO3 structure. In contrast, the O_STO film exhibits a nearly uniform strain of εyy ≈ 5.2% averaged over the measured region, which is a direct result of the planar defects formed from WO6 octahedral distortion in stoichiometric WO3 matrix.16 These results, in conjunction with our other observations, clearly show that local strain can vary greatly as a result of oxygen deficiency. While the trend of oxygen deficiency along the c direction in O2_STO sample is clearly identified from SIMS and STEM, we expect

a similar strain release

mechanism, albeit to a much less extent, in the O2_LAO sample due to a much smaller inplane lattice mismatch (1.7% vs. 4.5%). This result qualitatively agrees with the ToF-SIMS findings and explains why sample reduction is not detectable by XPS, which has a probe depth of ≤ ~5 nm. It should also be noted that the high degree of similarity in the c lattice parameter for O2_LAO and O2_STO means that this quantity is not highly sensitive to the extent of oxygen deficiency. As revealed by the SIMS and STEM studies, the major


Page 13 of 22 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


difference in oxygen deficiency between O2_LAO and O2_STO samples is manifested in the in-plane oxygen defects induced by interfacial strain.

CONCLUSION Using insight from several experimental methods, we present a detailed investigation into the effects of substrate strain and oxidizing power on the creation and distribution of oxygen vacancy defects in epitaxial WO3−δ. For WO3 deposited on STO in O2, the surface is slightly oxygen deficient, but the interface is substantially reduced, resulting in ordering of oxygen vacancies and the formation of semi-aligned Magnéli phases. It is shown that these highly defective structures exhibit lower resistivity and coloration in the films. Thus, by controlling the growth parameters, it is possible to tune the physical properties of the films. In contrast, WO3 deposited on LAO in O2 is substantially less reduced. From XRD, films grown on both substrates in O2 exhibit a similar degree of expansion along the c axis, indicating that this is not a highly sensitive indicator of oxygen deficiency. ToF-SIMS and STEM studies, combined with XPS and XRD, yield a more complete description of the origin of oxygen vacancies and their spatial distributions, which are critical to better understanding and control over the properties of functional oxide materials. Deposition in activated oxygen results in complete oxidation of films grown on both substrates. We note that a comprehensive characterization suite is needed to fully understand the interplay between composition, structure and defect distribution. Relying on one or even two techniques may lead to incomplete or contradictory conclusions. We have shown that oxygen vacancies can be tuned by interfacial stain and deposition conditions, and they can directly impact the electronic and optical properties of the resultant WO3-δ films. Furthermore, the significant difference in strain state and nanoscale structure found in the as-grown films may exhibit different performance/response (i.e., electrochromism, thermochromism, and redox activity) when external stimuli are applied. 13 ACS Paragon Plus Environment


AUTHOR INFORMATION Corresponding Authors *

Email: [email protected] Notes The authors declare no competing financial interest.

ACKNOWLEDGMENTS This work was supported by the U.S. Department of Energy (DOE), Office of Science, Office of Basic Energy Sciences, Early Career Research program under award # 68278. K.Z. acknowledges the support by the Thousand Youth Talents Programof China. Y.D. also acknowledges the support by EMSL’s Intermural Research and Capability Development Program for the preliminary data. The authors would like to thank Dr. Scott A. Chambers and Tim Droubay for many insightful discussions. A portion of the work was performed at the W. R. Wiley Environmental Molecular Sciences Laboratory, a DOE User Facility sponsored by the Office of Biological and Environmental Research located at the Pacific Northwest National Laboratory (PNNL). PNNL is a multi-program national laboratory operated for DOE by Battelle under contract DE-AC05-76RL01830.

REFERENCES 1. Chambers, S. A., Epitaxial growth and properties of thin film oxides. Surface Science Reports 2000, 39 (5-6), 105-180. 2. Bibes, M.; Barthelemy, A., Oxide spintronics. Ieee Transactions on Electron Devices 2007, 54 (5), 1003-1023. 3. Jeen, H.; Choi, W. S.; Freeland, J. W.; Ohta, H.; Jung, C. U.; Lee, H. N., Topotactic Phase Transformation of the Brownmillerite SrCoO2.5 to the Perovskite SrCoO3-delta. Advanced Materials 2013, 25 (27), 3651-3656. 4. Zhang, K. H. L.; Sushko, P. V.; Colby, R.; Du, Y.; Bowden, M. E.; Chambers, A., Reversible nano-structuring of SrCrO3-delta through oxidation and reduction at low temperature. Nature Communications 2014, 5, 7. 5. Diebold, U., The surface science of titanium dioxide. Surface Science Reports 2003, 48 (5-8), 53-229. 6. Chambers, S. A., Ferromagnetism in doped thin-film oxide and nitride semiconductors and dielectrics. Surface Science Reports 2006, 61 (8), 345-381. 7. Martin, L. W.; Chu, Y. H.; Ramesh, R., Advances in the growth and characterization of magnetic, ferroelectric, and multiferroic oxide thin films. Materials Science & Engineering R-Reports 2010, 68 (4-6), III-133. 8. Wang, L.; Chang, L.; Yin, X.; You, L.; Zhao, J.-L.; Guo, H.; Jin, K.; Ibrahim, K.; Wang, J.; Rusydi, A., Self-powered sensitive and stable UV-visible photodetector based on GdNiO3/Nb-doped SrTiO3 heterojunctions. Applied Physics Letters 2017, 110 (4), 043504. 14 ACS Paragon Plus Environment

Page 14 of 22

Page 15 of 22 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


9. Rougier, A.; Portemer, F.; Quede, A.; El Marssi, M., Characterization of pulsed laser deposited WO3 thin films for electrochromic devices. Applied Surface Science 1999, 153 (1), 1-9. 10. Migas, D. B.; Shaposhnikov, V. L.; Rodin, V. N.; Borisenko, V. E., Tungsten oxides. I. Effects of oxygen vacancies and doping on electronic and optical properties of different phases of WO3. Journal of Applied Physics 2010, 108 (9), 093713. 11. Du, Y.; Zhang, K. H. L.; Varga, T.; Chambers, S. A., Reflection high-energy electron diffraction beam-induced structural and property changes on WO3 thin films. Applied Physics Letters 2014, 105 (5), 4. 12. Adler, S. B., Chemical expansivity of electrochemical ceramics. Journal of the American Ceramic Society 2001, 84 (9), 2117-2119. 13. Niklasson, G. A.; Granqvist, C. G., Electrochromics for smart windows: thin films of tungsten oxide and nickel oxide, and devices based on these. Journal of Materials Chemistry 2007, 17 (2), 127-156. 14. Zheng, H. D.; Ou, J. Z.; Strano, M. S.; Kaner, R. B.; Mitchell, A.; Kalantar-Zadeh, K., Nanostructured Tungsten Oxide - Properties, Synthesis, and Applications. Advanced Functional Materials 2011, 21 (12), 2175-2196. 15. Li, M.; Altman, E. I.; Posadas, A.; Ahn, C. H., The p(4 x 2) surface reconstruction on epitaxial WO3 thin films. Surface Science 2003, 542 (1-2), 22-32. 16. Du, Y. G.; Gu, M.; Varga, T.; Wang, C. M.; Bowden, M. E.; Chambers, S. A., Strain Accommodation by Facile WO6 Octahedral Distortion and Tilting during WO3 Heteroepitaxy on SrTiO3(001). Acs Applied Materials & Interfaces 2014, 6 (16), 1425314258. 17. Wang, M.; Shen, S.; Ni, J.; Lu, N.; Li, Z.; Li, H. B.; Yang, S.; Chen, T.; Guo, J.; Wang, Y., Electric‐Field‐Controlled Phase Transformation in WO3 Thin Films through Hydrogen Evolution. Advanced Materials 2017, 29 (46), 1703628. 18. Yang, J.-t.; Ma, C.; Ge, C.; Zhang, Q.-h.; Du, J.-y.; Li, J.-k.; Huang, H.-y.; He, M.; Wang, C.; Meng, S., Effects of line defects on the electronic and optical properties of strainengineered WO 3 thin films. Journal of Materials Chemistry C 2017, 5 (45), 11694-11699. 19. Li, Y. B.; Bando, Y.; Golberg, D., Quasi-aligned single-crystalline W18O49 nanotubes and nanowires. Advanced Materials 2003, 15 (15), 1294-+. 20. Migas, D. B.; Shaposhnikov, V. L.; Borisenko, V. E., Tungsten oxides. II. The metallic nature of Magneli phases. Journal of Applied Physics 2010, 108 (9), 093714. 21. Wang, Z.; Liu, B.; Zhao, E. W.; Jin, K.; Du, Y.; Neeway, J. J.; Ryan, J. V.; Hu, D.; Zhang, K. H.; Hong, M., Argon cluster sputtering source for ToF-SIMS depth profiling of insulating materials: high sputter rate and accurate interfacial information. Journal of The American Society for Mass Spectrometry 2015, 26 (8), 1283-1290. 22. Jesson, D.; Pennycook, S., Incoherent imaging of crystals using thermally scattered electrons. Proc. R. Soc. Lond. A 1995, 449 (1936), 273-293. 23. Koch, C. T., Determination of core structure periodicity and point defect density along dislocations. 2002. 24. Jones, F.; Rawlings, K.; Foord, J.; Egdell, R.; Pethica, J.; Wanklyn, B.; Parker, S.; Oliver, P., An STM study of surface structures on WO3 (001). Surface science 1996, 359 (13), 107-121. 25. Moulzolf, S.; LeGore, L.; Lad, R., Heteroepitaxial growth of tungsten oxide films on sapphire for chemical gas sensors. Thin solid films 2001, 400 (1-2), 56-63. 26. Tanner, R. E.; Altman, E. I., Effect of surface treatment on the γ-WO 3 (001) surface: a comprehensive study of oxidation and reduction by scanning tunneling microscopy and lowenergy electron diffraction. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 2001, 19 (4), 1502-1509. 15 ACS Paragon Plus Environment


27. Al Mohammad, A., Effect of substrate structures on epitaxial growth and electrical properties of WO3 thin films deposited on (1¯ 012) and (0001) α-Al2O3 surfaces. Vacuum 2009, 83 (11), 1326-1332. 28. Fleisch, T. H.; Mains, G. J., AN XPS STUDY OF THE UV REDUCTION AND PHOTOCHROMISM OF MOO3 AND WO3. Journal of Chemical Physics 1982, 76 (2), 780786. 29. Xie, F. Y.; Gong, L.; Liu, X.; Tao, Y. T.; Zhang, W. H.; Chen, S. H.; Meng, H.; Chen, J., XPS studies on surface reduction of tungsten oxide nanowire film by Ar+ bombardment. Journal of Electron Spectroscopy and Related Phenomena 2012, 185 (3-4), 112-118. 30. Alamri, S. N., Study of thermocolored WO(3) thin film under direct solar radiation. Smart Materials & Structures 2009, 18 (2), 6. 31. Wang, G. M.; Ling, Y. C.; Wang, H. Y.; Yang, X. Y.; Wang, C. C.; Zhang, J. Z.; Li, Y., Hydrogen-treated WO3 nanoflakes show enhanced photostability. Energy & Environmental Science 2012, 5 (3), 6180-6187. 32. Santato, C.; Odziemkowski, M.; Ulmann, M.; Augustynski, J., Crystallographically oriented Mesoporous WO3 films: Synthesis, characterization, and applications. Journal of the American Chemical Society 2001, 123 (43), 10639-10649. 33. Kim, Y. M.; He, J.; Biegalski, M. D.; Ambaye, H.; Lauter, V.; Christen, H. M.; Pantelides, S. T.; Pennycook, S. J.; Kalinin, S. V.; Borisevich, A. Y., Probing oxygen vacancy concentration and homogeneity in solid-oxide fuel-cell cathode materials on the subunit-cell level. Nature Materials 2012, 11 (10), 888-894. 34. Cong, S.; Geng, F.; Zhao, Z., Tungsten Oxide Materials for Optoelectronic Applications. Advanced Materials 2016, 28 (47), 10518-10528. 35. Wang, F.; Di Valentin, C.; Pacchioni, G., Semiconductor-to-metal transition in WO 3− x: Nature of the oxygen vacancy. Physical Review B 2011, 84 (7), 073103. 36. Pergolesi, D.; Fabbri, E.; Cook, S. N.; Roddatis, V.; Traversa, E.; Kilner, J. A., Tensile Lattice Distortion Does Not Affect Oxygen Transport in Yttria-Stabilized Zirconia–CeO2 Heterointerfaces. ACS Nano 2012, 6 (12), 10524-10534. 37. De Souza, R. A.; Martin, M., Using O-18/O-16 exchange to probe an equilibrium space-charge layer at the surface of a crystalline oxide: method and application. Physical Chemistry Chemical Physics 2008, 10 (17), 2356-2367. 38. Zhao, T.; Chen, F.; Lu, H. B.; Yang, G. Z.; Chen, Z. H., Thickness and oxygen pressure dependent structural characteristics of BaTiO3 thin films grown by laser molecular beam epitaxy. Journal of Applied Physics 2000, 87 (10), 7442-7447. 39. Dobson, M. M.; Tilley, R. J. D., A NEW PSEUDO-BINARY TUNGSTEN-OXIDE, W17O47. Acta Crystallographica Section B-Structural Science 1988, 44, 474-480. 40. Hÿtch, M.; Houdellier, F., Mapping stress and strain in nanostructures by highresolution transmission electron microscopy. Microelectronic engineering 2007, 84 (3), 460463. 41. Spurgeon, S. R.; Sloppy, J. D.; Kepaptsoglou, D. M.; Balachandran, P. V.; Nejati, S.; Karthik, J.; Damodaran, A. R.; Johnson, C. L.; Ambaye, H.; Goyette, R., Thicknessdependent crossover from charge-to strain-mediated magnetoelectric coupling in ferromagnetic/piezoelectric oxide heterostructures. ACS nano 2013, 8 (1), 894-903.


Page 16 of 22

Page 17 of 22 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


FIGURES AND FIGURE CAPTIONS

Figure 1. RHEED patterns for LaAlO3(001) and SrTiO3(001) before (a, d) and after 6 nm WO3(001) (b, e), and 50nm (c, f) WO3(001) film growth. The incident beam was oriented along [100] direction. Core-level W 4f XPS spectra for two WO3 films grown in O2 on LaAlO3(001) (g), and SrTiO3(001) (h). For the film grown on SrTiO3, a detectable density of states is observed near the Fermi level in the valence band spectrum, as seen in the inset of (h).



Figure 2. a) XRD θ-2θ scans for WO3 films grown on STO in O plasma, STO in O2, LAO in O plasma, and LAO in O2. b) (113) in-plane peaks for two WO3 films of the same thickness grown on LAO and STO(001). c) Temperature-dependent resistivity curves for different WO3 films grown in O2. AA denotes the air annealing. d) Tauc-plot results for WO3 films grown on STO substrates measured before and after air annealing.


Page 18 of 22

Page 19 of 22 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 3. (a) SIMS depth profiles for a 150 nm O2_STO film grown in 18O2 on STO. (b) Normalized O/W ratio for two 150 nm thick O2_LAO and O2_STO films during SIMS depth profiling.



Figure 4. (a) STEM-HAADF image of a 150 nm WO3 film grown on STO in O2. High resolution STEM-HAADF images of (b) the vacuum/film interface, (c) 8 nm away from, and at (d, e) the film/substrate interface. Panel (f) shows that the defect density is higher at the film/substrate interface regions.


Page 20 of 22

Page 21 of 22 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 5. Geometric phase analysis (GPA) of O2_STO and O_STO films (top and bottom rows, respectively). (a, d) Adaptive Wiener-filtered STEM-HAADF micrographs used to calculate the strain maps. (b, e) Maps of the in-plane strain component (εεxx), showing significant fluctuations around the defects running through the film, as indicated by the arrows. The rectangle marks the reference region used to calculate the strain maps. (c,f) Maps of the out-of-plane strain component (εεyy), overlaid with the average profile integrated parallel to the interface. The O2_STO film exhibits a broad, gradual relaxation of strain from εyy ≈ -5.5 to -3.5%, as indicated by the direction of the arrow, while the O_STO film appears to be uniformly strained (εεyy ≈ -5.2%).



TOC graphic


Page 22 of 22

and Perovskite Interfaces - American Chemical Society

Recommend Documents