Ultrathin amorphous titania on nanowires: Optimization of conformal

4 days ago - The best approach, low pressure O2 plasma treatment, results in significantly more uniform titania films and a conformal coating...
1 downloads 0 Views 2MB Size
Subscriber access provided by UNIV OF SOUTHERN QUEENSLAND

Communication

Ultrathin amorphous titania on nanowires: Optimization of conformal growth and elucidation of atomic-scale motifs Danhua Yan, Mehmet Topsakal, Sencer Selcuk, John L. Lyons, Wenrui Zhang, Qiyuan Wu, Iradwikanari Waluyo, Eli Stavitski, Klaus Attenkofer, Shinjae Yoo, Mark S. Hybertsen, Deyu Lu, Dario J. Stacchiola, and Mingzhao Liu Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b04888 • Publication Date (Web): 03 May 2019 Downloaded from http://pubs.acs.org on May 5, 2019

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 21 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

Nano Letters

Ultrathin amorphous titania on nanowires: Optimization of conformal growth and elucidation of atomic-scale motifs Danhua Yan,†,⊥ Mehmet Topsakal,†,⊥ Sencer Selcuk,†,⊥ John L. Lyons,‡ Wenrui Zhang,† Qiyuan Wu,¶ Iradwikanari Waluyo,§ Eli Stavitski,§ Klaus Attenkofer,§ Shinjae Yoo,k Mark S. Hybertsen,† Deyu Lu,∗,† Dario J. Stacchiola,† and Mingzhao Liu∗,† †Center for Functional Nanomaterials, Brookhaven National Laboratory, Upton, NY 11973, USA ‡Center for Computational Materials Science, Naval Research Laboratory, Washington, DC 20375, USA ¶Department of Materials Science and Chemical Engineering, Stony Brook University, Stony Brook, NY 11794, USA §National Synchrotron Light Source II, Brookhaven National Laboratory, Upton, NY 11973, USA kComputational Science Initiative, Brookhaven National Laboratory, Upton, NY 11973, USA ⊥Contributed equally to this work E-mail: [email protected]; [email protected] Phone: +1 (631)344-2089; +1 (631)344-2569

1

ACS Paragon Plus Environment

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

Abstract Due to its chemical stability, titania (TiO2 ) thin films increasingly have significant impact when applied as passivation layers. However, optimization of growth conditions, key to achieving essential film quality and effectiveness, is challenging in the few-nanometer thickness regime. Furthermore, the atomic-scale structure of the nominally amorphous titania coating layers, particularly when applied to nanostructured supports, is difficult to probe. In this letter the quality of titania layers grown on ZnO nanowires is optimized using specific strategies for processing of the nanowire cores prior to titania coating. The best approach, low pressure O2 plasma treatment, results in significantly more uniform titania films and a conformal coating. Characterization using X-ray absorption near edge structure (XANES) reveals the titania layer to be highly amorphous, with features in the Ti spectra significantly different from those observed for bulk TiO2 polymorphs. Analysis based on first-principles calculations suggests that the titania shell contains a substantial fraction of undercoordinated Ti4+ ions. The best match to the experimental XANES spectrum is achieved with a “glassy” TiO2 model that contains ∼ 50% of undercoordinated Ti4+ ions, in contrast to bulk crystalline TiO2 that only contains 6-coordinated Ti4+ ions in octahedral sites. KEYWORD: amorphous titania; zinc oxide; photocatalysis; XANES; density functional theory

2

ACS Paragon Plus Environment

Page 2 of 21

Page 3 of 21 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

Nano Letters

Thin films of metal oxides have been extensively studied for a wide array of applications including those in microelectronics, high Tc superconductivity, devices for energy conversion and storage, and catalysis. 1,2 For instance, development of high-κ dielectrics has been essential to progress in CMOS technology and surface passivation layers play a key role in photovoltaic devices. 3,4 To illustrate, ultrathin layers of Al2 O3 and TiO2 (titania) have been deposited over crystalline Si films in solar cells to remove charge traps and to surpress parasitic carrier recombination. 4–6 In dye-sensitized solar cells (DSSC) and perovskite solar cells, the wide band gap of such oxides has led to the routine use of ultrathin oxide films as blocking layers (BLs) to prevent the backflow of photoelectrons. 7,8 The chemical inertness of certain metal oxides make them particularly useful for protecting critical materials layers in devices against adverse surrounding environments. For example, photocorrosion has been a significant limitation in the development of efficient materials and devices for photoelectrocatalysis. 9,10 Target materials that efficiently absorb solar radiation are often unstable when used as a photoelectrode in contact with the electrolyte under conditions required for key reactions such as water splitting. 10,11 Recently, the strategy of incorporating titania coatings has proved fruitful. Key design criteria for photocatalytic systems were decoupled, performance was improved, and the use of narrow band gap semiconductors such as Si and GaAs was enabled. 11–16 For example, we have reported that an ultrathin titania shell (TiO2 , ∼ 2 nm) deposited on zinc oxide nanowire (ZnO NW) arrays protected the nanowires against photocorrosion and significantly enhanced the overall photocatalytic activity (Figure 1). 12 This core/shell photocatalyst took advantage of the high carrier mobilities, the high crystallinity of the ZnO core, and the chemical robustness of the ultrathin titania shell. The net effect was to protect the ZnO core from photocorrosion and to increase photocatalytic activity. Through operando time-resolved photoluminescence spectroscopy studies, it was determined that the titania reduced the overall surface recombination rate by ∼ 40%. 15 To date, atomic layer deposition (ALD) has been the most common technique used for

3

ACS Paragon Plus Environment

Nano Letters

fabricating such layers due to a low processing temperature and excellent reproducibility. 17 Due to the self-limiting chemisorption of monolayers or submonolayers of alternating precursors, ALD can produce thin films that are uniform and conformal over target substrates. 18 However, in the limit of ultrathin films, the details of the film nucleation process during the first few reaction cycles can dominate the film morphology with the consequence that island growth can be observed to varying degrees. 19 If not carefully addressed, this problem can limit the degree of conformity achieved for films of few nanometers thickness. Also, even under optimal growth conditions, ALD grown conformal films are typically amorphorous due to the low temperatures used. This significantly complicates understanding the atomic-scale motifs. In particular, there is a missing link to the electronic structure of the amorphous titania layers with implications for such important characteristics as the carrier transport mechanisms. Resolving atomic scale structure in disordered materials is always challenging. Constructive characterization is further complicated for coatings on nanostructured target

a

b

0.8 TiO2 ZnO

0.6

2.83 Å

[002]

Jph (mA cm-2)

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 21

[100]

10 nm

0.4

0.2 ZnO/TiO2 As-grown ZnO 0.0

200 nm 0.0

0.5

1.0

1.5

E (VRHE) Figure 1: (a) Scanning electron microscopy (SEM) image of as grown ZnO nanowires (ZnOag). (Inset) High resolution transmission electron microscopy (TEM) image of an as grown ZnO nanowire. (b) Polarization curves of ZnO and ZnO/TiO2 nanowire photoanodes under simulated AM1.5 illumination, in an aqueous electrolyte of 0.1 M KOH. The higher photocurrent for ZnO/TiO2 clearly demonstrates its improved photocatalytic activity.

4

ACS Paragon Plus Environment

Page 5 of 21 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

Nano Letters

material, such as nanoparticles or nanowires. Here we report the optimization of ALD grown, ultrathin (2 - 3 nm) titania films to achieve conformal layers on nanowires. Systematic changes to the surface processing techniques applied to the ZnO nanowire cores are investigated. We observe that island growth dominates the TiO2 film morphology for growth over untreated ZnO nanowires. However, plasma sputtering applied to the ZnO core results in subsequent growth of TiO2 films that are significantly more uniform and conformal. We then turn to the atomic-scale structure of the titania shell and show that it is distinctively different from the bulk forms of titania. In particular, X-ray absorption near edge structure (XANES) studies, combined with analysis based on first-principles calculations, show that there is a high fraction (∼ 50%) of under-coordinated Ti sites in the titania shells. This contrasts with the octahedral coordination shell around Ti centers found in all three crystalline polymorphs of titania, i.e., rutile, anatase, and brookite. The ZnO NW arrays studied in this work are synthesized using a low temperature, seedmediated hydrothermal method that is slightly modified from the one developed by Greene et al. 20,21 On average the nanowires have a diameter of 40 − 50 nm and a length of 500 nm. They are wurtzite single crystals (space group P 63 mc) growing along the c-axis ([002]), with clearly resolved atomic lattice planes and well-defined facets in the high resolution TEM image (Figure 1a, inset). A thin shell of titania (TiO2 ) is deposited on the nanowires by atomic layer deposition (ALD) at 250 ◦ C, using titanium isopropoxide (TTIP) vapor and water vapor as precursors. Typically, 100 ALD cycles are performed to achieve a nominal thickness of 2 nm. The conformal coating of the nanowires by titania is confirmed by transmission electron microscopy (TEM). The titania shell appears amorphous and is easily distinguishable from the crystalline ZnO core, suggesting that the core/shell interdiffusion is negligible (Figure 2, lower panel ). The titania coating deposited on as-grown ZnO nanowires (ZnO-agT) appears rather uneven. Overall, it is thin (1 − 2 nm) and amorphous, but also exhibits islands that are

5

ACS Paragon Plus Environment

Nano Letters

As grown ZnO nanowires

ZnO-ag

ZnO-P

Plasma sputtering

ALD TiOx

Thermal annealing

ZnO-AP

ALD TiOx

ZnO-PT

ZnO-A

ALD TiOx

Plasma sputtering

ALD TiOx

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 21

ZnO-agT

ZnO-AT

ZnO-APT

Figure 2: (Upper panel) Post-growth treaments, including annealing in O2 at 500 ◦ C (A) and sputtering in low pressure O2 (P), are applied to as grown ZnO NWs prior to titania ALD coating (T). Samples were prepared by different sequences of processing followed by growth and denoted accordingly. (Lower panel) TEM images for each corresponding ZnO/TiO2 heterostructure that resulted, with the TiO2 shells false colored in cyan. False color is based on the Z -contrast between ZnO and TiO2 . Scale bars = 5 nm. thicker and contain crystalline domains of 3 − 4 nm in diameter (Figure 2, bottom left and Figure S1(b, c)). The coating uniformity can be significantly improved if post-growth treatments are applied to the ZnO core prior to titania deposition. As summarized in Figure 2, the treatments include various combinations of thermal annealing in O2 at 500 ◦ C (A) and sputtering by low pressure (< 25 mTorr) O2 plasma (P), and the samples are denoted accordingly. For example, sample ZnO-APT stands for ZnO nanowires sequentially treated by thermal annealing (A) and plasma sputtering (P), and finally coated with titania (T). Comparing across the four samples shown in Figure 2, the titania shells over the O2 plasma treated NWs (ZnO-PT and ZnO-APT) are thicker and more uniform than the ones without plasma treatment (ZnO-agT and ZnO-AT). On average, the titania shell thickness is ∼ 3 nm for ZnO-PT and ZnO-APT, but only 1 − 2 nm for ZnO-agT and ZnO-AT. 6

ACS Paragon Plus Environment

Page 7 of 21

b 1

ZnO-agT

ZnO-PT

0

ZnO-AT

Intensity (a.u.)

Ti 2p1/2 Ti 2p3/2

0 1

c 3

ZnO-APT

ZnO-agT

ZnO-PT

Zn 2p1/2 Zn 2p3/2

0 3

ZnO-AT

Atomic Fractions (%)

a

Intensity (a.u.)

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

Nano Letters

ZnO-APT

0 465

460

455 465

460

Binding energy (eV)

455

30 Ti Zn 20

10

0 1040

1020

1040

1020

Binding energy (eV)

ZnO-ag -agT

-PT

-AT

-APT

Samples

Figure 3: Ti 2p (a) and Zn 2p (b) X-ray photoelectron spectra of selected ZnO and ZnO/TiO2 nanowire samples processed and grown as shown in Fig. 2. For each spectrum the experimental data is fitted to two Gaussian peaks that correspond to 2p1/2 and 2p3/2 states respectively. (c) Surface atomic fractions for Ti and Zn in each sample, normalized to the sum of Zn 2p3/2 , Ti 2p3/2 , and O 1s (not shown) peak areas (Wagner sensitivity factors applied). The differences in the titania shell characteristics are confirmed at the ensemble level by X-ray photoelectron spectroscopy (XPS) (Figure 3). All four samples show distinctive Ti 2p3/2 and Ti 2p1/2 peaks (Figure 3a), with the peak positions corresponding to titanium in the 4+ valence state. 22 On the other hand, the Zn 2p3/2 and Zn 2p1/2 peaks corresponding to Zn 2+ are observed across all four samples (Figure 3b). The XPS line widths (FWHM) vary little across the four samples, which are 1.7 eV (2p3/2 ) and 2.3 eV (2p1/2 ) for Ti, and 2.1 eV (2p3/2 ) and 2.2 eV (2p1/2 ) for Zn. The Zn signal for samples sputtered by O2 plasma (ZnO-PT and ZnO-APT) are significantly weaker than the rest (ZnO-agT and ZnO-AT), which corresponds to a decrease of surface zinc atomic fraction by over one order of magnitude (Figure 3c). Given that the probe depth of XPS is only few nanometers, this result indicates that samples ZnO-PT and ZnO-APT do have thicker and more conformal titania coatings, so that fewer zinc atoms are exposed. Previously, it was reported that the solar water splitting activity of ZnO-APT is significantly improved over ZnO-agT, although it was entirely attributed to the improvement to the ZnO core in the earlier work. 12 Given the new finding based on high resolution TEM and XPS, we may infer that the improvement is partially due to the higher quality, more uniform titania coating over the plasma sputtered ZnO NWs, which offers a chemically stable surface and reaction sites that would not be consumed during

7

ACS Paragon Plus Environment

Nano Letters 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

photocatalysis. It is intriguing that the quality and thickness of titania coatings may vary so much depending on the post-growth treaments received by the ZnO core, despite the otherwise identical conductions under which all the titania coatings are deposited. In the case of as grown ZnO nanowires (ZnO-ag), one may speculate that the surface is contaminated by the organic reactants or products from the hydrothermal growth, which could hinder the ALD nucleation and growth, leading to island formation and overall slow growth. However, the titania coating quality is not improved for sample ZnO-AT, which has the ZnO core annealed in O2 at 500 ◦ C, a condition that should have removed all organic contaminants. On the other hand, it is the plasma sputtering that significantly improves the wetting of the ZnO surface by titania during ALD growth. To understand this, we note that ZnO nanowires grown by the hydrothermal technique are terminated by the nonpolar [1010] facets that has a very low surface energy. As such, the ALD precursors, namely titanium(IV) isopropoxide and water, may not be chemisorbed strongly over the pristine ZnO [1010] facets, thus leading to delayed nucleation and growth. In our previous work, through XANES studies at the Zn K -edge, we have discovered that the low pressure O2 plasma is more of a sputtering process that introduces a high level of oxygen vacancies and an enhanced surface dipole on the ZnO nanowire surfaces, 23 in contrast to the oxidizing effect borne by O2 plasma at higher pressures (> 25 mTorr). 24 We suggest that the increased polar character of the surface increases the coverage of the precursor adsorbates and decreases the nucleation barrier for titania ALD growth. As such, nucleation can happen more uniformly with shorter delay, thus improving the coating thickness and quality for samples ZnO-PT and ZnO-APT. Comparing the X-ray diffraction (XRD) pattern before and after titania growth shows that the titania shell contributes to no identifiable peaks (Supporting Information Figure S1). This suggests that the titania shell is highly amorphous. XANES spectra at the Ti L2,3 -edge (Figure 4a) further confirm this conclusion. The spectrum of the titania shell (ZnO-APT) is clearly different from those of crystalline bulk TiO2 in either of the anatase or rutile crystal

8

ACS Paragon Plus Environment

Page 8 of 21

Page 9 of 21 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

Nano Letters

structure. In particular, focusing on the L3 edge near ∼ 461 eV, the spectrum for the titania shell shows a single, broad peak while spectra from the crystalline polym rphs show a clear doublet. This difference has been previously observed in Ti L-edge for amorphous titania prepared by other means. 25 Simulations of spectra for cluster models based on crystalline fragments reveal that the splitting observed in spectra from crystalline samples only appears when the cluster has more than 30 atoms, 26 which corresponds to a diameter over 1 nm. In otherwords, the splitting is a clear signature for at least intermediate range order in the sample. This analysis suggests that the titania shell is predominantly rather amorphous with internal crystalline domains having a size of 1 nm or smaller. This is consistent with our TEM measurements where some larger nanocrystalline regions are seen, but these represent a small fraction of the film. At the Ti-K edge, the XANES spectra from the titania shells are also clearly distinguishable from those of rutile and anatase (Figure 4b), both in the white line region (4980 to 5010 eV) and the pre-edge region (near 4970 eV). The XANES spectra from the titania shells in the four differently prepared samples are very similar (Figure 4c), although there are some small variations to discuss. First, the intensity ratio between the white line doublet (D1 ∼ 4987 eV and D2 ∼ 5000 eV) varies slightly between the four samples, following the order of ZnOPT (1.06) > ZnO-APT (1.05) > ZnO-agT ∼ ZnO-AT (1.04). According to Harada et al., feature D1 corresponds to a ground state and a final state that are both in a charge transfer − 1s1 3d1 4p1 L), i.e., one electron in Ti 3d orbital and one hole in configuration (CT, 1s2 3d1 L → ligand (L), while feature D2 does not involve CT (1s2 3d0 → − 1s1 3d0 4p1 ). 27 It is the core hole interaction that substantially lowers the energy for the CT final state (1s1 3d1 4p1 L). Given that more prominent CT corresponds to higher electron density and lower average oxidation state at the center atom (Ti), the observed trend suggests that the oxidation state of the titania shell is lowered for plasma sputtered ZnO cores. Since plasma sputtering introduces more surface oxygen vacancies to the ZnO surface and leaves the surface electron-rich, 23 the ZnO interfacial region will have excess electrons to donate to the titania shell in the

9

ACS Paragon Plus Environment

Nano Letters

b ZnO-APT Anatase Rutile

with offsets

Ti K-edge

Normalized χµ(E)

ZnO-APT Anatase Rutile

0.5

see (d)

Ti L2,3-edge 455

460

465

470

475

5000

4950

Energy (eV)

c

d Normalized χµ(E)

D1 D2

with offsets

ZnO-APT

ZnO-AT

5000

0.1 4970

0.5

ZnO-PT

ZnO-agT

4950

5100

Energy (eV)

Ti K-edge

4965

5050

ZnO-APT

Anatase

Rutile

4975

5050

with offsets

5100

4965

Pre-edge 4970

0.1

Normalized absorption (a.u.)

a

Normalized χµ(E)

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 21

4975

Energy (eV)

Energy (eV)

Figure 4: (a) The Ti L2,3 -edge and (b) K -edge X-ray absorption near edge structure (XANES) spectrum of ZnO-APT is compared to those of anatase and rutile. (c) Ti K edge XANES spectra of ZnO-agT, ZnO-AT, ZnO-PT and ZnO-APT. Inset, a zoom-in of the pre-edge region. (d) The Ti K -edge XANES spectrum of ZnO-APT is compared to those of anatase and rutile in the pre-edge region. Offsets are applied along the y-axis if noted. ZnO-PT as compared to ZnO-agT samples, and similarly in the ZnO-APT as compared to the ZnO-AT samples, through interfacial electron spill-over. Second, in the pre-edge region

10

ACS Paragon Plus Environment

Page 11 of 21 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

Nano Letters

(Figure 4c, inset), a closer examination reveals small variations in the spectra. These may indicate less of an influence of interfacial effects in the samples with thicker, more conformal titania shells and better developed spectral signatures of the amorphous titania as such. Between the titania shell and crystalline TiO2 , the difference in Ti-K edge spectra is particularly pronounced in the pre-edge region (4963 − 4977 eV, Figure 4d). Specifically, the ZnO-APT spectrum is dominated by a single, asymmetric feature that has stronger intensity than the maxima in either rutile or anatase. Also, there is a ∼ 1 eV redshift of peak position relative to that of the strongest pre-edge peak in either rutile or anatase. Similar pre-edge feature was previously observed in TiO2 –SiO2 glasses 28 and heat treated Xerogels. 29 The transitions in this energy region are formally dipole-forbidden due to inversion symmetry in the case of octahedral Ti centers. In practice, quadrapole transitions and coupling with near-neighbor Ti sites play a role, 30,31 as do static structural distortions of Ti octahedra and thermal vibrations, 32 resulting in the sequence of relatively weak absorption peaks in this region for the anatase and rutile crystalline samples. On the other hand, studies of other titanium oxides have shown more intense and red-shifted pre-edge peaks for crystal structures with tetrahedral and five-coordinated Ti centers that lack inversion symmetry. The result is a trend correlating red-shift and intensity increase with the decrease of the coordination number of the absorbing atom. 33 This broadly suggests that the observed XANES spectra from the titania shells in the present samples contain a high level of undercoordinated Ti atoms and Ti octahedra with a large degree of structural distortion. However, the lack of measured Ti-K edge spectra from confirmed sources of undercoordinated Ti atoms in titania limits further analysis on the basis of experimental spectra. To quantify the analysis of the spectra and the deductions concerning local structure motifs, two approaches based on first-principles theory are developed. In the first approach, a simulated annealing procedure is used, 34 together with Density Functional Theory based molecular dynamics implemented in the VASP suite, 35,36 to develop a model structure for amorphous TiO2 . The resulting structure consists of 108 atoms in total and represents a

11

ACS Paragon Plus Environment

Nano Letters

b Normalized χµ(E)

a Normalized χµ(E)

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 21

ZnO-APT Best fit Best fit, Ti6c only

ZnO-APT Amorphous Model

4960

4970

4980

4990

5000

4960

Energy (eV)

4970

4980

4990

5000

Energy (eV)

Figure 5: (a) The experimental Ti K -edge spectrum of the sample ZnO-APT (black), compared to a computed spectrum for an amorphous TiO2 model (inset) that contains 50% Ti5c , 47% Ti6c , and 3% Ti7c . (b) Same experimental spectrum compared to non-negative least squares fitting to a representative dataset of computed spectra for a set of crystalline TiO2 polymorphs. One fit includes spectra for 4-, 5- and 6-coordinated Ti atoms (solid red), and the other fit is restricted to using Ti6c spectra only (dotted red). The best fit corresponds to 10% Ti4c , 32% Ti5c and 58% Ti6c atoms. constructive model of the amorphous structure that contains 50% Ti5c , 47% Ti6c and 3% Ti7c , where the subscript refers to local coordination number (inset to Figure 5a). In a second approach, we generalize the widely used procedure of fitting an unknown spectrum to reference spectra. Naturally, this approach is properly applicable to a polyphase sample with micro or nanocrystalline content. However, to the extent that the Ti K -edge XANES spectra are largely determined by the local structure, this approach may still give insight even in the limit of an amorphous sample. In both approaches, we simulate the XANES spectra. To assure the robustness of the analysis, the Ti K-edge XANES spectra are computed with different methods, including the core-hole pseudopotential method implemented in the XSpectra module of Quantum ESPRESSO, 37–40 and the Bethe-Salpeter equation-based method implemented in exciting 41,42 and OCEAN. 43,44 Results are systematically com-

12

ACS Paragon Plus Environment

Page 13 of 21 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

Nano Letters

pared. Details on methods and validation are described in the Supporting Information. 45–49 The computational spectrum of the amorphous model based on XSpectra, including contributions from all the Ti sites, is shown in Figure 5a in comparison to the experimental spectrum. The overall agreement for the features of the spectrum between the model and the experiment is excellent. In the pre-edge region, the computed peaks are slightly closer to the main edge, as compared to experiment. This is also observed when this computational method is tested against rutile and anatase (Supporting Information, Figures S2 and S3). An even better agreement in the pre-edge region is found between the OCEAN spectrum and the experimental spectrum as shown in the Supporting Information (Figure S4). These results suggest that an amorphous structural model with no nanocrystalline regions can account for the primary features observed in the spectra for the titania shells formed on the ZnO nanowires. To fit the unknown spectrum with reference spectra in the second approach, a large number of titania polymorphs listed in the Materials Project database 50 are used to build a database of titania structures with energy optimized internal structure at the density functional therory level, for which Ti K -edge XANES spectra are calculated. The resulting sample of 334 symmetry distinct local Ti structure motifs show a distribution of coordinations spanning Ti4c , Ti5c , and Ti6c centers. Using a clustering approach, a representative set of 11 reference spectra are developed using XSpectra spanning 4-, 5-, and 6-fold coordinated Ti centers, as detailed in the Supporting Information (Figure S5). 51 Linear regression is used to match the experimental spectrum as a (non-negative) linear combination of the computational XANES basis set. 52 The best fit obtained from the whole basis set is compared to the one from the basis set that only contains Ti6c elements in Figure 5b. The results are consistent with the broad expectations described above and semiquantitatively with the results found based on the synthetic amorphous model structure. Inclusion of Ti4c and Ti5c spectra significantly improves the agreement in the pre-edge region, which is particularly sensitive to the first coordination shell. The best fit consists 10% Ti4c , 32% Ti5c and 58%

13

ACS Paragon Plus Environment

Nano Letters 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

Ti6c . Both computational approaches capture the changes in the pre-edge region and in the white line region of the spectrum, while the full spectrum range has been considered during the comparison between theory and experiment. Both approaches demonstrate that the sample contains a significant proportion of undercoordinated Ti atoms. The glassy nature of the TiO2 shell found here is comparable to that observed in molten TiO2 53 and in a bulk phase TiO2 glass prepared at low temperature. 54,55 Both types of sample were found to contain undercoordinated Ti atoms, at various fractions depending on the processing parameters. In our case, although the precise breakdown of the distribution of local Ti atom coordination deduced does vary somewhat with the approach, we have carefully explored the factors involved, giving confidence in the main result, namely that roughly a half of the Ti atoms in the titania shell are undercoordinated. Compared to bulk phase TiO2 glass, the atomic motif of ultrathin TiO2 shell is more affected by surface and interfacial energy. For example, although it was reported that a relatively well defined anatase phase emerges from a bulk phase TiO2 glass when heated above 200◦ C, 54 there is no evidence for the formation of the anatase phase in our TiO2 shell, despite a growth temperature of 250◦ C. Interestingly, our second computational approach to analyze the Ti K -edge data identifies about 10% fourcoordinated Ti centers as contributing. Considering that the interface to ZnO represents a transition from four-coordinated to six-coordinated metal oxides, the structure in this region could have characteristics similar to those found in Zn-Ti-O ternary oxides where lower coordinated Ti has been observed. 56 This may also hint at a templating effect near the interface that stabilizes lower coordination numbers in the titania shell. In summary, we report that the quality of ALD titania coatings over ZnO nanowires is strongly correlated to the post-processing procedures performed on the NWs, including thermal annealing and plasma sputtering. We find that plasma sputtering in O2 wets the ZnO nanowires and significantly improves the uniformity and growth rate of the titania shell. Ti K -edge XANES studies at high spectral resolution indicate that the titania shell

14

ACS Paragon Plus Environment

Page 14 of 21

Page 15 of 21 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

Nano Letters

over ZnO has a significantly different structure from those of crystalline TiO2 . Ti L-edge XANES studies suggest that the titania shell is highly amorphous with crystalline domains limited to a size of 1 nm or smaller. Based on two different first-principle computational approaches to analyze the Ti K -edge data, we demonstrate that the experimental spectrum can be satisfactorily fitted only by introducing a large fraction (∼ 50%) of undercoordinated Ti atoms. Trends in the intensity ratio between the white line doublet in the Ti K -edge spectra measured for different post-processing procedures point to a lower oxidation state in the titania shell due to the plasma sputtering of the ZnO cores. In turn, this may play a role in the observation of improved photocatalytic performance with this processing approach.

Supporting Information The Supporting Information is available free of charge on the ACS Publications website. X-ray diffraction pattern of ZnO/TiO2 nanowire arrays, computational methods, and details for simulated Ti K -edge X-ray absorption near edge structure (XANES) spectra.

Acknowledgements This research was carried out at Brookhaven National Laboratory (BNL) under Contract No. DE-SC0012704 where facilities used included those of the Center for Functional Nanomaterials (CFN) and the beamlines 8-ID ISS (Inner Shell Spectroscopy) and 23-ID-2 IOS (In situ and Operando Soft X-ray Spectroscopy) in the National Synchrotron Light Source II, U.S. Department of Energy Office of Science User Facilities, and the Scientific Data and Computing Center, a component of the BNL Computational Science Initiative. This research also used resources of the Natinal Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility operated under Contract No. DE-AC02-05CH11231. The effort on computation by M.T. was supported by LDRD Project at BNL (No. 16-039). The primary research contribution of J.L.L. was carried out while 15

ACS Paragon Plus Environment

Nano Letters 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

at CFN supported as noted above, and J.L.L. now acknowledges support from the Office of Naval Research through the Naval Research Laboratory’s Basic Research Program.

References (1) Ramanathan, S., Ed. Thin Film Metal-Oxides; Springer: Boston, MA, 2010. (2) Pryds, N., Esposito, V., Eds. Metal Oxide-Based Thin Film Structures; Metal Oxides; Elsevier, 2018. (3) Kingon, A. I.; Maria, J.-P.; K., S. S. Nature 2000, 406, 1032–1038. (4) Hoex, B.; Heil, S. B. S.; Langereis, E.; van de Sanden, M. C. M.; Kessels, W. M. M. App. Phys. Lett. 2006, 89, 042112. (5) Richards, B. S.; Cotter, J. E.; Honsberg, C. B. App. Phys. Lett. 2002, 80, 1123–1125. (6) Bonilla, R. S.; Hoex, B.; Hamer, P.; Wilshaw, P. R. Phys. Status Solidi A 214, 1700293. (7) Li, L.; Xu, C.; Zhao, Y.; Chen, S.; Ziegler, K. J. ACS Appl. Mater. Interfaces 2015, 7, 12824–12831. (8) Jeong, I.; Park, Y. H.; Bae, S.; Park, M.; Jeong, H.; Lee, P.; Ko, M. J. ACS Appl. Mater. Interfaces 2017, 9, 36865–36874. (9) Frank, S. N.; Bard, A. J. J. Phys. Chem. 1977, 81, 1484–1488. (10) Gerischer, H. J. Electroanal. Chem. 1977, 82, 133 – 143. (11) Chen, Y. W.; Prange, J. D.; Duehnen, S.; Park, Y.; Gunji, M.; Chidsey, C. E. D.; McIntyre, P. C. Nature Mater. 2011, 10, 539–544. (12) Liu, M.; Nam, C. Y.; Black, C. T.; Kamcev, J.; Zhang, L. J. Phys. Chem. C 2013, 117, 13396–13402. 16

ACS Paragon Plus Environment

Page 16 of 21

Page 17 of 21 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

Nano Letters

(13) Scheuermann, A. G.; Prange, J. D.; Gunji, M.; Chidsey, C. E. D.; McIntyre, P. C. Energy Environ. Sci. 2013, 6, 2487–2496. (14) Hu, S.; Shaner, M. R.; Beardslee, J. A.; Lichterman, M.; Brunschwig, B. S.; Lewis, N. S. Science 2014, 344, 1005–1009. (15) Appavoo, K.; Liu, M.; Black, C. T.; Sfeir, M. Y. Nano Lett. 2015, 15, 1076–1082. (16) Siddiqi, G.; Luo, Z.; Xie, Y.; Pan, Z.; Zhu, Q.; Röhr, J. A.; Cha, J. J.; Hu, S. ACS Appl. Mater. Interfaces 2018, 10, 18805–18815, PMID: 29668253. (17) George, S. M. Chem. Rev. 2010, 110, 111–131. (18) Bayn, I.; Chen, E. H.; Trusheim, M. E.; Li, L.; Schröder, T.; Gaathon, O.; Lu, M.; Stein, A.; Liu, M.; Kisslinger, K.; Clevenson, H.; Englund, D. Nano Lett. 2015, 15, 1751–1758. (19) Lee, H.-B.-R.; Bent, S. F. Chem. Mater. 2012, 24, 279–286. (20) Greene, L. E.; Law, M.; Goldberger, J.; Kim, F.; Johnson, J. C.; Zhang, Y.; Saykally, R. J.; Yang, P. Angew. Chem. Int. Ed. 2003, 42, 3031–3034. (21) Greene, L. E.; Law, M.; Tan, D. H.; Montano, M.; Goldberger, J.; Somorjai, G.; Yang, P. Nano Lett. 2005, 5, 1231–1236. (22) Yan, D.; Tao, J.; Kisslinger, K.; Cen, J.; Wu, Q.; Orlov, A.; Liu, M. Nanoscale 2015, 7, 18515–18523. (23) Yan, D.; Zhang, W.; Cen, J.; Stavitski, E.; Sadowski, J. T.; Vescovo, E.; Walter, A.; Attenkofer, K.; Stacchiola, D. J.; Liu, M. Appl. Phys. Lett. 2017, 111, 231901. (24) Khan, W.; Ajmal, H. M. S.; Khan, F.; Huda, N. U.; Kim, S.-D. Nanomaterials 2018, 8.

17

ACS Paragon Plus Environment

Nano Letters 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

(25) Kucheyev, S. O.; van Buuren, T.; Baumann, T. F.; Satcher, J. H.; Willey, T. M.; Meulenberg, R. W.; Felter, T. E.; Poco, J. F.; Gammon, S. A.; Terminello, L. J. Phys. Rev. B 2004, 69, 245102. (26) Krüger, P. J. Phys.: Conf. Ser. 2009, 190, 012006. (27) Harada, I.; Suzuki, T.; Kotani, A. J. Phys. Soc. Jpn. 1996, 65, 3016–3020. (28) Greegor, R. B.; Lytle, F. W.; Sandstrom, D. R.; Wong, J.; Schultz, P. J. Non-Cryst. Solids 1983, 55, 27 – 43. (29) Mountjoy, G.; Pickup, D. M.; Wallidge, G. W.; Anderson, R.; Cole, J. M.; Newport, R. J.; Smith, M. E. Chem. Mater. 1999, 11, 1253–1258. (30) Uozumi, T.; Okada, K.; Kotani, A.; Durmeyer, O.; Kappler, J. P.; Beaurepaire, E.; Parlebas, J. C. EPL (Europhys. Lett.) 1992, 18, 85. (31) Wu, Z. Y.; Ouvrard, G.; Gressier, P.; Natoli, C. R. Phys. Rev. B 1997, 55, 10382–10391. (32) Durmeyer, O.; Beaurepaire, E.; Kappler, J.-P.; Brouder, C.; Baudelet, F. J. Phys.: Condens. Matter 2010, 22, 125504. (33) Farges, F.; Brown, G. E.; Rehr, J. J. Phys. Rev. B 1997, 56, 1809–1819. (34) Pham, H. H.; Wang, L.-W. Phys. Chem. Chem. Phys. 2015, 17, 541–550. (35) Kresse, G.; Furthmüller, J. Phys. Rev. B 1996, 54, 11169–11186. (36) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758–1775. (37) Gougoussis, C.; Calandra, M.; Seitsonen, A.; Brouder, C.; Shukla, A.; Mauri, F. Phys. Rev. B 2009, 79, 045118. (38) Taillefumier, M.; Cabaret, D.; Flank, A.-M.; Mauri, F. Phys. Rev. B 2002, 66, 195107. (39) Giannozzi, P. et al. Journal of Physics: Condensed Matter 2009, 21, 395502. 18

ACS Paragon Plus Environment

Page 18 of 21

Page 19 of 21 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

Nano Letters

(40) England, A. H.; Duffin, A. M.; Schwartz, C. P.; Uejio, J. S.; Prendergast, D.; Saykally, R. J. Chem. Phys. Lett. 2011, 514, 187–195. (41) Gulans, A.; Kontur, S.; Meisenbichler, C.; Nabok, D.; Pavone, P.; Rigamonti, S.; Sagmeister, S.; Werner, U.; Draxl, C. J. Phys.: Condens. Matter 2014, 26, 363202. (42) Vorwerk, C.; Cocchi, C.; Draxl, C. Phys. Rev. B 2017, 95, 155121. (43) Vinson, J.; Rehr, J. J.; Kas, J. J.; Shirley, E. L. Phys. Rev. B 2011, 83, 115106. (44) Vinson, J.; Jach, T.; Elam, W. T.; Denlinger, J. D. Phys. Rev. B 2014, 90, 205207. (45) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865–3868. (46) Lejaeghere, K. et al. Science 2016, 351 . (47) Sarkar, K.; Topsakal, M.; Holzwarth, N.; Wentzcovitch, R. M. J. Comput. Phys. 2017, 347, 39 – 55. (48) Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188–5192. (49) Prandini, G.; Marrazzo, A.; Castelli, I.; Mounet, N.; Marzari, N. Materials Cloud Archive 2018, (50) Jain, A.; Ong, S. P.; Hautier, G.; Chen, W.; Richards, W. D.; Dacek, S.; Cholia, S.; Gunter, D.; Skinner, D.; Ceder, G.; Persson, K. A. APL Mater. 2013, 1, 011002. (51) Ester, M.; Kriegel, H.-P.; Sander, J.; Xu, X. A density-based algorithm for discovering clusters in large spatial databases with noise. 1996; pp 226–231. (52) NNLS

by

SciPy.

https://docs.scipy.org/doc/scipy/reference/generated/

scipy.optimize.nnls.html, Accessed: 2018-08-24. (53) Alderman, O. L. G.; Skinner, L. B.; Benmore, C. J.; Tamalonis, A.; Weber, J. K. R. Phys. Rev. B 2014, 90, 094204. 19

ACS Paragon Plus Environment

Nano Letters 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

(54) Bersani, D.; Lottici, P. P.; Braghini, M.; Montenero, A. Phys. Status Solidi B 1992, 170, K5–K10. (55) Landmann, M.; Köhler, T.; Köppen, S.; Rauls, E.; Frauenheim, T.; Schmidt, W. G. Phys. Rev. B 2012, 86, 064201. (56) Caretti, I.; Yuste, M.; Torres, R.; SÃąnchez, O.; JimÃľnez, I.; Escobar Galindo, R. RSC Adv. 2012, 2, 2696–2699.

20

ACS Paragon Plus Environment

Page 20 of 21

Page 21 of 21 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

Nano Letters

ACS Paragon Plus Environment