Bi2Pt2O7 pyrochlore is thought to be one of the most promising oxide catalysts for application in fuel cell technology. Unfortunately, direct film growth of Bi2Pt2O7 has not yet been achieved, owing to the difficulty of oxidizing platinum metal in the precursor material to Pt4+. In this work, in order to induce oxidation of the platinum, we annealed pulsed laser deposited films consisting of epitaxial δ–Bi2O3 and co-deposited, comparatively disordered platinum. We present synchrotron x-ray diffraction results that show the nonuniform annealed films contain the first epitaxial crystals of Bi2Pt2O7. We also visualized the pyrochlore structure by scanning transmission electron microscopy, and observed ordered cation vacancies in the epitaxial crystals formed in a bismuth-rich film but not in those formed in a platinum-rich film. The similarity between the δ–Bi2O3 and Bi2Pt2O7 structures appears to facilitate the pyrochlore formation. These results provide the only route to date for the formation of epitaxial Bi2Pt2O7.

Pyrochlore oxides,1 A2B2O7, that exhibit catalytic activity toward oxygen reduction and mixed ionic-electronic conduction are prospective cathode materials for various types of fuel cells.2–6 Specifically, pyrochlore powders have been studied as cathode materials in solid oxide fuel cells (SOFCs).5–7 Additionally, certain pyrochlores are among the best oxygen catalysts in alkaline media,2 and some of them have also been envisaged as electrodes for direct methanol fuel cells. In particular, motivated by the nearly complete tolerance toward poisoning by CO of nanoparticles of the ordered intermetallic phase PtBi,8 Beck et al. prepared powder electrodes of Bi2Pt2O7 pyrochlore.9 They found its oxygen reduction activity, after electrochemical activation, essentially identical to pure Pt.

Epitaxial oxide thin films can have superior properties compared to powders. For instance, enhanced catalytic activity has been observed in epitaxially grown perovskite oxides and structural derivatives for application as SOFC cathodes.10–13 This enhancement has been related to tensile lattice strain.14 Thus, there is an increasing interest in epitaxial films of materials with crystal structures different from perovskite, such as pyrochlore, not yet studied in epitaxial form for fuel cell applications. Furthermore, fundamental studies on single-crystals displaying well-defined surface composition and crystallographic orientation can shed light upon electrode reactions relying on the electrochemical properties of the materials at the nanometer scale.15–17 

In this context, exploring epitaxial Bi2Pt2O7 is of great interest. Nanocrystalline powders of Bi2Pt2O7 by thermal oxidation of Bi-Pt nanoparticles have been reported.18 However, direct film growth of epitaxial Bi2Pt2O7 pyrochlore has not yet been achieved. Here, we report the first, successful formation of epitaxial crystals (∼100 nm long) of Bi2Pt2O7(111) on YSZ(111) single-crystal substrates. Our synthesis strategy consists of a deposition step followed by a post-growth anneal. During growth, epitaxial δ–Bi2O3 and co-deposited, comparatively disordered platinum are formed. Annealing in air transforms the δ–Bi2O3 fluorite and a portion of disordered platinum into epitaxial crystals of Bi2Pt2O7 pyrochlore that are ∼100 nm long.

We studied the out-of-plane orientation of the films by x-ray diffraction (XRD). Fig. 1(a) shows θ–2θ scans of an as-deposited film (blue pattern). The growth was carried out at an oxygen pressure of 10−4 Torr, 640 °C, and a laser fluence of 3 J/cm2.19 The as-grown pattern reveals intense peaks that can be assigned to (111), (222), and (333) (not shown) cubic δ–Bi2O3. Distinct thickness fringes are observed surrounding the (111) Bragg peak of δ–Bi2O3 in the as-grown films (Fig. S119) which give evidence of a smooth film surface. Rocking curve measurements for the (111) δ–Bi2O3 peak (Fig. S219) show a FWHM of 0.100 ± 0.001° in omega. The measured out-of-plane lattice parameter for δ–Bi2O3 is 5.519 Å, in good agreement with previous reported values.20,21 Although the cubic δ phase of Bi2O3 in bulk form is stable only from 729 °C up to its melting point at 825 °C and transforms to other phases upon cooling, it has previously been stabilized at room temperature on Au substrates,20 on polycrystalline YSZ substrates,22 and as δ–Bi2O3 nanostructures on perovskite substrates.21 

FIG. 1.

Structural rearrangement of δ–Bi2O3 and disordered platinum through post-growth anneal. (a) XRD θ–2θ scans of a film as-grown (blue), and after post-growth anneal (orange), showing the formation of the pyrochlore phase during the anneal. (b) Normalized grazing incidence in situ x-ray fluorescence (XRF). Bi and Pt x-ray emissions in the range 8.5 keV – 12 keV are marked with vertical lines whose heights are proportional to their relative intensity. XRF analysis shows a Bi/Pt ratio of 0.88 ± 0.01. The Hf impurity of the YSZ substrate is also included. (c) Pt L3 edge X-ray absorption near edge structure (XANES) spectra of an annealed film (orange); and, PtO2 (green) and Pt metal foil (black) used as reference standards. The intense peak confirms the oxidation of platinum in the annealed film. (d) Atomic force microscopy (AFM) image of the annealed film in panels (a) and (b). (e) The order of Bi3+ and Pt4+ along 〈110〉 determines two types of fluorite subcells in the pyrochlore structure. Cubic δ–Bi2O3 with fluorite structure where each anion site has an average occupancy of 3/4.23 

FIG. 1.

Structural rearrangement of δ–Bi2O3 and disordered platinum through post-growth anneal. (a) XRD θ–2θ scans of a film as-grown (blue), and after post-growth anneal (orange), showing the formation of the pyrochlore phase during the anneal. (b) Normalized grazing incidence in situ x-ray fluorescence (XRF). Bi and Pt x-ray emissions in the range 8.5 keV – 12 keV are marked with vertical lines whose heights are proportional to their relative intensity. XRF analysis shows a Bi/Pt ratio of 0.88 ± 0.01. The Hf impurity of the YSZ substrate is also included. (c) Pt L3 edge X-ray absorption near edge structure (XANES) spectra of an annealed film (orange); and, PtO2 (green) and Pt metal foil (black) used as reference standards. The intense peak confirms the oxidation of platinum in the annealed film. (d) Atomic force microscopy (AFM) image of the annealed film in panels (a) and (b). (e) The order of Bi3+ and Pt4+ along 〈110〉 determines two types of fluorite subcells in the pyrochlore structure. Cubic δ–Bi2O3 with fluorite structure where each anion site has an average occupancy of 3/4.23 

Close modal

The weak diffraction peak at q = 2.77 Å−1 in Fig. 1 (a, blue pattern) is consistent with that of the (111) plane of a trace amount of platinum metal with a fcc structure. Apart from the minute amounts of (111)-oriented platinum, these data are in agreement with either amorphous platinum or ∼1 nm platinum nanocrystals that diffract similarly weakly to amorphous platinum.

The phase segregation of the platinum metal and Bi2O3 is a major difficulty of working with this system. Like the as-deposited film, the pulsed laser deposition (PLD) target exhibits phase separation of the platinum from the bismuth oxide. The precursor material exhibits the pyrochlore phase (Bi3+, Pt4+) when prepared by solid state reaction at 650 °C from stoichiometric mixtures of Bi2O3 and platinum metal (Fig. S319). However, as expected, targets sintered at this temperature are not dense enough for proper ablation. Sintering temperatures higher than 650 °C are required to synthesize high-density PLD targets, but at these temperatures, Pt4+ reduces to platinum metal. Thus, PLD targets sintered at 820 °C (slightly below the melting point of Bi2O3) contain Bi2O3 in its monoclinic α phase upon cooling, and platinum (Fig. S319). Consequently, platinum metal has to oxidize to Pt4+ so that the pyrochlore phase can form on the substrate. Furthermore, control of the Bi/Pt cation stoichiometry of the film is complicated due to the high volatility of the bismuth and the large difference between the melting temperatures of the two components of the target (Bi2O3, 825 °C; Pt, 1770 °C). Bi-based perovskite films with the correct stoichiometry have been grown by PLD from bismuth-rich targets24,25 as well as from stoichiometric targets.26,27 There is no previous work on epitaxial films of Bi2Pt2O7. We used in situ XRF at grazing incidence to track the Bi/Pt ratio of the films while tuning the deposition parameters. We chose to use a stoichiometric target, and compensate for the preferential ablation of bismuth that we observed in this system by working at a pressure of 10−4 Torr, lower than the bismuth vapor pressure at 640 °C (bismuth vapor pressure is 10−4 Torr at 517 °C; 10−2 Torr at 672 °C). This leads to the sublimation of a portion of the deposited bismuth. Fig. 1(b) displays the XRF spectrum for the film shown in panel (a) measured in situ once the substrate temperature was cooled down. The quantitative analysis19 provides an estimate Bi/Pt ratio of 0.88 ± 0.01 within a 95% confidence level.

In order to induce oxidation of the platinum, the film shown in Fig. 1 (a, blue scan) was annealed in a tube furnace in air at 640 °C for 8 h. Fig. 1 (a, orange scan) displays the θ–2θ scan of the film after annealing. This pattern shows two new peaks that can be attributed to (222) and (444) pyrochlore with an out-of-plane lattice parameter of 10.57 ± 0.02 Å. The reflections assigned in the as-grown film to {111} planes of δ–Bi2O3 have disappeared while the weak peak attributed to some trace of (111) platinum metal is still observable. These results, together with the Bi/Pt ratio of 0.88 ± 0.01 suggest, that the Pt (111) peak in the XRD scan of Fig. 1 (a, blue pattern) cannot fully account for all the deposited platinum in the film. Rocking curve measurements for the Bi2Pt2O7(222) peak (Fig. S419) reveal a FWHM of 0.098 ± 0.002° in omega, comparable to that of the (111) δ–Bi2O3 peak (Fig. S219).

We investigated the oxidation state of platinum in the annealed films. Fig. 1(c) shows Pt L3 edge XANES spectrum (orange) of a bismuth-rich annealed film (Bi/Pt = 1.62 ± 0.04) with pyrochlore phase (Figs. S5 and S619). A platinum metal foil (black), and PtO2 (green) were used as reference standards. The higher threshold energy (E0) of the film in relation to that of the platinum reference foil is consistent with a higher oxidation state in the annealed film, most probably Pt4+. The intense peak observed in the spectrum of the film (Fig. 1(c), orange) is also indicative of a metal oxide. It is thus clear that a portion of the platinum oxidized during the post-growth anneal.

The smoothness of the surface of the annealed film in Figs. 1(a) and 1(b) was analyzed by AFM. The AFM image shown in Fig. 1(d) reveals a rms roughness of 0.84 nm. It was grown on a YSZ (111) stepped surface with atomically flat terraces (Fig. S719).

The structural models shown in Fig. 1(e) reveal striking similarity between the cubic δ–Bi2O3 and the pyrochlore structures, both based on an ordered oxygen deficient fluorite structure. This strongly suggests the stabilization of the δ phase of Bi2O3 is essential for the formation of the pyrochlore phase during the ex situ post-growth anneal. In fact, films that contained the monoclinic (α) phase of Bi2O3 still exhibit α-Bi2O3 as the majority phase after annealing (Fig. S819).

Both the substrate temperature and the oxygen pressure during growth determine the rate of sublimation of the deposited bismuth, thus defining the Bi/Pt ratio of the films. The increased bismuth desorption at high temperatures limits the growth conditions for this system where a Bi/Pt ratio close to 1 can be achieved. The supply of excess of bismuth may facilitate the control of the stoichiometry. This approach has been successfully applied in the epitaxial growth of bismuth containing oxides by molecular beam epitaxy.28 The working pressure also defines the structure of the deposited platinum. When the growth is carried out at a higher oxygen pressure (10−2 Torr) platinum crystallizes as Pt(111) and Pt(200) phases instead of being deposited as mostly disordered platinum, which hinders the structural rearrangement that leads to the pyrochlore phase (Fig. S919).

Synchrotron-based off-specular, grazing incidence x-ray diffraction results (Fig. 2) show definitively that the annealed films contain the first epitaxial crystals of (111)-oriented bismuth platinum pyrochlore oxide. Along with high intensity peaks characteristic of the parent fluorite structure (h + k, k + l, and l + h are all multiples of four), we observed weak pyrochlore superstructure peaks (their intensity is expected to be three orders of magnitude lower), such as ( 33 3 ̄ ) and (331).

FIG. 2.

Synchrotron x-ray reciprocal space maps (RSMs) and ϕ-scans. (a) Calculated positions for reflections of Bi2Pt2O7 (gray filled circles), assuming the Bi2Pt2O7 bulk lattice constant, 10.371 Å. Substrate reflections (red hollow triangles) are also expected to occur for the fluorite-like reflections of the pyrochlore lattice, where h + k, k + l, and l + h are all multiples of four. h′, k′, l′ stand for the coordinates of reciprocal vectors with respect to the set of reciprocal lattice vectors generated from { ( 1 ̄ 01 ) / 2 , ( 1 2 ̄ 1 ) / 6 , ( 111 ) / 3 } . r.l.u = reciprocal lattice units of the Bi2Pt2O7 bulk lattice. RSMs combined into a single image for a bismuth-rich film (b), and a platinum-rich film (c). The insets show an enlargement of reciprocal space for the superstructure reflections (331) and (33 3 ̄ ) of the pyrochlore phase. The intensity of all the figures is presented on a logarithmic scale. (d) Azimuthal ϕ scan of the (200) reflection of the δ–Bi2O3 phase in an as-grown film, and (e) of the ( 22 2 ̄ ) reflection of the Bi2Pt2O7 phase of an annealed film. The incident angle (α) was fixed at 0.25° and 0.275°, respectively. The exit angle, β = δα, was collected in a range of 12°.

FIG. 2.

Synchrotron x-ray reciprocal space maps (RSMs) and ϕ-scans. (a) Calculated positions for reflections of Bi2Pt2O7 (gray filled circles), assuming the Bi2Pt2O7 bulk lattice constant, 10.371 Å. Substrate reflections (red hollow triangles) are also expected to occur for the fluorite-like reflections of the pyrochlore lattice, where h + k, k + l, and l + h are all multiples of four. h′, k′, l′ stand for the coordinates of reciprocal vectors with respect to the set of reciprocal lattice vectors generated from { ( 1 ̄ 01 ) / 2 , ( 1 2 ̄ 1 ) / 6 , ( 111 ) / 3 } . r.l.u = reciprocal lattice units of the Bi2Pt2O7 bulk lattice. RSMs combined into a single image for a bismuth-rich film (b), and a platinum-rich film (c). The insets show an enlargement of reciprocal space for the superstructure reflections (331) and (33 3 ̄ ) of the pyrochlore phase. The intensity of all the figures is presented on a logarithmic scale. (d) Azimuthal ϕ scan of the (200) reflection of the δ–Bi2O3 phase in an as-grown film, and (e) of the ( 22 2 ̄ ) reflection of the Bi2Pt2O7 phase of an annealed film. The incident angle (α) was fixed at 0.25° and 0.275°, respectively. The exit angle, β = δα, was collected in a range of 12°.

Close modal

The expected positions for the reflections of the epitaxial Bi2Pt2O7 ∥ Y SZ system are depicted in Fig. 2(a). RSMs are plotted in Figs. 2(b) and 2(c) for a bismuth-rich film and for a slightly platinum-rich film, respectively. δ–Bi2O3 still present in the bismuth-rich film shows up in the RSM of the pyrochlore phase, revealing again the similarity between the fluorite and pyrochlore structures (Fig. S1019). Superstructure peaks indicative of the pyrochlore phase are highlighted.

In-plane (a) and out-of-plane (a) lattice constants of the epitaxial pyrochlore phase, along [ 0 1 ̄ 1 ] and [111], respectively, are a = 10.46 ± 0.03 Å and a = 10.423 ± 0.006 Å, for the film in Fig. 2(b), and, a = 10.53 ± 0.08 Å and a = 10.53 ± 0.04 Å (in accordance, within the error, with the value obtained from the θ–2θ scan in Fig. 1(a)) for the film in Fig. 2(c). These figures suggest the pyrochlore is relaxed with respect to the substrate lattice.

Azimuthal ϕ scans on the (200) peak of δ–Bi2O3 of an as-grown film, Fig. 2(d), and on the ( 22 2 ̄ ) peak of Bi2Pt2O7 in an annealed film, Fig. 2(e), confirm the epitaxial nature of both phases. The FWHM of (200) peaks of δ–Bi2O3 is 0.50° ± 0.01, and that of the ( 22 2 ̄ ) peaks separated 120° in the ϕ scan is 0.57° ± 0.02. In-plane orientation relationships of either [ 10 1 ̄ ] YSZ [ 10 1 ̄ ] Bi 2 Pt 2 O 7 or [ 10 1 ̄ ] YSZ [ 1 1 ̄ 0 ] Bi2Pt2O7, rotated 60° respect to each other, can be built with the structural model of Bi2Pt2O7(111)∥ YSZ(111) shown in Fig. S11.19 However, no twin domains were observed in our films.

Fig. 3 shows cross-sectional HAADF scanning TEM images for the films whose RSMs are displayed in Figs. 2(b) and 2(c). These studies reveal ∼100 nm long regions of ordered epitaxial pyrochlore. A wider field of view scanning TEM images showing non-pyrochlore regions of the films are shown in Fig. S12.19 Fig. 3(a) exhibits the expected columns of a pyrochlore structure with cation ordered vacancies (no noticeable contrast in the HAADF is expected between Bi and Pt columns). Fig. 3(b) shows a schematic of the bismuth and platinum columns of the Bi2Pt2O7 pyrochlore structure viewed along [ 1 ̄ 01 ] . Black positions mark where the ordered cation vacancies are expected, according to the STEM image in Fig. 3(a). A region of epitaxial pyrochlore formed in the annealed platinum-rich film is shown in Fig. 3(c). This pyrochlore structure does not contain cation vacancies. Thus, its composition is expected to be stoichiometric, Bi2Pt2O7.

FIG. 3.

High-angle annular dark field (HAADF) scanning TEM images. (a) HAADF STEM image of the ∼100 nm long regions of ordered epitaxial pyrochlore in the annealed bismuth-rich film shown in Fig. 2(b), viewed along [ 1 ̄ 01 ] in the NION UltraSTEM. The dark, ordered atomic locations in the image are the result of cation vacancies. (b) Schematic of the bismuth and platinum columns of the Bi2Pt2O7 pyrochlore structure viewed along [ 1 ̄ 01 ]. Black positions mark where the ordered cation vacancies of the image in panel (a) are expected, matching the model in Fig. S11.19 (c) HAADF STEM image of the pyrochlore structure in the annealed platinum-rich film shown in Fig. 2(c), viewed along [ 1 ̄ 01 ] in the Tecnai F20. Cation vacancies are not observed in this structure. A wider field of view scanning TEM images showing non-pyrochlore regions of the films are shown in Fig. S12.19 

FIG. 3.

High-angle annular dark field (HAADF) scanning TEM images. (a) HAADF STEM image of the ∼100 nm long regions of ordered epitaxial pyrochlore in the annealed bismuth-rich film shown in Fig. 2(b), viewed along [ 1 ̄ 01 ] in the NION UltraSTEM. The dark, ordered atomic locations in the image are the result of cation vacancies. (b) Schematic of the bismuth and platinum columns of the Bi2Pt2O7 pyrochlore structure viewed along [ 1 ̄ 01 ]. Black positions mark where the ordered cation vacancies of the image in panel (a) are expected, matching the model in Fig. S11.19 (c) HAADF STEM image of the pyrochlore structure in the annealed platinum-rich film shown in Fig. 2(c), viewed along [ 1 ̄ 01 ] in the Tecnai F20. Cation vacancies are not observed in this structure. A wider field of view scanning TEM images showing non-pyrochlore regions of the films are shown in Fig. S12.19 

Close modal

The presence of vacancies in the pyrochlore crystal shown in Fig. 3(a) is not surprising. The pyrochlore structure (A2B2O7, also represented as A2B2O6O′) is well known to tolerate anion (in one of the two nonequivalent oxygen positions, O′) and A-site cation vacancies. The BO6 octahedra are the basic structural units, and the A and O′ ions of the A2O′ chains (Fig. S1119) are not essential for the stabilization of the pyrochlore structure.1 Moreover, non-stoichiometry accommodated by cation disorder has also been observed in pyrochlores. For instance, cation antisite disorder (A- and B-site mixing) has been reported in non-stoichiometric Gd2Zr2O7 pyrochlore.29 Our data, however, do not provide information about cation disorder in the defect structure (Fig. 3(a)) nor the deviation from stoichiometry (Bi/Pt ratio) in its cation sublattice.

An amorphous layer 1.2–1.4 nm thick is observed at the interface between the pyrochlore phase and the YSZ substrate. This may be due to solid-state reactions at the interface pyrochlore∥Y SZ. A fundamental understanding of the pyrochlore∥Y SZ interface is lacking since only very recently the growth of high quality single crystals of pyrochlore has been reported for the first time on YSZ substrates (Ho2Ti2O7∥Y SZ),30 and non-commercially available pyrochlore substrates (Dy2Ti2O7∥Y 2Ti2O7).31 

In conclusion, these results provide the only currently known method to form epitaxial Bi2Pt2O7 pyrochlore, thought to be one of the most promising oxide catalysts for fuel cell applications, and a step forward in the study of thin films of complex oxides with crystal geometry different of perovskite. The electrochemical activity of the ∼100 nm long epitaxial Bi2Pt2O7 crystals obtained here could be probed on the scale of several nanometers by the novel approach electrochemical strain microscopy.16 Alternatively, uniform epitaxial films grown by this approach with the correct stoichiometry could be used to investigate the oxygen reduction activity of the (111) surface of Bi2Pt2O7 at a macroscopic level. In this case, the difficulty of achieving the correct Bi/Pt stoichiometry suggests that independent control of volatile bismuth and non-volatile platinum sources may be required for routine growth of pristine epitaxial films.

We acknowledge H. Paik for helpful discussions, D. Schlom for making the equipment in his laboratory available to prepare YSZ substrates, and R. Burns and F. DiSalvo for providing us an initial amount of pyrochlore powder to start this work. This work is based upon research conducted at the Cornell High Energy Synchrotron Source (CHESS) which is supported by the National Science Foundation and the National Institutes of Health/National Institute of General Medical Sciences under NSF Award Nos. DMR-0936384 and DMR-1332208. This work also made use of the Cornell Center for Materials Research Shared Facilities supported through the NSF MRSEC program (DMR-1120296). A.G.L. acknowledges financial support from the Spanish Ministry of Education under research grant PRX12/00405; the CajaMadrid Foundation (Spain) under a research grant, 2012 call; and, the Energy Materials Center at Cornell (emc2), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Award No. DE-SC0001086.

A.G.L. designed the experiments and led this research project; synthesized the targets, prepared the substrates, and grew the films; performed the XRD symmetrical scans, the in situ XRF, and the AFM measurements; analyzed the XRD, RSMs, and XRF data; prepared the figures and wrote the manuscript. M.C.S., H.J., and A.G.L. tuned the PLD system up. A.W. and A.G.L. performed area scans for asymmetrical reflections at G2. A.W. wrote the MATLAB code to transform area scans carried out at G2 into RSMs, and to combine several RSMs into a single figure, and began XRF studies for this system on Si substrates. M.E.H. and D.A.M. performed STEM studies. M.C.S. assisted in preliminary growths of films consisting of Bi2O3 ant Pt phases. M.J.W. and H.J. measured XANES at Pt L3 edge. J.D.B. made available the resources and equipments of his group to develop this research. All the authors discussed the manuscript.

1.
M. A.
Subramanian
,
G.
Aravamudan
, and
G. V. S.
Rao
,
Prog. Solid State Chem.
15
,
55
(
1983
).
2.
H. S.
Horowitz
,
J. M.
Longo
, and
H. H.
Horowitz
,
J. Electrochem. Soc.
130
,
1851
(
1983
).
3.
4.
B. J.
Wuensch
,
K. W.
Eberman
,
C.
Heremans
,
E. M.
Ku
,
P.
Onnerud
,
E. M.
Yeo
,
S. M.
Haile
,
J. K.
Stalick
, and
J. D.
Jorgensen
,
Solid State Ionics
129
,
111
(
2000
).
5.
A.
Jaiswal
and
E. D.
Wachsman
,
J. Electrochem. Soc.
152
,
A787
(
2005
).
6.
G.
Ehora
,
S.
Daviero-Minaud
,
M. C.
Steil
,
L.
Gengembre
,
M.
Frere
,
S.
Bellayer
, and
O.
Mentre
,
Chem. Mater.
20
,
7425
(
2008
).
7.
R.
Martínez-Coronado
,
J.
Alonso
,
V.
Cascos
, and
M.
Fernández-Díaz
,
J. Power Sources
247
,
876
(
2014
).
8.
C.
Roychowdhury
,
F.
Matsumoto
,
P. F.
Mutolo
,
H. D.
Abruna
, and
F. J.
DiSalvo
,
Chem. Mater.
17
,
5871
(
2005
).
9.
N. K.
Beck
,
B.
Steiger
,
G. G.
Scherer
, and
A.
Wokaun
,
Fuel Cells
6
,
26
(
2006
).
10.
G. J.
la O
,
S. J.
Ahn
,
E.
Crumlin
,
Y.
Orikasa
,
M. D.
Biegalski
,
H. M.
Christen
, and
Y.
Shao-Horn
,
Angew. Chem., Int. Ed.
49
,
5344
(
2010
).
11.
J.
Liu
,
G.
Collins
,
M.
Liu
,
C.
Chen
,
J.
He
,
J.
Jiang
, and
E. I.
Meletis
,
Appl. Phys. Lett.
100
,
193903
(
2012
).
12.
H.
Jeen
,
Z.
Bi
,
W. S.
Choi
,
M. F.
Chisholm
,
C. A.
Bridges
,
M. P.
Paranthaman
, and
H. N.
Lee
,
Adv. Mater.
25
,
6459
(
2013
).
13.
M.
Kubicek
,
Z.
Cai
,
W.
Ma
,
B.
Yildiz
,
H.
Hutter
, and
J.
Fleig
,
ACS Nano
7
,
3276
(
2013
).
15.
K. A.
Stoerzinger
,
M.
Risch
,
J.
Suntivich
,
W. M.
Lu
,
J.
Zhou
,
M. D.
Biegalski
,
H. M.
Christen
,
Ariando
,
A. T.
Venkatesan
, and
Y.
Shao-Horn
,
Energy Environ. Sci.
6
,
1582
(
2013
).
16.
A.
Kumar
,
F.
Ciucci
,
A. N.
Morozovska
,
S. V.
Kalinin
, and
S.
Jesse
,
Nat. Chem.
3
,
707
(
2011
).
17.
S. V.
Kalinin
,
A.
Borisevich
, and
D.
Fong
,
ACS Nano
6
,
10423
(
2012
).
18.
F.
Dawood
,
B. M.
Leonard
, and
R. E.
Schaak
,
Chem. Mater.
19
,
4545
(
2007
).
19.
See supplementary material at http://dx.doi.org/10.1063/1.4908103 E-AMPADS-3-008502 for a detailed experimental description and Figures S1-S12.
20.
J.
Switzer
,
M. G.
Shumsky
, and
E.
Bohannan
,
Science
5412
,
293
(
1999
).
21.
D. L.
Proffit
,
G.-R.
Bai
,
D. D.
Fong
,
T. T.
Fister
,
S. O.
Hruszkewycz
,
M. J.
Highland
,
P. M.
Baldo
,
P. H.
Fuoss
,
T. O.
Mason
, and
J. A.
Eastman
,
Appl. Phys. Lett.
96
,
021905
(
2010
).
22.
T.
Takeyama
,
N.
Takahashib
,
T.
Nakamurab
, and
S.
Itoh
,
Surf. Coat. Technol.
200
,
4797
(
2006
).
23.
M.
Yashima
and
D.
Ishimura
,
Chem. Phys. Lett.
378
,
395
(
2003
).
24.
S.
Havelia
,
S.
Wang
,
M.
Skowronski
, and
P. A.
Salvador
,
J. Appl. Phys.
106
,
123509
(
2009
).
25.
V.
Shelke
,
V. N.
Harshan
,
S.
Kotru
, and
A.
Gupta
,
J. Appl. Phys.
106
,
104114
(
2009
).
26.
L.
You
,
N. T.
Chu
,
K.
Yao
,
L.
Chen
, and
J.
Wang
,
Phys. Rev. B
80
,
024105
(
2009
).
27.
Y.
Kim
,
A.
Morozovska
,
E.
Eliseev
,
M. P.
Oxley
,
R.
Mishra
,
S. M.
Selbach
,
T.
Grande
,
S. T.
Pantelides
,
S. V.
Kalinin
, and
A. Y.
Borisevich
,
Nat. Mater.
13
,
879
(
2014
).
28.
S.
Stoughton
,
M.
Showak
,
Q.
Mao
,
P.
Koirala
,
D. A.
Hillsberry
,
S.
Sallis
,
L. F.
Kourkoutis
,
K.
Nguyen
,
L. F. J.
Piper
,
D. A.
Tenne
,
N. J.
Podraza
,
D. A.
Muller
,
C.
Adamo
, and
D. G.
Schlom
,
APL Mater.
1
,
042112
(
2013
).
29.
D. J.
Gregg
,
Z.
Zhang
,
G. J.
Thorogood
,
B. J.
Kennedy
,
J. A.
Kimpton
,
G. J.
Griffiths
,
P. R.
Guagliardo
,
G. R.
Lumpkin
, and
E. R.
Vance
,
J. Nucl. Mater.
452
,
474
(
2014
).
30.
D. P.
Leusink
,
F.
Coneri
,
M.
Hoek
,
S.
Turner
,
H.
Idrissi
,
G. V.
Tendeloo
, and
H.
Hilgenkamp
,
APL Mater.
2
,
032101
(
2014
).
31.
L.
Bovo
,
X.
Moya
,
D.
Prabhakaran
,
Y.-A.
Soh
,
A.
Boothroyd
,
N.
Mathur
,
G.
Aeppli
, and
S.
Bramwell
,
Nat. Commun.
5
,
3439
(
2014
).

Supplementary Material