Molybdenum oxide films were deposited on α-Al2O3 (0001) at 580 °C using MoO3 from a conventional molecular beam epitaxy Knudsen cell. A relatively smooth film (RMS roughness 1.1 nm) was deposited in 1 min at 580 °C using a Knudsen cell temperature of 620 °C; however, after 15 min deposition under these conditions, isolated islands (30–50 nm wide × 10–20 nm tall) develop that are stable to annealing at 600 °C for 60 min. XPS evidenced that the films are oxygen deficient with an average formula of MoO2.7. The authors infer that this oxygen deficiency is responsible for their thermal stability and may have significant effects on their catalytic and electronic properties. In contrast, stoichiometric MoO3 films deposited at 400 °C sublime completely during annealing at 600 °C.

MoO3 has attracted much attention due to its unique structural and electronic properties that are advantageous in many applications, such as catalysis,1–4 active elements in conductance-type gas sensors for CO, H2, methanol, and NH3,5 and cathode materials for lithium-ion batteries.6 The morphology of films and particles profoundly affects behavior in these applications; therefore, methods and conditions that allow one to precisely manipulate phase composition, defect density, and morphology are of great interest. Special attention has been given to the preparation of thin films and nanoparticles of various morphologies such as nanorods,5 microballs,7 nanoflowers,8 nanoribbons,9 and even forklike rods for self-assembly.10 For such a goal, a wide repertoire of preparation techniques exists including microwave hydrothermal growth,8 precipitation with hydrotreatment,9 electrodeposition,11 reactive sputtering,12 and molecular beam epitaxy (MBE).13 

MoO3 exists in one of the three polymorphic forms: thermodynamically stable orthorhombic α-phase MoO3, metastable monoclinic β-phase MoO3 that thermally converts to α-phase above 350 °C, and metastable hexagonal h-phase.12,13 The latter phase is rare, and most films reported to date consist of the α or β phase. α-MoO3 has a unique layered structure with 0.7-nm thick bilayers. Each bilayer is comprised of two sheets of distorted MoO6 octahedra. There is strong cohesion within the bilayers dominated by covalent and ionic bonding, but the bilayers are held together only by weak van der Waals forces. On the other hand, β-MoO3 forms a monoclinic structure comprising a 3D network of corner-sharing MoO6 octahedra, and van der Waals forces do not play any significant structural role.5,13

Because of its relatively high vapor pressure, MoO3 can be easily deposited via MBE using a conventional Knudsen cell. MoO3 sublimes readily at above 600 °C in vacuo; the gas phase molecules are thought to be mostly trimers (MoO3)3.14 Typically, Knudsen cell temperatures of 570–620 °C are used to achieve growth rates of 80–100 nm/h.13 The crystallinity of the resulting MoO3 film can be tuned by the substrate temperature. Films deposited below 150 °C are amorphous, and polycrystalline films are deposited at 200–350 °C. Films deposited below 200 °C consist mainly of β-MoO3, whereas at higher temperatures (>300 °C) only α-MoO3 is deposited.13 

The aim of this work was to prepare and characterize MoO3 films deposited on α-Al2O3 (0001) (c-plane sapphire) as model catalysts for oxidative dehydrogenation (ODH) of ethane. MoO3 supported on γ-Al2O3 is a well-known catalyst for ethane ODH.2,3,15–19 The ODH reaction at <650 °C is dominated by heterolytic C-H cleavage and follows a Mars-van Krevelen mechanism;20,21 however, a consensus regarding the active Mo species in supported MoO3 catalysts has not been reached. Planar catalysts present many advantages for the investigation of catalytic behavior when compared to powders. Structure and composition of the planar sample can be more easily controlled and probed; thus, they have the potential to provide a clearer and deeper understanding of catalytic processes from a mechanistic standpoint. Koike et al.13 grew MoO3 films on c-plane sapphire by MBE and reported that above 400 °C the sublimation becomes dominant and limits further film growth. The same conclusion was reached by Du et al.22 for films grown on SrTiO3. Because temperatures of interest for ethane ODH are above 500 °C, this obviously brings into question whether it is possible to grow MoOx films on sapphire that will be stable under reaction conditions. Our results demonstrate the successful preparation of nanostructured MoOx films on c-plane sapphire at 580 °C that are stable to annealing at 600 °C for 60 min.

A schematic of the experimental apparatus is shown in Fig. 1. It consists of four interconnected, independently pumped vacuum chambers isolated by UHV gate valves.

Fig. 1.

Schematic of thin film deposition and surface analysis cluster.

Fig. 1.

Schematic of thin film deposition and surface analysis cluster.

Close modal

A substrate is attached with Ag paste to an Mo sample holder and introduced via a load lock evacuated by a turbomolecular pumping station (Pffeifer HiCube 80 Eco) with a base pressure of 10−7 Torr. After the load lock is evacuated, the sample enters a transfer line evacuated to 1 × 10−8 Torr with a turbomolecular pump (Pfeiffer HiPace 300 M). The sample can be transferred between adjacent chambers via magnetically coupled transfer rods which allow in vacuo transfer for online angle-resolved x-ray photoelectron spectroscopy (ARXPS), Auger electron spectroscopy (AES), temperature-programmed desorption, and low-energy electron diffraction (LEED).

The ARXPS chamber is equipped with a PHI 3057 XPS system with a spherical capacitor analyzer (PHI 80-865A), a dual-anode x-ray source (PHI 04-548), a tilt stage, and an Ar+ ion gun for sample cleaning. The chamber is pumped by a Perkin-Elmer TNBX ion pump to a base pressure of 10−10 Torr. Al Kα radiation was used exclusively in this work. Multiplex scans were recorded using a 23.50-eV pass energy and a 0.1-eV step size. Binding energies are referred to the Al 2p peak at 74.1 eV.23–25 Using such a reference was necessary because no C 1s signal was detectable in the deposited films. The Al 2p reference used in this work is equivalent with reference to C 1s at 284.6 eV.24 In order to quantitatively estimate the Mo oxidation states present in the films, the Mo 3d spectral region was deconvoluted in CasaXPS software using a method reported by Baltrusaitis.26 Briefly, Mo 3d5/2 binding energies of 232.6, 231.0, and 229.2 eV for Mo oxidation states of +VI, +V, and +IV, respectively, were measured on the same instrument for similar samples, and these values were used as constraints in peak fitting using CasaXPS software. Binding energy constraints were allowed to vary within 0.1 eV to account for experimental error and the energy resolution of the instrument. Additional constraints were peak shape 90% Gaussian/10% Lorentzian, area ratio of Mo 3d5/2 and Mo 3d3/2 peaks equal 3/2, and peak splitting equal 3.15 eV. FWHM (found by fitting) were 2.0, 2.5, and 1.3 eV for Mo +VI, +V, and +IV, respectively.

The deposition chamber is equipped with a sample stage capable of rotational and translational motion with a small tilt adjustment. The sample holder is heated from backside by radiation from a pyrolytic boron nitride (PBN)-coated graphite heating element. Power is supplied by a low-voltage DC power supply (Sorensen DC 40-75) connected to the heating element via two copper electrical feedthroughs. The sample stage is equipped with a K-type thermocouple located in the proximity of the sample holder. The sample surface temperature (above 550 °C) can be measured directly by a pyrometer (Ultimax UX-20/600) through a window on the front of the chamber. The temperature of the substrate below 550 °C was estimated from thermocouple readings via a multipoint calibration established at higher temperatures using the pyrometer. A conventional Knudsen effusion cell (SVT Associates) was used as the MoO3 source. The cell is equipped with a 20-ml PBN crucible and a manual shutter for on/off control of the flux. The lip of the crucible is approximately 13 cm from the substrate, and the angle between the Knudsen cell axis and the substrate normal is 60°. Film morphology can be monitored in situ via reflection high-energy electron diffraction (RHEED) using a Fisons Instruments (LEG110) electron gun and phosphor screen. The chamber is pumped by a turbomolecular pump (Pfeiffer HiPace800) to a base pressure of 5 × 10−9 Torr. During in vacuo depositions, the pressure in the chamber was 3–7 × 10−8 Torr. The chamber can be backfilled with O2 via a UHV leak valve to 10−4 Torr (max.) as a reaction background or for sample annealing.

The LEED/AES chamber contains a single-pass cylindrical mirror analyzer with a coaxial electron gun for AES (PHI 10-155), an Ar sputter gun for depth profiling, and reverse-view LEED optics (PHI 11-020).

The Knudsen cell was loaded with MoO3 (99.95%, Alfa Aesar), and the cell temperature was maintained at 620 °C during deposition experiments. The sapphire substrate (1 × 1 cm2, AdValue Technology) was washed and sonicated in isopropanol for 15 min and then mounted on an Mo sample holder using Ag paste. Prior to MoO3 deposition, the sapphire substrate was outgassed and cleaned by heating at 700 °C in 5 × 10−6 Torr O2 ambient for 30 min. Subsequently, the substrate temperature was decreased to 400 or 580 °C, and deposition was initiated by opening the Knudsen cell shutter. Depositions were performed in vacuo (10−8 Torr) and in an O2 ambient (5 × 10−6 Torr) achieved by backfilling via a UHV leak valve. MoO3 was deposited for 1–15 min after which the shutter was closed, and sample heating was stopped. After cooling to ambient temperature, the sample was analyzed in situ using RHEED, transferred in vacuo for online XPS analysis, and then removed from the apparatus for ex situ atomic force microscopy (AFM). AFM measurements were conducted using a Digital Electronics Dimension 3000 scanning probe microscope using an Si tip in a tapping mode. Some samples also underwent annealing in an O2 ambient to test sample stability at elevated temperatures.

Figure 2 shows RHEED patterns and AFM images of molybdenum oxide films deposited in vacuo on α-Al2O3 (0001) at 580 °C. The sapphire substrate after in situ annealing at 700 °C shows an RHEED streak pattern characteristic of a clean flat single-crystal surface27,28 that was further evidenced by a featureless atomically smooth AFM image [Fig. 2(a)]. The RHEED pattern [Fig. 2(b)] for a film deposited in 1 min at 580 °C shows spots characteristic of a rough surface with 3D islands (asperities), the faint diagonal streaks that may arise from reflections of crystal facets on the side walls of the islands.29 The spotty RHEED pattern developed more fully at longer deposition times indicating relatively uniform coverage of nanoscale 3D islands on the surface. This was further evidenced by AFM. The film deposited in 1 min shows a relatively smooth and uniform surface (RMS roughness 1.1 nm) comprised of nanocrystallites. These grow to a larger size after a 5-min deposition (RMS roughness 3.02 nm) and finally form relatively uniform 3D islands (20–40 nm wide × 10–20 nm tall) with bare sapphire exposed in between (RMS roughness 4.75 nm) at 15 min deposition time. Figure 3 shows height profiles along the white lines shown in Fig. 2. It is apparent that an initially continuous relatively smooth film evolves over time into tall isolated granules with space between them that appears to be bare sapphire, although there might be present some small (<1 nm) features that cannot be resolved by AFM. Evidence of such features is visible in AFM height profile of sample deposited at 15 min although this could also be attributed to “flying tip” artifacts that are unavoidable with such a dramatic change in the slope of the landscape caused by the large islands.

Fig. 2.

Molybdenum oxide films deposited in vacuo on c-plane sapphire at 580 °C: RHEED patterns and (1 × 1) μm2 AFM morphology images. (a) Sapphire annealed at 700 °C for 60 min (RMS 172 pm), (b) 1 min deposition time (RMS 1.10 nm), (c) 5 min deposition time (RMS 3.02 nm), and (d) 15 min deposition time (RMS 4.75 nm).

Fig. 2.

Molybdenum oxide films deposited in vacuo on c-plane sapphire at 580 °C: RHEED patterns and (1 × 1) μm2 AFM morphology images. (a) Sapphire annealed at 700 °C for 60 min (RMS 172 pm), (b) 1 min deposition time (RMS 1.10 nm), (c) 5 min deposition time (RMS 3.02 nm), and (d) 15 min deposition time (RMS 4.75 nm).

Close modal
Fig. 3.

AFM height profiles along the white lines depicted in (a) Fig. 2 and (b) Fig. 5.

Fig. 3.

AFM height profiles along the white lines depicted in (a) Fig. 2 and (b) Fig. 5.

Close modal

At 580 °C, net film deposition results from the incident MoO3 flux and sublimation (desorption). The substrate receives the incident MoO3 flux, and at the same time, part of the deposited material sublimes due to the high vapor pressure of MoO3 at the substrate temperature.13,14 This growth regime together with the fact that our Knudsen cell is located at a 60° angle with respect to the sample normal may explain the unusual growth morphology of the films. Moreover, MoO3 reduces at 580 °C making the deposited films oxygen deficient (shown in Sec. III B). We speculate that tall isolated 3D islands result from the shadowing effect of the crystallites, as depicted schematically in Fig. 4. The film initially deposits in a relatively smooth layer which grows in time. As the roughness of the surface increases, the valleys no longer receive the incident MoO3 flux due to the shadowing effect of the roughness, but due to the high temperature material from the valley sublimes. This leads to a morphology of 3D islands separated by areas of the exposed substrate, as shown in Fig. 1(d) for the 15 min deposited sample. Shadowing is a relatively well-known phenomenon affecting thin film growth morphology.30,31

Fig. 4.

Possible effect of shadowing on MoOx film deposition at 580 °C.

Fig. 4.

Possible effect of shadowing on MoOx film deposition at 580 °C.

Close modal

In an attempt to deposit smoother films, the temperature of the substrate was lowered to 400 °C. Figures 5(a) and 5(b) show RHEED patterns and AFM images of a film deposited for 1 min in an O2 ambient (5 × 10−6 Torr O2) and after annealing in an oxygen ambient at 600 °C for 60 min. Deposition at 400 °C led to a thicker film with an RHEED pattern typical of polycrystalline materials.13,29 The AFM image reveals rather large MoO3 grains seemingly randomly oriented in multiple layers. Most strikingly, upon annealing at 600 °C in an oxygen ambient, the entire MoO3 film was lost due to sublimation. RHEED shows only streaky reflections of bare sapphire, and AFM revealed an atomically flat surface with no crystallites present. Thus, in agreement with conclusions made by Koike et al.,13 it is clear that MoO3 deposited at 400 °C is not stable at temperatures that are of interest (500–600 °C) for future ODH reaction testing.

Fig. 5.

Molybdenum oxide films deposited on c-plane sapphire at 400 °C for 1 min and at 580 °C for 15 min, both deposited in an O2 ambient 5 × 10−6 Torr. (a) 400 °C, 1 min (RMS 6.16 nm), (b) 400 °C, 1 min, annealed at 600 °C for 60 min (RMS 42.1 pm), (c) 580 °C, 15 min (RMS 3.30 nm), and (d) 580 °C, 15 min, annealed at 600 °C for 60 min (RMS 3.46 nm).

Fig. 5.

Molybdenum oxide films deposited on c-plane sapphire at 400 °C for 1 min and at 580 °C for 15 min, both deposited in an O2 ambient 5 × 10−6 Torr. (a) 400 °C, 1 min (RMS 6.16 nm), (b) 400 °C, 1 min, annealed at 600 °C for 60 min (RMS 42.1 pm), (c) 580 °C, 15 min (RMS 3.30 nm), and (d) 580 °C, 15 min, annealed at 600 °C for 60 min (RMS 3.46 nm).

Close modal

RHEED and AFM results for a film that was deposited in 15 min at 580 °C in an O2 ambient and then annealed at 600 °C for 60 min are shown in Figs. 5(c) and 5(d), respectively. The as-deposited film had the morphology of separated 3D islands with areas of the exposed substrate virtually identical with the in vacuo deposited films from Fig. 2; therefore, we infer that the effect of O2 ambient is not significant. Koike et al. found the effect of supplying oxygen radicals negligible to the film growth.13 More importantly, the film deposited at 580 °C withstood annealing at 600 °C [Fig. 5(d)] with qualitatively the same morphology to the as-deposited film. MoOx islands on the annealed sample are not as tall as in the as-deposited sample and have a very narrow distribution of sizes with almost all islands in the range of maximum height 10–15 nm, with 30–50 nm diameter at the base [Fig. 3(b)]. We conclude that MoOx films in the form of nanoscale 3D islands can be grown on c-plane sapphire at 580 °C and are thermostable at 600 °C. It is clear that when compared to films deposited at 400 °C, the nature of MoOx in samples deposited at 580 °C has to be somehow different, which provides them with increased thermal stability. We hypothesized that this is due to oxygen deficiency of the films deposited at 580 °C which is supported with an XPS evidence in Sec. III B.

XPS analysis of in vacuo deposited films and films deposited in an O2 ambient at 580 °C revealed that they are oxygen deficient (Table I, Fig. 6). The XPS binding energies were referenced to Al 2p at 74.1 eV. In all cases, the O 1s peak was located at 530.6 ± 0.1 eV, and the Mo 3d5/2 peaks for Mo+VI, Mo+V, and Mo+IV were located at 232.6, 231.0, and 229.2 eV (±0.1 eV), respectively, in good agreement with literature.26,32 Carbon was not detected in deposited (or annealed) films.

Fig. 6.

Mo 3d XPS of molybdenum oxide films on c-plane sapphire deposited at (a) 580 °C for 15 min and (b) 400 °C for 1 min. Mo 3d5/2 component represented by a full line and 3d3/2 component dashed in the CasaXPS deconvolutions.

Fig. 6.

Mo 3d XPS of molybdenum oxide films on c-plane sapphire deposited at (a) 580 °C for 15 min and (b) 400 °C for 1 min. Mo 3d5/2 component represented by a full line and 3d3/2 component dashed in the CasaXPS deconvolutions.

Close modal
Table I.

Summary of XPS results for MoOx films on c-plane sapphire.

Deposition conditionsMo oxidation states (%)
Mo+VIMo+VMo+IVAverage formula
1 min, 580 °C 55.3 28.6 16.2 MoO2.70 
5 min, 580 °C 52.7 29.9 17.4 MoO2.68 
15 min, 580 °C 50.3 32.9 16.7 MoO2.67 
1 min, 400 °C 100.0 — — MoO3 
Deposition conditionsMo oxidation states (%)
Mo+VIMo+VMo+IVAverage formula
1 min, 580 °C 55.3 28.6 16.2 MoO2.70 
5 min, 580 °C 52.7 29.9 17.4 MoO2.68 
15 min, 580 °C 50.3 32.9 16.7 MoO2.67 
1 min, 400 °C 100.0 — — MoO3 

The Mo 3d spectral region of a film deposited at 580 °C for 15 min in an O2 ambient is shown in Fig. 6(a). Besides expected oxidation state Mo+VI, the experimental envelope clearly contains also oxidation states Mo+V (32.9%) and Mo+IV (16.7%), which corresponds to the average formula MoO2.67. In contrast, the experimental envelope of the sample deposited at 400 °C contains only an Mo+VI component [Fig. 6(b)]. In order to evaluate if the reduction of Mo is uniform throughout the sample, ARXPS was performed on the films, and no observable gradient in the Mo oxidation state was found. It is known that vapor pressure of Mo oxides increases with the oxidation state;33 thus, the Mo reduction that takes place during deposition at 580 °C is most likely responsible for the increased thermal stability of the films. The reduction of MoO3 to Mo+V and Mo+IV oxides under a nitrogen atmosphere above 450 °C has been previously reported by Spevack and McIntyre,32 and evidence of this phenomena in vacuo has been reported by others as well,26,34–36 although usually as a result of electron beam heating. The reduction may be initiated by the creation of oxygen vacancies, e.g., in the vicinity of chemical impurities/residues from preparation.35 Oxygen deficiency of MoO3 has a profound effect on its catalytic and electronic properties. For instance, this may be undesired for catalytic application such as ethane ODH because the presence of lower Mo oxidation states increases catalytic activity but triggers ethane hydrogenolysis resulting in a substantial loss of selectivity.37 On the other hand, many electronic applications of MoO3 films rely solely on substoichiometric Mo oxide, which is an effective way of decreasing the otherwise wide bandgap (3–3.4 eV) of α-MoO3.38–42 

In addition, we performed XRD on two selected films [Figs. 5(a) and 5(c)] to probe the phase of MoO3 present at the surface. The data can be found in Fig. S1 in the supplementary material.43 The film deposited at 400 °C [Fig. 5(a)] shows reflections corresponding (0k0) planes of α-MoO3 only, which suggests that the film is iso-oriented in this direction. This is in good agreement with the literature.13 However, it should be noted that the intensity of those reflections is very low; therefore, it remains possible that reflections of other crystalline planes are present but below the detection limit of the instrument. In the case of the film deposited at 580 °C [Fig. 5(c)], no reflections besides those from sapphire were observed. As shown above, films deposited at 580 °C are oxygen deficient; therefore, the α-MoO3 phase is not expected to be present; also, the amount of material on the substrate is rather low.

MoOx films were deposited on c-plane sapphire at 580 °C via MBE using MoO3 from a conventional Knudsen cell. The film morphology evolves with deposition time (1–15 min) from a continuous film comprised of nanocrystallites to isolated nanoscale islands (20–40 nm wide by 10–20 nm tall) that are stable upon annealing at 600 °C for 60 min. We suggest that the film morphology evolves into 3D islands due to the shadowing effect of the developing grains (or islands) with concurrent MoO3 sublimation due to its high vapor pressure at 580 °C. We infer that films grown at 580 °C are thermally more stable due to their oxygen deficiency. The oxygen deficiency of the films was shown by XPS, and the Mo oxide can be described by the average formula of MoO2.67–2.70. This oxygen deficiency is expected to have a profound effect on catalytic and electronic properties that will be evaluated in follow-up studies. In contrast, films deposited at 400 °C are stoichiometric MoO3 and completely sublime during annealing at 600 °C.

This research was sponsored by the National Science Foundation (NSF) (No. CBET-1604605).

1.
P.
Novotný
,
S.
Yusuf
,
F.
Li
, and
H. H.
Lamb
,
Catal. Today
317
,
50
(
2018
).
2.
E.
Heracleous
,
F.
Lee
,
I.
Vasalos
, and
A. A.
Lemonidou
,
Catal. Lett.
88
,
47
(
2003
).
3.
A.
Christodoulakis
,
E.
Heracleous
,
A. A.
Lemonidou
, and
S.
Boghosian
,
J. Catal.
242
,
16
(
2006
).
4.
Y. V.
Plyuto
,
I. V.
Babich
,
I. V.
Plyuto
,
A. D.
Van Langeveld
, and
J. A.
Moulijn
,
Appl. Surf. Sci.
119
,
11
(
1997
).
5.
E.
Comini
,
L.
Yubao
,
Y.
Brando
, and
G.
Sberveglieri
,
Chem. Phys. Lett.
407
,
368
(
2005
).
6.
L.
Zhou
,
L.
Yang
,
P.
Yuan
,
J.
Zou
,
Y.
Wu
, and
C.
Yu
,
J. Phys. Chem. C
114
,
21868
(
2010
).
7.
Q. P.
Ding
,
H. B.
Huang
,
J. H.
Duan
,
J. F.
Gong
,
S. G.
Yang
,
X. N.
Zhao
, and
Y. W.
Du
,
J. Cryst. Growth
294
,
304
(
2006
).
8.
G.
Wei
,
W.
Qin
,
D.
Zhang
,
G.
Wang
,
R.
Kim
,
K.
Zheng
, and
L.
Wang
,
J. Alloys Compd.
481
,
417
(
2009
).
9.
X. W.
Lou
and
H. C.
Zeng
,
Chem. Mater.
14
,
4781
(
2002
).
10.
X. W.
Lou
and
H. C.
Zeng
,
J. Am. Chem. Soc.
125
,
2697
(
2003
).
11.
D.
Di Yao
,
J. Z.
Ou
,
K.
Latham
,
S.
Zhuiykov
,
A. P.
O’Mullane
, and
K.
Kalantarzadeh
,
Cryst. Growth Des.
12
,
1865
(
2012
).
12.
P. F.
Carcia
and
E. M.
McCarron
,
Thin Solid Films
155
,
53
(
1987
).
13.
K.
Koike
,
R.
Wada
,
S.
Yagi
,
Y.
Harada
,
S.
Sasa
, and
M.
Yano
,
Jpn. J. Appl. Phys.
53
,
3
(
2014
).
14.
E. A.
Gulbransen
,
K. F.
Andrew
, and
F. A.
Brassart
,
J. Electrochem. Soc.
110
,
242
(
1963
).
15.
M. V.
Martínez-Huerta
,
X.
Gao
,
H.
Tian
,
I. E.
Wachs
,
J. L. G.
Fierro
, and
M. A.
Bañares
,
Catal. Today
118
,
279
(
2006
).
16.
T.
Blasco
,
A.
Galli
,
J. M.
López Nieto
, and
F.
Trifiró
,
J. Catal.
169
,
203
(
1997
).
17.
J.
Le Bars
,
A.
Auroux
,
M.
Forissier
, and
J. C.
Vedrine
,
J. Catal.
162
,
250
(
1996
).
18.
K.
Chen
,
A. T.
Bell
, and
E.
Iglesia
,
J. Phys. Chem. B
104
,
1292
(
2000
).
19.
B.
Solsona
,
A.
Dejoz
,
T.
Garcia
,
P.
Concepción
,
J. M. L.
Nieto
,
M. I.
Vázquez
, and
M. T.
Navarro
,
Catal. Today
117
,
228
(
2006
).
20.
P.
Mars
and
D. W.
van Krevelen
,
Chem. Eng. Sci.
3
,
41
(
1954
).
21.
C. A.
Gärtner
,
A. C.
van Veen
, and
J. A.
Lercher
,
ChemCatChem
5
,
3196
(
2013
).
22.
24.
J. A.
Rotole
and
P. M. A.
Sherwood
,
Surf. Sci. Spectra
5
,
11
(
1998
).
25.
P. A.
Spevack
and
N. S.
McIntyre
,
J. Phys. Chem.
97
,
11031
(
1993
).
26.
J.
Baltrusaitis
,
B.
Mendoza-Sanchez
,
V.
Fernandez
,
R.
Veenstra
,
N.
Dukstiene
,
A.
Roberts
, and
N.
Fairley
,
Appl. Surf. Sci.
326
,
151
(
2015
).
27.
W.
Wang
,
W.
Yang
,
Z.
Liu
,
Y.
Lin
,
S.
Zhou
,
H.
Qian
,
H.
Wang
,
Z.
Lin
, and
G.
Li
,
CrystEngComm
16
,
7626
(
2014
).
28.
S.
Bera
,
Y.
Sumiyoshi
, and
Y.
Yamada-Takamura
,
J. Appl. Phys.
106
,
63531
(
2009
).
29.
S.
Hasegawa
, in
Characterization of Materials
, 2nd ed., edited by
E. N.
Kaufmann
(
Wiley
, New York,
2012
), p.
1925
.
30.
M.
Pelliccione
,
T.
Karabacak
,
C.
Gaire
,
G.-C.
Wang
, and
T.-M.
Lu
,
Phys. Rev. B
74
,
125420
(
2006
).
31.
J. T.
Drotar
,
Y.-P.
Zhao
,
T.-M.
Lu
, and
G.-C.
Wang
,
Phys. Rev. B
62
,
2118
(
2000
).
32.
P. A.
Spevack
and
N. S.
McIntyre
,
J. Phys. Chem.
96
,
9029
(
1992
).
33.
P. E.
Blackburn
,
M.
Hoch
, and
H. L.
Johnston
,
J. Phys. Chem.
62
,
769
(
1958
).
34.
L. A.
Bursill
,
W. C. T.
Dowell
,
P.
Goodman
, and
N.
Tate
,
Acta Crystallogr. A
34
,
296
(
1978
).
35.
M.
Łabanowska
,
Phys. Chem. Chem. Phys.
1
,
5385
(
1999
).
36.
L. A.
Bursill
and
A. J.
Stuart
,
Proc. R. Soc. London. A. Math. Phys. Sci.
311
,
267
(
1969
).
37.
A.
Hanif
,
T.
Xiao
,
A. P. E.
York
,
J.
Sloan
, and
M. L. H.
Green
,
Chem. Mater.
14
,
1009
(
2002
).
38.
K.
Inzani
,
M.
Nematollahi
,
F.
Vullum-Bruer
,
T.
Grande
,
T. W.
Reenaas
, and
S. M.
Selbach
,
Phys. Chem. Chem. Phys.
19
,
9232
(
2017
).
39.
H.-S.
Kim
,
J. B.
Cook
,
H.
Lin
,
J. S.
Ko
,
S. H.
Tolbert
,
V.
Ozolins
, and
B.
Dunn
,
Nat. Mater.
16
,
454
(
2016
).
40.
C.
Battaglia
 et al.,
Nano Lett.
14
,
967
(
2014
).
41.
J. J.
Jasieniak
,
J.
Seifter
,
J.
Jo
,
T.
Mates
, and
A. J.
Heeger
,
Adv. Funct. Mater.
22
,
2594
(
2012
).
42.
H.
Simchi
,
B. E.
McCandless
,
T.
Meng
,
J. H.
Boyle
, and
W. N.
Shafarman
,
J. Appl. Phys.
114
,
13503
(
2013
).
43.
See supplementary material at https://doi.org/10.1116/1.5100752 for XRD profiles of selected samples.

Supplementary Material