Decoupling the Effects of High Crystallinity and Surface Area on the Photocatalytic Overall Water Splitting over Β-Ga2O3 Nanoparticles by Chemical Vapor Synthesis

Article abstract of DOI:10.1002/cssc.201701309

Chemical vapor synthesis (CVS) is a unique method to prepare well-defined photocatalyst materials with both large specific surface area and a high degree of crystallinity. The obtained β-Ga₂O₃ nanoparticles were optimized for photocatalysis by reductive photodeposition of the Rh/CrO_x co-catalyst system. The influence of the degree of crystallinity and the specific surface area on photocatalytic aqueous methanol reforming and overall water splitting (OWS) was investigated by synthesizing β-Ga₂O₃ samples in the temperature range from 1000 °C to 1500 °C. With increasing temperature, the specific surface area and the microstrain were found to decrease, whereas the degree of crystallinity and the crystallite size increased. Whereas the photocatalyst with the highest specific surface area showed the highest aqueous methanol reforming activity, the highest OWS activity was that for the sample with an optimum ratio between high degree of crystallinity and specific surface area. Thus, it was possible to show that the facile aqueous methanol reforming and the demanding OWS have different requirements for high photocatalytic activity.

Full text of DOI:10.1002/cssc.201701309

Accepted Article
Title: Decoupling the Effects of High Crystallinity and Surface Area
on the Photocatalytic Overall Water Splitting over β-Ga2O3
Nanoparticles by Chemical Vapor Synthesis
Authors: Sasa Lukic, Jasper Menze, Philipp Weide, Wilma Busser,
Markus Winterer, and Martin Muhler
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201701309
Link to VoR: http://dx.doi.org/10.1002/cssc.201701309
A Journal of
www.chemsuschem.org

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
Decoupling the Effects of High Crystallinity and Surface Area on
the Photocatalytic Overall Water Splitting over β-Ga2O3
Nanoparticles by Chemical Vapor Synthesis
Sasa Lukic,[a] Jasper Menze,[b] Philipp Weide,[b] G. Wilma Busser,[b]
Markus Winterer*,[a] Martin Muhler,*[b]
Abstract: Chemical vapor synthesis (CVS) is a unique method to
prepare well-defined photocatalyst materials with both large specific
surface area and a high degree of crystallinity. The obtained β-
Ga2O3 nanoparticles were optimized for photocatalysis by reductive
photodeposition of the Rh/CrOx cocatalyst system. The influence of
the degree of crystallinity and the specific surface area on
photocatalytic aqueous methanol reforming and overall water
splitting (OWS) was investigated by synthesizing β-Ga2O3 samples
in the temperature range from 1000˚C to 1500˚C. With increasing
temperature, the specific surface area and the microstrain were
found to decrease, whereas the degree of crystallinity and the
crystallite size increased. While the photocatalyst with the highest
specific surface area showed the highest aqueous methanol
reforming activity, the highest OWS activity was observed for the
sample with an optimum ratio between high degree of crystallinity
and specific surface area. Thus, it was possible to show that the
facile aqueous methanol reforming and the demanding OWS have
different requirements for high photocatalytic activity.
UV light to evolve H2/O2 simultaneously in a stoichiometric ratio
from H2O when loaded with suitable cocatalysts. Ga2O3 is a
highly suitable model catalyst to study the dependence of the
photocatalytic activity on properties such as phase composition,
degree of crystallinity, and specific surface area of the particles.
[4]
Ga2O3 can also be used to develop and optimize new co-
catalyst systems. A wide range of metals and metal oxides has
been reported as cocatalyst for H2 evolution, including Pt, Rh,
Au, Ni and CuOx.[3, 5-9] Mainly metal oxides such as RuOx, IrOx
and CuOx have been applied for the oxygen evolution.[3,5,8-11]
A
cocatalyst system showing an outstanding overall water splitting
(OWS) activity is the Rh/CrOx system. Domen and co-workers
ascribed the high efficiency of the Rh/CrOx cocatalyst to the
formation of a core/shell structure containing a noble metal Rh
core and a CrOx (most likely Cr2O3) shell,[12,13] which is assumed
to be permeable to protons and H2, but impermeable to O2.
Arranged in this way, the back reaction of H2 and O2 on the
noble metal would be inhibited. Busser et al.[14] showed that the
formation of a CrOx shell is rather unlikely in a CuOy-containing
cocatalyst. They assumed that CrOx enables the forward
reaction by acting as an O2 evolution site rather than
suppressing the back reaction. Recently, this role of CrOx was
supported by Soldat et al.,[15] who observed OWS over
Ba5Ta4O15 and Ga2O3 using only CrOx as cocatalyst, in this way
confirming the assumption that CrOx mainly provides sites for
the oxygen evolution reaction (OER). The Rh/CrOx cocatalyst
can be deposited by impregnation[16] followed by subsequent
thermal decomposition of the precursor or by in situ
photodeposition.[17] Methanol is used as a sacrificial agent in the
reductive photodeposition acting as electron donor. The
advantage is that the cocatalyst particles are reductively
deposited on those sites, which are most accessible for the
generated photoelectrons. These sites should be the optimum
position for the hydrogen evolution reaction (HER). The H2
evolution rate during reductive photodeposition directly
correlates with the number of H2 evolution sites allowing a quick
and efficient determination of the optimum cocatalyst loading.
Simultaneously, methanol is oxidized to CO2 by holes in the
valence band.
Introduction
One of the key scientific challenges in the 21st century is direct
overall water splitting (OWS) using solar irradiation yielding
hydrogen and oxygen. It would be an ideal way to produce clean
and renewable energy on a large scale, for which powder
photocatalysts applied in slurry reactors are highly suitable.[1]
The highest quantum yields for single particle photocatalysts
were reported for metal oxides.[2] The most promising materials
involve cationic species of d0 or d10 electronic configuration such
as Ga3+.[3] Ga2O3 has a large band gap of about 4.8 eV requiring
[a]
[b]
S. Lukic, Prof. Dr. M. Winterer
Nanoparticle Process Technology (NPPT) & CENIDE
University of Duisburg-Essen
47057 Duisburg (Germany)
J. Menze, P. Weide, G. W. Busser, Prof. Dr. M. Muhler
Laboratory of Industrial Chemistry
Ruhr-University Bochum
44780 Bochum (Germany)
E-mail: muhler@techem.rub.de
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
In addition to the optimum cocatalyst loading, the OWS activity
is determined by other factors such as the specific surface and
the degree of crystallinity, which depend on the growth rate
relative to the rate by which the ions find their ideal lattice
position. Defects may act as charge carrier recombination sites
that reduce the concentration of free charge carriers available
for the HER and OER, thus lowering the OWS activity. Hence, a
high degree of crystallinity enhances the number of charge
carriers reaching the surface of the semiconductor nanoparticles.
The influence of the specific surface is discussed controversially.
A high surface area leads to an increased number of HER and
OER sites, but also to an increasing number of surface defect
sites.[18] A high OWS activity would therefore be expected for
materials having both a high degree of crystallinity and a high
specific surface area. Nanostructuring of photocatalysts has
both advantages and disadvantages regarding the OWS activity.
Nanoscaling has been proven to enhance charge carrier
extraction for materials with low carrier mobility and short carrier
lifetime. At the same time, the ability of an absorber to generate
chemical energy from solar energy is lowered.[18] In this work we
exploit the advantages of chemical vapor synthesis (CVS),
which is a modification of chemical vapor deposition (CVD),
where process parameters are adjusted such that nanoparticles
are formed in the gas phase instead of forming films.[19] Due to
the synthesis in the air-borne state, high temperatures can be
applied to eliminate defects, which would lead to strong
agglomeration and sintering in conventional synthetic
approaches. Ga2O3 nanoparticles were synthesized and by
adjusting the process parameters of CVS like temperature and
pressure, we tuned the characteristics of the particles aiming
simultaneously at high purity, crystallinity and surface area.
decisive. Therefore, crystallites at higher temperatures
increased in size from 9 nm to 31 nm, while microstrain
decreased from 0.00059 to 0.0016 (Figure 1b).
Figure 1. a) X-Ray diffraction patterns (all diffractograms are normalized to the
individual maximum intensity for graphical clarity) and b) crystallite size and
microstrain determined by Rietveld refinement of the synthesized samples as
function of the reaction temperature in the system.
Results and Discussion
Since the specific surface area and the degree of crystallinity are
closely correlated to crystal growth and sintering, such a relation
is illustrated in Figure 2a. Within the considered temperature
range from 1000°C to 1500°C, the specific surface area
decreased from 120 m2 g-1 to 41 m2 g-1, whereas the degree of
crystallinity increased from 31% to 85%. CVS samples have up
to 10 times larger specific surfaces areas when compared with a
commercial Ga2O3 sample (Chempur, 10 m2 g-1). Based on the
structural properties the optimum synthesis temperature was
found to be in the range from 1100°C to 1200°C for sample 4
Properties of CVS Ga2O3 samples synthesized at different
temperatures
Figure 1a shows the diffracton patterns of samples synthesized
at different temperatures. Line broadening is generally observed
due to small crystallite sizes, microstrain and stacking and
intrinsic faults as well as dislocations and point defects.[20,21] At
higher temperatures growth defects are annealed during the
synthesis by fast solid-state and surface diffusion processes,[22]
the sintering kinetic is enhanced and coalescence becomes
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
characterized by both a high degree of crystallinity as well as a
high specific surface area as summarized in Table 1. Both
properties have a large influence on the photocatalytic activity.
The refined XRD pattern of sample 4 confirms the presence of
single monoclinc β-Ga2O3 phase with Ga3+ ions in distorted
octahedral and tetrahedral sites and O2- ions in a distorted cubic
close packing arrangement with Ga-O bond distances between
1.83 Å and 2.0 Å (Figure 2b). As shown in Table 1, the
experimentally determined lattice parameters by Rietveld
refinement are in a good agreement with literature values.[23]
Figure 3. a) UV-Vis diffuse reflectance spectrum and b) bandgap energy
determined by the Kubelka-Munk function for sample 4 obtained by CVS at
1200°C and 300 mbar.
Figure 4 shows high-resolution TEM images. It is clearly
observed that the nanoparticles present in sample 4 are highly
crystalline, supporting the calculated value derived from XRD for
the degree of crystallinity of 78%. Particles of commercial Ga2O3
powder (Chempur) have a rod-like shape with a length of
several μm, while those obtained by CVS have a more or less
round shape with a size of a few nanometers.
Figure 2. a) Crystallinity and specific surface area as function of the reaction
temperature. b) Rietveld refinement of the XRD pattern of sample
synthesized at 1200°C and 300 mbar.
4
Bandgap energies of the Ga2O3 samples obtained by CVS at
temperatures between 1000°C and 1500°C were derived from
the Kubelka-Munk function [F(R∞)·hν]2 calculated from the
reflectance spectra R vs. hν plot. An example is shown in Figure
3. The bandgap energies were found to be within the range of
4.6-4.9 eV, consistent with values reported for bulk β−Ga2O3.[24]
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
Figure 5. HRTEM of irregularities in the planar stacking sequence of atoms in
the crystal, like twin or intrinsic faults of sample 4 (red arrows), while some
surface terminations look amorphous (yellow arrows).
Figure 4. HRTEM of sample 4 obtained by CVS at 1200°C and 300 mbar with
highly crystalline nanoparticles (top) and commercial Ga2O3 powder
(Chempur) with rod-like particles (bottom).
However, some particles surfaces contain irregularities due to
changes in the planar stacking sequence of atoms in the crystal,
like stacking or twin faults. Some of the imperfections in the
crystal structures are visible in Figure 5.
Figure 6. EELS measurements of sample 4 and 10 synthesized under exactly
the same conditions (1200°C, 300 mbar).
Samples 4 and 10 have very similar crystallite sizes of 22 nm
and 23 mm with degrees of crystallinity of 77% and 73%,
respectively, and the same specific surface area of 62 m2 g-1
indicating a high degree of reproducibility. Thus, it is possible to
synthesize samples by CVS with almost identical particle
characteristics.
SEM/EDX analysis of sample 4 (not shown) indicated that traces
of copper might be present, which would influence their
photocatalytic activity. The synthesis of sample 4, which showed
the best photocatalytic performance, was repeated (sample 10)
using exactly the same process parameters as shown in Table 1.
Detailed EELS analysis was performed to investigate the
possible presence of copper in both samples prepared on Au
grids in order to check reproducibility. The O-K and Ga-L edges
of both samples were found to have the same structure and
relative intensity, suggesting that they have the same
composition and electronic structure (Figure 6). The Cu-L edge
was below the detection limit for both samples, and Cu was not
identified in other parts of the TEM samples, either.
Photocatalytic activity of CVS samples
The CVS Ga2O3 samples obtained at different temperatures
were modified by consecutively performed reductive
photodeposition of the binary Rh(0.1 wt.%)/CrOx(0.09 wt.%)
cocatalyst system by aqueous methanol reforming:
CH3OH + H2O → CO2 + 3 H2
(1)
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
which seems to be important for a high methanol reforming
activity as further shown in Figure 8. An increase in specific
surface area may affect the number of HER sites or increase the
adsorption capacity of methanol. For the HER in an aqueous
methanol solution the degree of crystallinity does not seem to be
the decisive factor. Obviously, mainly the specific surface is
relevant for the methanol reforming activity of CVS Ga2O3. The
effect of an increased hydrogen evolution during aqueous
methanol reforming due to a larger specific surface area was
also demonstrated by Grewe and Tüysüz [25] for tantalate-based
catalysts and for TiO2.
Figure 8. H2 evolution after reductive photodeposition of the Rh/CrOx
cocatalyst by aqueous methanol reforming in dependence of the specific
surface area (the solid line is intended as a guide to the eye).
Figure 7. H2 evolution (top) and CO2 evolution (bottom) during reductive
photodeposition of the Rh/CrOx cocatalyst by aqueous methanol reforming on
β-Ga2O3 nanoparticles obtained by CVS at different temperatures.
The deposition of CrOx in the first step does not increase the
methanol reforming activity. In case of the samples obtained at
1000°C, 1300°C and 1500°C the activity even decreases slightly
by deposition of CrOx. The reductive deposition of CrOx occurs
at surface sites where preferentially electrons accumulate, i.e.,
on sites active for the HER. Since CrOx does not increase the
HER, the deposition of CrOx seems to block HER sites resulting
in a decreased methanol reforming activity. However, the
additional deposition of Rh in the second deposition step
increased the activity significantly demonstrating the ability of Rh
to enhance the HER. The highest methanol reforming activity
was observed for the sample obtained at 1000°C with H2
evolution rates of 15.3 mmol h-1 (pure Ga2O3), 14.6 mmol h-1
(CrOx/Ga2O3) and 25.6 mmol h-1 (Rh/CrOx/Ga2O3) in the three
photodeposition steps. The subsequent gradual decrease as a
function of time on stream has also been observed previously
The observed rates of H2 and CO2 evolution during the
sequential photodeposition procedure are shown in Figure 7. H2
evolution was observed immediately after the lamp had been
switched on in each deposition step showing that Ga2O3 has
catalytically active sites for the HER. Observation of CO2
evolution is temporally delayed as a result of the higher CO2
solubility in water and the adsorption of the different methanol
oxidation products on the Ga2O3 nanoparticles. All CVS Ga2O3
samples show aqueous methanol reforming activity without
cocatalysts deposited when irradiated with a 500 W Hg lamp. An
increase in the CVS temperature clearly results in a decrease in
aqueous methanol reforming activity. This correlation is
assumed to be related to the decrease in specific surface area,
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
during photocatalytic methanol reforming[14, 26] suggesting photo-
degradation of the Rh/CrOx sites due to strongly adsorbing
organic intermediates. The relatively slow increase in CO2
evolution indicates that the formation of CO2 occurs in
consecutive reactions: first methanol is oxidized to intermediates
like formates, which are subsequently oxidized to CO2.[26]
importance of the surface area for the methanol reforming
reaction, since the specific surface area of commercial Ga2O3 is
very low (10 m2 g-1). The OWS activity is comparable for both
materials. Here, it should be mentioned that the applied low
cocatalyst loadings of only 0.1 wt.% Rh and 0.09 wt.% CrOx
have been optimized for commercial Ga2O3 with a low specific
surface area (10 m² g-1). Hence the activity of the CVS Ga2O3
with significantly higher specific surface areas may be further
enhanced by increasing the cocatalyst loading.[14]
The OWS measurements were performed after filtration,
washing and freeze-drying of the cocatalyst-modified Ga2O3
samples. The results of the OWS measurements are shown in
Figure 9. In the first 60 min of irradiation CO2 evolution was
observed for all catalysts, which can be attributed to methanol
reforming of adsorbed methanol residues that cannot be
removed by washing and drying. Due to this reforming, the initial
H2 evolution rate was relatively high compared with the oxygen
evolution rate. The highest H2 and O2 evolution rates were
observed for the sample obtained at 1200°C showing an initial
H2 and O2 evolution rate of 4.9 mmol h-1 and 2.1 mmol h-1,
respectively.
The
decreasing
water
splitting
activity
characterized by the decrease in O2 evolution rate in the first two
hours is probably caused by photocorrosion of the photocatalyst
involving processes such as particle growth and leaching of the
cocatalyst. After 2 h TOS the photocatalytic activity did not
further decrease and almost stationary H2 and O2 evolution rates
of 2.06 mmol h-1 and 0.86 mmol h-1 were detected, respectively.
The almost stoichiometric H2/O2 ratio of 2:1 is retained for all
samples initially as well as after 5 h TOS indicating OWS to
proceed. While mainly a high surface is required for a high
methanol reforming activity, both the degree of crystallinity and
the specific surface seem to determine the OWS activity, as the
sample obtained at 1200°C has a high degree of crystallinity as
well as a high surface area. These properties result in high
charge carrier separation efficiency and in a high number of
HER and OER sites leading to a high OWS activity. The
relationship between crystallinity, surface area, charge
recombination and the way how these parameters affect overall
photocatalytic activity have been described by Kudo et al.[5]
Additionally, it has also been pointed out by Domen et al.[27] that
charge carrier recombination and, correspondingly, crystallinity
is an important issue in demanding uphill reactions such as
overall water splitting.
Figure 9. H2 and O2 temporal evolution curves for sample 4 (top) and initial
and long-term (after 5 h) evolution rate of the CVS samples obtained at
different temperatures between 1000°C and 1500 °C (bottom).
An apparent quantum efficiency of about 27% (+/- 3%) for OWS
was estimated for the most active sample using the method
described in ref. [17]. This value is in excellent agreement with
the apparent quantum yield reported for pure Ga2O3 by Sakata
The observed methanol reforming activities of photocatalysts
based on CVS-Ga2O3 are up to twice as high compared with
photocatalysts based on commercial Ga2O3 containing the same
cocatalyst and loading.[26] This finding again underlines the
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
et al.[28]. They show that pure Ga2O3 optimized with Rh0.5Cr1.5O3
as a cocatalyst shows an apparent quantum yield of 24%, which
can be further increased to 71% by doping with 3 mol% Zn and
preparation in a diluted CaCl2 solution.
photocatalytic activity. The obtained materials had a high degree
of crystallinity as well as specific surface areas up to more than
120 m² g-1. With increasing CVS temperature, the specific
surface area and the microstrain decreased, while the degree of
crystallinity and crystallite size increased. The methanol
reforming activity observed during reductive cocatalyst
deposition decreased as a function of CVS temperature, which
is ascribed mainly to the lower specific surface area. The degree
of crystallinity seems to have a minor influence on the facile H2
evolution from methanol. When loaded with the Rh/CrOx
cocatalyst all samples evolved H2 and O2 in the stoichiometric
ratio of 2:1. OWS experiments showed an increasing OWS
activity with increasing CVS temperature up to 1200°C, but for
higher temperatures the OWS activity decreased. Thus, the
photocatalytic activity of the CVS samples in the demanding
OWS is demonstrated to depend on the optimum ratio of
crystallinity and surface area. The highest OWS activity was
observed for the Ga2O3 sample synthesized at 300 mbar and
1200°C, which has both a high specific surface and a high
degree of crystallinity.
In addition to the CVS temperature, the applied pressure
strongly influences the characteristics of the CVS Ga2O3
samples and, therefore, their photocatalytic OWS activity. Based
on these observations it is possible to create a combined map
correlating structural characteristics of Ga2O3 samples and
process parameter fields with photocatalytic OWS performance
as shown in Figure 10.
Experimental Section
CVS setup and synthesis of Ga2O3 nanoparticles
A CVS system is mainly a combination of four modules: a gas supply
(gases used: He (Air Liquide, Alfagaz 1, 99.999%) and O2 (Air Liquide,
Purity 5.1, 99.9991%)) with precursor delivery system, a hot wall reactor,
a thermophoretic particle collector and a pressure control unit as shown
in Figure 11.
Figure 10. Combined map correlating structural characteristics of Ga2O3
samples and CVS process parameter fields with photocatalytic OWS
performance.
By decreasing the pressure we strongly influence the residence
times in the reaction zone. Below 40 mbar the residence time is
not sufficient for decomposing the precursor and no reaction
occurred. In general, at lower pressures samples are observed
with lower degrees of crystallinity and higher specific surface
areas. In this way, it is demonstrated that CVS is an excellent
method for the preparation of materials with well-defined particle
characteristics.
The precursor delivery unit consists of an alumina tube and a heating
system to control the evaporation rate of the precursor in a temperature
range of 195-230ºC. The precursor used for the synthesis of Ga2O3
nanoparticles was Ga(III)acetylacetonate (99.99% STREM, Germany). It
was placed into a boat, made of MACOR, a machineable glass-ceramic,
stable up to a temperature of 1000°C with very little thermal expansion.
The precursor vapor was carried by 600 sccm of helium to the reaction
zone and reacted with 1700 sccm of oxygen in a reaction tube with a
length of 840 mm heated by a tube furnace (STF Carbolite Furnaces) in
a temperature range of 1000-1500ºC. Oxygen was added just before the
reactor to prevent preliminary oxidation or combustion of the precursor.
The decomposition of Ga(III)acetylacetonate and the formation of Ga2O3
during the reaction with oxygen can be described as
Conclusions
We investigated the influence of the CVS process parameters
such as system pressure, precursor partial pressure and in
particular temperature on β-Ga2O3 particle properties and
2 Ga(O2C5H7)3 + 36 O2 → Ga2O3 + 30 CO2 + 21 H2O
(2)
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
Table 1. CVS process parameters and particle characteristics.
Process parameters
Particle characteristics
Sample
T (°C)
1200
1200
1200
1200
1000
1100
1300
1400
1500
1200
P (mbar)
40
CS (nm)
5
MS
Г (%)
SSA (m2 g-1)
LP (Å)
1
2
2.10∙10-3
4.78∙10-4
2.42∙10-4
2.10 ∙10-4
5.80∙10-4
2.14∙10-4
2.51∙10-4
1.52∙10-4
1.06∙10-4
2.30 ∙10-4
10
45
48
78
31
71
80
84
85
73
/
a = 12.426; b = 2.995; c = 5.781
100
7.5
12
108
92
62
120
62
45
45
41
62
a = 12.276; b = 3.023; c = 5.815
a = 12.219; b = 3.018; c = 5.781
a = 12.258; b = 3.039; c = 5.816
a = 12.359; b = 3.041; c = 5.856
a = 12.267; b = 3.040; c = 5.826
a = 12.285; b = 3.049; c = 5.832
a = 12.230; b = 3.040; c = 5.800
a = 12.257; b = 3.043; c = 5.818
a = 12.255; b = 3.039; c = 5.815
a = 12.232; b = 3.041; c = 5.801
3
150
4
300
22
5
300
9
6
300
17
7
300
25
8
300
30.5
31
9
300
10
300
23
Bulk literature values of the lattice parameters for β-Ga2O3 by Geller et.al.[23]
*CS – crystallite size, MS – microstrain, Г – degree of crystallinity, SSA – specific surface area, LP – lattice parameters
The synthesized nanoparticles were collected in
a
thermophoretic
online 3-channel gas analyzer (Emerson) containing detectors for H2
(thermal conductivity detector, detection limit 100 ppm), O2
(paramagnetic detector, detection limit 100 ppm) and CO2 (IR detector,
detection limit 10 ppm). The apparatus and procedure were described in
detail before.[26] N2 (99.9999 %, Air Liquide) as inert carrier and purge
gas was introduced by a PTFE tube placed at the bottom of the reactor.
The binary Rh/CrOx cocatalyst system was consecutively photodeposited
reductively from aqueous solutions of Na3RhCl6 ∙ 2H2O (99 %, Sigma
Aldrich) and K2CrO3 (≥ 98%, Sigma Aldrich). 300 mg of the CVS Ga2O3
sample was ultrasonicated in 50 ml H2O for 10 min and then suspended
in an aqueous solution containing approximately 8 vol.% of methanol (40
ml MeOH (Sigma-Aldrich, purity p.a. (>99.8%)) in 460 ml total H2O
(deionized) as sacrificial agent for the oxidation reaction. The suspension
was homogeneously stirred by a magnetic stirrer and purged with 96
sccm N2 for approximately 30 min to remove residual H2, O2 and CO2.
The Hg lamp was switched on without adding cocatalyst solution to
investigate the activity of pure Ga2O3 towards aqueous methanol
reforming. During irradiation, the N2 flow was adjusted to 50 sccm. After
the H2 evolution had reached steady state, the lamp was switched off.
The reactor was purged with N2 after each irradiation step and addition of
metal salt solution. 5 ml K2CrO4 solution was added and irradiated for 45
min, resulting in a nominal CrOx loading of 0.09 wt.% Cr based on the
metal. After CrOx deposition, 5 ml Na3RhCl6 solution was added and
irradiated for 2 h, resulting in a nominal Rh loading of 0.1 wt.%. The
solution was filtrated and the filter cake was washed thoroughly with
distilled water to remove residual metal salt solution and methanol before
drying overnight using a freeze-dryer. OWS experiments were carried out
using 150 mg of each Ga2O3 sample synthesized by CVS and modified
by cocatalysts. The catalyst was suspended in 50 ml distilled water,
ultrasonicated for 10 min, transferred to the reactor and 450 ml distilled
water was added. After the suspension had been homogeneously stirred
by a magnetic stirrer and purged with 96 sccm N2, the lamp was switched
on for 5 h and the evolved gas products in the 50 sccm N2 carrier gas
flow were monitored by the online gas analyzer. Commercial Ga2O3 for
comparison measurements was obtained from ChemPUR.
collector where they were separated from the gas phase and gaseous
byproducts by a temperature gradient established by quartz lamps in the
channels of the collector and water-cooled stainless steel walls. Low
pressures in the system and continuous gas flow were generated using a
central microprocessor (MKS Multi-Gas Controller 647C) and pumping
unit containing a vacuum pump (Busch Cobra screw pump, 570 m3/h), a
capacitive absolute pressure gauge (MKS Baratron) and a butterfly valve
for maintaining the desired pressure in the system.
Figure 11. Schematic presentation of the CVS system used to synthesize
Ga2O3 nanoparticles.
A first set of experiments was carried out by keeping the temperature
constant at 1200˚C and varying the pressure in the system from 40 to
300 mbar, while in the second series the pressure was kept constant at
300 mbar and the temperature of the hot-wall reactor was varied from
1000-1500˚C as summarized in Table 1. The synthesis of the sample
with highest OWS activity was repeated in order to check reproducibility.
Photocatalytic experiments and cocatalyst photodeposition
Photocatalytic experiments and cocatalyst photodeposition were
performed in a 500 ml double-walled stirred semi-continuous quartz glass
reactor equipped with an inner-radiation type 500 W Hg lamp (Peschl
UV-consulting) with an intensity of 59 mW cm-² as UV-light source and an
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
Characterization methods and data analysis
Center Ruhr (MERCUR) and Deutsche Forschungsgemeinschaft (DFG)
within the SPP 1613 (SolarH2).
The crystal structure of the samples was determined by X-Ray Diffraction
(XRD) using a PANalytical X-Ray diffractometer (X'Pert PRO) with Ni-
filtered Cu Kα (1.54056 Å) radiation produced at 40 kV and 40 mA in the
2θ interval from 10 to 120° with steps of 0.03° and a counting time of 200
s/step. XRD Rietveld refinements were performed using the MAUD
software with initial values for the monoclinic crystal structure (space
group C1 2/m 1(12)) published by Geller et. al.[23] It was also possible to
derive the degree of crystallinity from the ratio of the coherently scattered
intensity to the total scattered intensity in X-ray diffraction. It can be
quantified as:
Keywords: Water splitting • Photocatalysis • Chemical Vapor
Synthesis • Gallium oxide • Nanoparticles
[1]
B. A. Pinaud, J. D. Benck, L. C. Seitz, A. J. Forman, Z. Chen, T. G.
Deutsch, B. D. James, K.N. Baum, G. N. Baum, S. Ardo, H. Wang, E.
Miller, T. F. Jaramillo, Energy Environ. Sci. 2013, 6, 1983-2002.
D. M. Fabian, S. Hu, N. Singh, F. A. Houle, T. Hisatomi, K. Domen, F. E.
Osterloh, S. Ardo, Energy Environ. Sci. 2015, 8, 2825-2850.
T. Takata, C. Pan, K. Domen, Sci. Technol. Adv. Mater. 2015, 16,
033506.
[2]
[3]
[4]
X. Wang, Q. Xu, M. Li, S. Shen, X. Wang, Y. Wang, Z. Feng, J. Shi, H.
Han, C. Li, Angew. Int. Ed. 2012, 51, 13089-13092.
⨜퐴푑휃
Г =
(3)
[5]
[6]
A. Kudo, Y. Miseki, Chem. Soc. Rev. 2009, 38, 253-278.
Y. Sakata, T. Nakagawa, Y. Nagamatsu, Y. Matsuda, R. Yasunaga, E.
Nakao, H. Imamura, J. Catal. 2014, 310, 45-50.
⨜(퐴+퐵)푑휃
where Г is the degree of crystallinity, A is the computed XRD signal
(coherent scattering), B is the background signal (incoherent scattering)
and θ is the scattering angle. The specific surface area (SSA) of the
samples was obatined by low-temperature nitrogen adsorption at 77 K
(Quantachrome Autosorb 1C instrument) using the Brunauer-Emmett-
Teller (BET) method. Prior to the measurement the samples were
degassed at a temperature of 200°C under vacuum for approximately 24
h, and helium was used as a backfill gas. Around 70 mg of each sample
was used for the physisorption analysis. Morphology, shape and grain
size of the synthesized samples were studied by transmission and
scanning electron microscopes (TEM and SEM). TEM was performed on
a FEI-TITAN 300-800 microscope with a maximal spatial resolution of 0.8
Å. The sample was prepared without any solvent (dry preparation) and
without plasma cleaning. Atomic composition was identified by using
Electron Energy Loss Spectroscopy (EELS). SEM images in STEM mode
were recorded on a field emission scanning electron microscope (JEOL
JSM-7500F) with an accelerating voltage of 30 kV. The samples were
prepared by dispersing 2 mg of powder into 10 ml of water. Diffuse
reflectance UV-Vis spectra were measured using a PerkinElmer Lambda
650 spectrophotometer and a Harrick Praying Mantis diffuse reflectance
accessory. Spectra were recorded at ambient conditions from 800 nm to
[7]
[8]
J. Foo, C. Zhang, G. W. Ho, Nanoscale, 2013, 5, 759-764.
R. M. N. Yerga, M. C. Á. Galván, F. d. Valle, J. A. V. d. l. Mano, J. L. G.
Fierro, ChemSusChem 2009, 2, 471-485.
[9]
J. Yang, D. Wang, H. Han, C. Li, Acc. Chem. Res. 2013, 46, 1900-1909.
[10] R. Abe, J. Photochem. Photobiol. C. 2010, 11, 179-209.
[11] K. Kalyanasundaram, E. Borgarello, D. Duinghong, M. Grätzel, Angew.
Chem. Int. Ed. 1981, 20, 987-988.
[12] K. Maeda, K. Teramura, D. Lu, N. Saito, Y. Inoue, K. Domen, Angew.
Chem. Int. Ed. 2006, 45, 7806-7809.
[13] K. Maeda, K. Teramura, D. Lu, N. Saito, Y. Inoue, K. Domen, J. Phys.
Chem. C 2007, 111, 7554-7560.
[14] G. W. Busser, B. Mei, A. Pougin, J. Strunk, R. Gutkowski, W.
Schuhmann, M.-G. Willinger, R. Schlögl, M. Muhler, ChemSusChem
2014, 7, 1030-1034.
[15] J. Soldat, G. W. Busser, M. Muhler, M. Wark, ChemCatChem 2016, 8,
153-156.
[16] K. Maeda, K. Teramura, D. Lu, T. Takata, N. Saito, Y. Inoue, K. Domen,
J. Phys. Chem. B 2006, 110, 13753-13758.
[17] G. W. Busser, B. Mei, P. Weide, P. C. K. Vesborg, K. Stührenberg, M.
Bauer, X. Huang, M.-G. Willinger, I. Chorkendorff, R. Schlögl, M.
Muhler, ACS Catal. 2015, 5, 5530-5539.
[18] F. E. Osterloh, Top Curr. Chem. 2016, 371, 105-142.
[19] M. Winterer in Nanocrystalline Ceramics - Synthesis and Structure,
Springer, Heidelberg, 2002, pp. 34.
[20] B. D. Cullity in Elements of X-ray diffraction, 2nd ed. Addison-Wesley,
Menlo Park, CA, 1978, pp. 431.
200 nm with
a resolution of 1 nm and 0.4 s integration time.
[21] M. Birkholz in Thin film analysis by X-ray scattering, Wiley-VCH,
Weinheim, 2006, 3. 85-86.
Measurements of the powder samples were performed as a dilution of 25
mg Ga2O3 samples in 500 mg BaSO4, which was used as the internal
100% reflectance standard. The band gap was determined from Tauc
plots assuming a direct bandgap transition for Ga2O3 based on the
Kubelka-Munk function calculated from the reflectance spectra.
[22] K. K. Akurati, S. S. Bhattacharya, M. Winterer, H. Hahn, J. Phys. D:
Appl. Phys. 2006, 39, 2248-2254.
[23] S. Geller, J. Chem. Phys. 1960, 33, 676-684.
[24] L. Li, W. Wei, M. Behrens, Solid State Sci. 2012, 14, 971-981.
[25] T. Grewe, H. Tüysüz, ChemSusChem 2015, 8, 3084-3091.
[26] G. W. Busser, B. Mei, M. Muhler, ChemSusChem 2012, 5, 2200-2206.
[27] T. Hisatomi, K. Takanabe, K. Domen, Catal. Lett. 2015, 145, 95-108.
[28] Y. Sakata, T. Hayashi, R. Yasunaga, N. Yanaga, H. Imamura, Chem.
Commun. 2015, 51, 12935-12938.
Acknowledgements
The authors would like to thank to Dr. Ramzi Farra from the Electron
Microscopy Group, Fritz Haber Institute of the Max Planck Society in
Berlin, for TEM analysis and Prof. Dr. Christina Scheu and Dr. Siyuan
Zhang, Max Planck Institute for Iron Research in Düsseldorf, for EELS
analysis. This work was financially supported by Mercator Research
This article is protected by copyright. All rights reserved.

10.1002/cssc.201701309
ChemSusChem
FULL PAPER
Sasa Lukic, Jasper Menze, Philipp Weide, G. Wilma Busser, Markus Winterer, Martin
Muhler
Decoupling the Effects of High Crystallinity and Surface Area on the Photocatalytic
Overall Water Splitting over β-Ga2O3 Nanoparticles by Chemical Vapor Synthesis
Ga2O3 nanoparticles were synthesized by chemical vapor synthesis (CVS) at different
temperatures and pressures. When loaded with Rh/CrOx cocatalysts all samples evolved H2
and O2 in the stoichiometric ratio of 2:1 in photocatalytic overall water splitting (OWS). The
OWS activity of the CVS samples is demonstrated to depend on the optimum ratio of
crystallinity and specific surface area.
This article is protected by copyright. All rights reserved.