Platinum loaded sodium tantalate photocatalysts prepared by a flux method for photocatalytic steam reforming of methane

Article abstract of DOI:10.1016/j.apcata.2015.10.031

Lanthanum-doped sodium tantalate samples (NaTaO₃:La) were prepared by a flux method using a sodium chloride flux with various parameters, such as presence or absence of the flux, solute concentration, hold temperature, and amount of lanthanum doping. SEM images showed cubic and rectangular shapes for the samples prepared by the flux method, somewhat rounded shape for the sample prepared in the absence of the flux, and large particles for the sample without lanthanum doping. Among the parameters, the lanthanum doping and solute concentration much influenced the crystallites size of the NaTaO₃:La samples. Most of these NaTaO₃:La samples loaded with platinum cocatalyst exhibited the photocatalytic activity in the photocatalytic steam reforming of methane around room temperature. Among them, the highest activity was obtained by the Pt/NaTaO₃:La sample prepared by the flux method with moderate solute concentration, enough high hold temperature, and moderate amount of platinum and lanthanum doping. A positive correlation was found between the crystallite size and the photocatalytic activity. When we compared the catalysts having the same crystallite size, the sample prepared by the flux method showed higher photocatalytic activity than the catalyst prepared without the flux. It is suggested that the difference in the shape of particle would be important factor for the photocatalytic activity.

Full text of DOI:10.1016/j.apcata.2015.10.031

G Model
APCATA-15607; No. of Pages8
ARTICLE IN PRESS
Applied Catalysis A: General xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Platinum loaded sodium tantalate photocatalysts prepared by a flux
method for photocatalytic steam reforming of methane
Akira Yamamotoa,b, Shota Mizuba , Yurina Saeki , Hisao Yoshida
a
a
a,b,∗
a
Department of Interdisciplinary Environment, Graduate School of Human and Environmental Studies, Kyoto University, Yoshida Nihonmatsu-cho,
Sakyo-ku, Kyoto 606-8501, Japan
Elements Strategy Initiative for Catalysts & Batteries (ESICB), Kyoto University, Kyotodaigaku Katsura, Nishikyo-ku, Kyoto 615-8520, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Lanthanum-doped sodium tantalate samples (NaTaO :La) were prepared by a flux method using a sodium
3
Received 31 August 2015
Received in revised form 17 October 2015
Accepted 20 October 2015
Available online xxx
chloride flux with various parameters, such as presence or absence of the flux, solute concentration,
hold temperature, and amount of lanthanum doping. SEM images showed cubic and rectangular shapes
for the samples prepared by the flux method, somewhat rounded shape for the sample prepared in the
absence of the flux, and large particles for the sample without lanthanum doping. Among the parameters,
the lanthanum doping and solute concentration much influenced the crystallites size of the NaTaO3:La
samples. Most of these NaTaO3:La samples loaded with platinum cocatalyst exhibited the photocatalytic
activity in the photocatalytic steam reforming of methane around room temperature. Among them, the
highest activity was obtained by the Pt/NaTaO3:La sample prepared by the flux method with moderate
solute concentration, enough high hold temperature, and moderate amount of platinum and lanthanum
doping. A positive correlation was found between the crystallite size and the photocatalytic activity.
When we compared the catalysts having the same crystallite size, the sample prepared by the flux method
showed higher photocatalytic activity than the catalyst prepared without the flux. It is suggested that
the difference in the shape of particle would be important factor for the photocatalytic activity.
Keywords:
Photocatalytic steam reforming of methane
Hydrogen production
Sodium tantalite
Flux method
Molten salt method
©
2015 Elsevier B.V. All rights reserved.
1
. Introduction
reaction even in the presence of catalysts. The high temperature
operation causes several problems such as the large energy con-
sumption, the irreversible carbon formation, and the necessity of
expensive reactor. Thus, lowering the operation temperature has
been highly desired in the seam reforming of methane.
Utilization of photocatalysts is expected to be a promising way
for the development of the steam reforming of methane at low
temperature [1–3]. A reaction between methane and water to form
methanol and hydrogen was reported by Taylor et al. over La-doped
Hydrogen is one of the clean and oncoming energy sources, and
its production technology have been developed and investigated by
many researchers. As a source of hydrogen, methane is an attractive
hydrogen source because of the highest H/C values among hydro-
carbons, and it is one of the most abundant natural resource as well
as a main component of renewable biogas. Methane can be catalyti-
cally converted to hydrogen by steam reforming of methane, which
is the practically employed method for the hydrogen production.
The overall chemical equation of the steam reforming of methane
with successive water-gas shift reaction can be shown in Eq. (1).
WO under UV and visible light irradiation [4,5] in an aqueous solu-
3
tion containing methyl viologen dichloride as an electron transfer
reagent. Gondal et al. also reported the photocatalytic conversion
of methane to methanol in a batch reactor using a visible laser
(514 nm) and a WO3 catalyst [6]. However, these systems might
not be suitable for hydrogen production.
On the other hand, our group first reported the photocatalytic
steam reforming of methane over heterogeneous photocatalysts
using a gas mixture of methane and water around room tempera-
ture to produce hydrogen and carbon dioxide directly as shown in
Eq. (1), where the photoenergy was added to the system for com-
pensation of the Gibbs free energy change of this reaction [7]. In
other words, the utilization of photoenergy enables us to operate
◦
2
= 113.5 kJ mol−1
CH + 2H O → 4H + CO
ꢀG
(1)
4
2
2
2
98 K
This is a highly endergonic reaction; therefore, high temper-
ature, typically more than 1073 K, is necessary to promote the
∗ Corresponding author at: Department of Interdisciplinary Environment, Grad-
uate School of Human and Environmental Studies, Kyoto University, Yoshida
Nihonmatsu-cho, Sakyo-ku, Kyoto 606-8501, Japan. Fax: +81 75 753 2988.
E-mail address: yoshida.hisao.2a@kyoto-u.ac.jp (H. Yoshida).
http://dx.doi.org/10.1016/j.apcata.2015.10.031
0
926-860X/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: A. Yamamoto, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.10.031

G Model
APCATA-15607; No. of Pages8
ARTICLE IN PRESS
A. Yamamoto et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
2
the reaction at low temperatures, which may solve the some prob-
lems mentioned above. Platinum-loaded titanium dioxide (Pt/TiO2)
photocatalysts showed the activity at mild temperature under pho-
toirradiation, and gave the stoichiometric ratio of the products
Eq. (2), T is the hold temperature, and y shows the aimed amount of
lanthanum doping (mol%), if necessary. On the other hand, a refer-
ence sample were prepared without the flux in a solid state reaction
method: the starting mixture except for the sodium chloride flux
(
H /CO = 4) [7,8]. Various photocatalysts have been developed
i.e., Ta O5, Na CO , and La O , where the solute concentration x
2
2
2 2 3 2 3
the photocatalytic steam reforming such as Pt-loaded La-doped
corresponded to be 100, was ground well, heated with the same
rate of 200 K h , maintained at 1273 K for 5 h, and then cooled,
followed by washing, in the similar way as mentioned above. This
−
1
NaTaO3 (Pt/NaTaO :La) [7,9], Pt-loaded CaTiO3 [10,11], Rh-loaded
3
K Ti O [12,13], and Pt-loaded or Rh-loaded ␤-Ga O3 [14]. It was
2
6
13
2
also found that this photocatalytic reaction could be further pro-
moted at higher temperature; i.e., thermal energy could assist this
photocatalytic reaction, which promised further development [15].
However, their photocatalytic activities have been not enough yet,
and the further improvement of the photocatalytic activity have
been desired for the practical use.
sample is referred to as NaTaO :La(100, 1273, 2).
Platinum co-catalyst was loaded onto the prepared NaTaO :La
samples by an impregnation method. The sample was soaked in an
3
3
aqueous solution of H PtCl6 (Wako, 99.9%), dried up and calcined
2
at 673 K for 2 h. The sample powder was granulated to the size of
300–600 ␮m before the photocatalytic activity test. The Pt loaded
Flux method (molten salt method) is effective for synthesis of
high quality crystals (with high crystallinity), and the prepared
crystals often have characteristic and uniformed shapes covered
with particular flat facets. It is considered that the high crystallinity
is an important property for the highly active photocatalyst because
crystal defects are believed to function as recombination sites for
excited electron and hole pairs. In addition to the crystallinity, the
shape of the crystals or the particles is considered to be one of
the key factors affecting the photocatalytic activity [16–19]. The
crystal facets have been pointed out to help in the separation of
photogenerated carriers (electrons and holes), which leads to the
high photocatalytic activity [20–22]. The suppression of the recom-
bination is one of the strategies to improve the photocatalytic
activity. Up to date, a number of crystals with characteristic shapes
were synthesized by flux methods and examined as photocatalysts
sample is referred to as Pt(z)/NaTaO :La(x, T, y), where z shows the
loading amount of Pt (wt%).
3
2.2. Characterization
X-ray diffraction (XRD) measurement was carried out at room
temperature using a Shimadzu Lab X XRD-6000 using Cu K␣ radi-
ation (40 kV, 30 mA). The crystallite size was determined by the
Scherrer equation using the full width at half maximum (FWHM)
◦
of the diffraction line at 2ꢁ = 22.8 in the XRD patterns of NaTaO .
3
Scanning electron microscopy (SEM) images were recorded by
a JEOL JSM-890. Diffuse reflectance (DR) UV–vis spectrum was
recorded on a JASCO V-670 equipped with an integrating sphere
covered with BaSO4 reference. The band gap was estimated from
the spectrum according to Tauc plot [36]. The specific surface area
[
23–34].
In the present paper, a flux method was employed to prepare the
was estimated from the amount of N adsorption at 77 K measured
2
using a Quantachrome Monosorb.
NaTaO :La crystalline samples and the photocatalytic activity was
3
investigated in the photocatalytic steam reforming of methane.
2.3. Photocatalytic activity tests
2
. Experimental
Photocatalytic steam reforming of methane was carried out with
a fixed-bed flow reactor as described in our previous studies [7–9].
Shortly, a mixture of the catalyst granules (0.5 g) and quartz sand
2
.1. Catalyst preparation
3
(
1.2 g) was put into a quartz reactor (ca. 50 × 20 × 1 mm ) and the
Various NaTaO :La samples were prepared by a flux method.
3
reaction gas of CH (25%) and H O (0.75%) with an argon carrier was
4
2
Solid reagents of Ta O5, Na CO (Rare Metallic, 99.99%), La O3
−1
2
2
3
2
introduced at a flow rate of 50 mL min without heating at atmo-
spheric pressure. Light irradiation was carried out from a 300 W
xenon lamp without using any optical filter, where the light inten-
(
Kishida, 99.99%), and NaCl (Kishida, 99.5%) were used as-
purchased. The mixture of Ta O5, Na CO , La O , and NaCl was
ground in an aluminum mortar for 15 min, where the molar ratio
of Na CO to Ta O5 was unity. Although it was reported that an
excess amount of Na CO3 in the mixture could compensate the
volatile sodium and function as a flux [35], in the present study a
stoichiometric ratio of Na CO3 was added and NaCl was used as a
flux, which can help us to discuss the role of a flux more clearly.
The aimed amount of La was 0–5 mol%. The solute concentration of
2
2
3
2
3
−1
sity was measured to be 14 mW cm in the range of 245 ± 10 nm.
2
3
2
The outlet gas was analyzed by online gas chromatography with a
thermal conductivity detector at an interval of ca. 30 min. Since the
sensitivity for CO2 in the argon carrier was low, the experimental
error for the values of CO2 production rate was relatively large.
2
2
3. Results and discussion
5
–90% is defined in the following equation:
Solute concentration , x (mol%)
3.1. Characterization
Amount of NaTaO 3 (mol )
= Amount of NaTaO
× 100 (2)
Fig. 1 shows XRD patterns of the Pt/NaTaO :La samples prepared
by the flux method at various hold temperatures. A diffraction pat-
3
(mol ) +Amount of NaCl (mol )
3
The mixture was heated in a platinum crucible using an electric fur-
tern of Ta O5 (ICSD No. 9112) [37] was observed for the sample
2
−1
nace with a heating rate of 200 K h to various target temperatures
1073–1473 K, typically 1273 K), successively heated at the same
temperature (hold temperature) for 5 h, and then cooled down once
heated at 1073 K (Fig. 1a), which indicates that the temperature
is not enough to generate the NaTaO3 phase as desired. At this
temperature, Na CO would be decomposed to form Na O through
(
2
3
2
−
1
to 773 K at a cooling rate 100 K h and then to room temperature
without controlling the temperature. The obtained powder was dis-
persed in hot ion-exchanged water (300 mL, 353 K) and filtrated
with suction to separate the powder from the flux. The washing
procedure was repeated four times, and dried at 323 K overnight
decarbonation. Considering the melting points of the NaCl flux to
be 1074 K [38], the hold temperature and time (1073 K, 5 h) would
not be enough for Na O to react with the Ta O5 particles. By heating
2
2
over 1173 K, the clear diffraction lines of the NaTaO3 phase (ICSD
No. 980) [39] appeared without any impurity phase. It was revealed
that the higher temperature than 1173 K was necessary to generate
the NaTaO3 crystallites in this method.
to obtain the NaTaO :La sample. These samples are referred to as
3
NaTaO :La(x, T, y), where x is the solute concentration defined in
3
Please cite this article in press as: A. Yamamoto, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.10.031

G Model
APCATA-15607; No. of Pages8
ARTICLE IN PRESS
A. Yamamoto et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
3
NaTaO
Ta
3
NaTaO
g)
3
1000 cps
1
000 cps
2 5
O
(
(
(
(
(
(
(
e)
d)
c)
(
f)
e)
(d)
c)
(
b)
a)
20
(
(
b)
a)
1
0
30
40
50
60
70
10
20
30
40
50
60
70
2θ / degree
2θ / degree
Fig. 1. XRD patterns of the Pt(0.2)/NaTaO3:La(70, T, 1) samples prepared by the flux
method at various hold temperature of (a) 1073, (b) 1173, (c) 1273, (d) 1373, and
Fig. 2. XRD patterns of the Pt(0.03)/NaTaO3:La(x, 1273, 2) samples prepared by the
flux method at various solute concentrations of (a) 5, (b) 30, (c) 50, (d) 70, (e) 90, and
(f) 100 mol%, and (g) the Pt(0.03)/NaTaO3(70, 1273, 0) sample without La doping.
The NaTaO3:La(100, 1273, 2) sample was prepared without the flux.
(
e) 1473 K.
Fig. 2 shows the XRD patterns of the Pt/NaTaO :La samples pre-
3
pared by the flux method at the hold temperature of 1273 K with
various solute concentrations. For all the samples (Fig. 2a–f), the
indicated that all the La cations introduced were doped inside the
bulk or on the surface of the NaTaO :La samples, which means the
3
diffraction patterns were assignable to that of NaTaO , and any
impurity phase was not observed. The sample without La doping
La cations were neither captured in the NaCl matrix nor released
from the NaTaO3 particles by washing during the preparation.
As for the samples prepared with higher hold temperature than
1173 K, the average size of the NaTaO3 crystallites was estimated
from the diffraction line at 22.8 by using Scherrer equation and
listed in Table 1. The crystallite size did not depend on the hold
3
also showed the diffraction lines from NaTaO crystallites (Fig. 2g).
3
XRF measurements on the Pt/NaTaO :La(x, 1273, 2) samples
3
◦
confirmed that the amount of the doped La was almost the same
as the desired value (Table 1 entries 7–12, 16, 17). This clearly
Fig. 3. SEM images of the NaTaO3:La(x, 1273, 2) samples prepared by the flux method with various solute concentrations of (a) 5, (b) 30, (c) 70, (d) 90, and (e) 100 mol%,
where the NaTaO3:La(100, 1273, 2) sample was prepared without the flux, as well as (f) the NaTaO3(70, 1273, 0) sample without La doping.
Please cite this article in press as: A. Yamamoto, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.10.031

G Model
APCATA-15607; No. of Pages8
ARTICLE IN PRESS
A. Yamamoto et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
4
Table 1
Physical and optical properties of the NaTaO3:La samples prepared by the flux method and the photocatalytic performance of the Pt/NaTaO3:La samples in the photocatalytic
steam reforming of methane.
Entry Ta (K)
Solute
Introduced
La contentd
(mol%)
Crystallite
size (XRD)
(nm)
Particle size
(SEM) (nm)
SBETg (m g
2
−1
)
Bandgaph (eV) Loading
amount of Pt production
H2
concentrationb amount of
e
f
c
ratei
␮mol min 1
(mol%)
La (mol%)
(wt%)
−
(
)
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
1073
1173
1273
1323
1373
1473
1273
1273
1273
1273
1273
1273
1273
1273
1273
1273
1273
70
70
70
70
70
70
5
30
50
70
90
100
70
70
30
30
30
1
1
1
1
1
1
2
2
2
2
2
2
0
2
0
1
5
–
–
–
–
–
–
1.9
2.1
2.2
2.2
2.0
2.1
–
2.2
–
1.1
4.6
–j
–
–
–
–
–
–
195
210
160
210
216
315
733
210k
–
–
–
–
–
–
–
–
–
0.2
0.2
0.2
0.2
0.2
0.2
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0
0.12
1.2
1.4
1.2
1.3
72
69
80
77
76
54
56
62
71
68
76
84
4.7
3.5
3.5
2.0
2.0
5.2
4.6
5.0
3.5
2.4
2.1
0.92
1.0
0.60
0.56
0.78
1.2
4.16
4.16
4.14
4.13
4.12
4.10
4.14k
4.06
4.14
4.15
1
1
1
1
1
1
1
1
1.1
0.85
0.22
0.26
0.02
1.2
k
k
k
71
3.5
88
65
50
0.5
4.9
5.5
0.03
0.03
0.03
–
–
0.55
a
b
c
d
e
f
Hold temperature during the preparation in the flux method.
Solute concentration in the flux.
Introduced amount in preparation.
Doping amount of La measured with XRF.
Average crystallite size calculated from a line width in the XRD patterns.
Average particle size estimated from the SEM images. Relative standard deviations were ca. 60%.
Specific surface area calculated in the BET method.
Band gap determined by UV–vis adsorption spectra.
The hydrogen production rate was evaluated at 4 h later from the start of photoirradiation.
NaTaO3 was not formed.
The same data as those in the entry 10.
g
h
i
j
k
temperature (Table 1, entries 2–6). On the other hand, the solute
concentration and the presence of the flux much affected the crys-
tallite size; the size of the NaTaO3 crystallites increased with an
increase of the solute concentration (Table 1, entries 7–11) and
the sample prepared without the flux had the largest crystallite
size among these samples (Table 1, entry 12). This means that the
NaCl flux would suppress the crystal growth to some extent to yield
smaller crystallites.
3
2
1
0
(
e)
(a)
(
g)
(d)
c)
The crystallite size was also decreased with increasing the
amount of La doping (Table 1, entries 10, 13, 15–17). Kato et al.
reported that La doping could suppress the crystal growth of
NaTaO :La during the preparation of the NaTaO sample in a solid
(f)
(
b)
(
3
3
state reaction (SSR) method [16]. Thus, the present result confirmed
that the crystal growth could be suppressed by La doping also in the
present flux method.
260
280
300
320
The BET specific surface area of the samples was measured and
listed in Table 1, entries 2–17 and also depicted in Fig. S1, where
a rough tendency could be confirmed that the BET specific surface
area decreased with increasing the crystallite size as expected.
Wavelength / nm
Fig. 4. DR UV–vis spectra of the NaTaO3:La(x, 1273, 2) samples prepared by the flux
method at various solute concentrations of (a) 5, (b) 30, (c) 50, (d) 70, (e) 90, and
(f) 100 mol%, where the NaTaO3:La(100, 1273, 2) sample was prepared without the
flux, and (g) that of the NaTaO3:La(70, 1273, 0) sample prepared by the flux method
without La doping.
Fig. 3 shows the SEM images of the NaTaO :La samples prepared
3
by the flux method at various solute concentrations from 5 to 100
(
i.e., with/without the flux), as well as the sample without La dop-
ing. The shape of the NaTaO :La particles was changed with the
3
from XRD profiles. The samples prepared without flux and without
La doping consisted of much larger average particle sizes such as
presence of the flux; i.e., the samples prepared with the flux had
the cubic or rectangular particles (Figs. 3a–d), while the sample pre-
pared in the absence of the flux had the rounded particles (Fig. 3b).
Among them, the sample prepared with the solute concentration
of 70% exhibited the most clear cubic shape (Fig. 3c).
The average particle sizes of these samples were estimated from
the SEM images and listed in Table 1. The average particle size of
the La doped samples prepared with the flux was in the range of
3
15 and 733 nm, respectively.
In DR UV–vis spectra of the NaTaO :La samples, the adsorption
3
band appeared at shorter wavelength than 320 nm (Fig. 4). From
these spectra the bandgap of these samples were evaluated from
Tauc plot [36] and listed in Table 1. The values for the NaTaO :La
3
samples were in the range of 4.12–4.16 eV. In addition the sample
without La doping showed lower values such as 4.06 and 4.10 eV
1
60–216 nm, which did not systematically vary with the solute con-
(entries 13, and 15). It is found that the bandgap was not signifi-
centration (entries 7–11). It is clear that the observed particle size
in the SEM images was larger than the crystallite size calculated
cantly but slightly increased with decreasing of the crystallite size
as shown in Fig. S2.
Please cite this article in press as: A. Yamamoto, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.10.031

G Model
APCATA-15607; No. of Pages8
ARTICLE IN PRESS
A. Yamamoto et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
5
1
.5
1
0
.5
1
1
.5
0
.5
0
0
1
000
1100
1200
1300
1400
1500
0
1
2
3
4
5
Temperature / K
La contenst (mol%)
Fig. 5. The hydrogen production rate in the photocatalytic steam reforming of
methane over the Pt(0.2)/NaTaO3(70, T, 1) photocatalysts prepared with various
hold temperatures. The hydrogen production rate was evaluated at 4 h later from
the start of photoirradiation.
Fig. 6. The hydrogen production rate in the photocatalytic steam reforming of
methane over the Pt(0.03)/NaTaO3:La(30, 1273, y) photocatalysts prepared by the
flux method with various La contents. The La contents were measured with XRF.
The hydrogen production rate was evaluated at 4 h later from the start of photoir-
radiation.
In the previous study, it was reported that the band gap and
specific surface area of the NaTaO :La photocatalysts prepared in
3
a SSR method increased with an increase of the La content [9]. The
present study confirmed that, in addition to the presence of the
La doping, the decrease of the crystallite size would also enlarge
This result clearly indicates that the generation of the NaTaO3:La
phase is essential for achieving the high activity in the photocat-
alytic steam reforming of methane. Among these photocatalysts,
the Pt(0.2)/NaTaO3:La(70, 1273, 1) photocatalyst exhibited the
the band gap. Among the NaTaO :La(x, 1273, 2) samples prepared
3
with various solute concentrations, as they had similar amount of
La doping as shown in Table 1, it is suggested that the shift of the
band gap would originate from the variation of the crystallite size
−1
highest hydrogen production rate i.e., 1.4 ␮mol min
.
(
Fig. S2). This means that the structure, both the crystallite size and
the band structure, could be controlled by the solute concentration
in the flux method.
3.2.2. Effect of the La doping
Effect of La doping on the hydrogen production rate was inves-
tigated in the photocatalytic steam reforming of methane over the
Pt(0.03)/NaTaO :La(30, 1273, y) photocatalysts prepared by the
3
.2. Photocatalytic activity
3
flux method with various La contents (Fig. 6, Table 1 entries 8,
−
1
1
5–17). The hydrogen formation rate was low (0.02 ␮mol min
)
3.2.1. Effect of the hold temperature
over the photocatalyst without La doping, while La doping drasti-
cally improved the activity. Among these samples, the maximum
hydrogen production rate (1.2 ␮mol min ) was achieved at the La
The photocatalytic activity of the prepared Pt loaded NaTaO :La
3
samples was evaluated. A representative time course of the prod-
ucts formation rate is shown in Fig. S3. The products were hydrogen
and carbon dioxide only. The ratio of these two products was 3.8
after 210 min, which was close to the ideal ratio of Eq. (1) within
the experimental error. Without irradiation, no product formation
was observed in all the catalysts. The reaction lasted for a long time
under photoirradiation without catalytic deactivation. Since plat-
inum loading enhanced the reaction such as 4.6 times as shown in
Table 1, entries 10 and 14, TON was tentatively calculated based
on Pt. The TON was 53 for 210 min, which was clearly larger than
−
1
doping amount of 1 mol%, and the further addition of La doping over
1
mol% decreased the activity. In literature, the positive effects of
the La doping on the NaTaO3 photocatalysts prepared by the SSR
method have been reported for the photocatalytic water splitting
[
16] and the photocatalytic steam reforming of methane [9]. Here,
it was found that the La doping was also effective for the improve-
ment of the activity of the NaTaO3 photocatalysts prepared in the
flux method.
As for the reason why the La doping enhanced the photocatalytic
activity, some possibilities have been pointed out. Kato et al. sug-
gested that the presence of La suppressed the crystal growth to form
unity. These clarified that this reaction over the Pt/NaTaO :La sam-
3
ples proceeded photocatalytically. The production rate varied with
the samples as listed in Table 1, indicating that the photocatalytic
activity of the photocatalyst varied with the structures and the
properties.
Fig. 5 shows the effect of the hold temperature among the prepa-
ration parameters on the hydrogen production rate. The hydrogen
production rate was evaluated after ca. 4 h from the start of the
smaller NaTaO crystals with fine steps on the surface, which would
3
be advantageous compared to the much larger particles of the pho-
tocatalyst without La doping [16]. Yamakata et al. pointed out that
the La doping could prolong the lifetime of photo-generated carri-
ers (electron and hole), which would result in the high activity [40].
In our previous study, the substitutionally doped La cations would
change the band structure as well as the specific surface area of
the photocatalyst [9]. Actually, since both the crystallite size and
the electronic band structure simultaneously varied with La dop-
ing, it is unclear which factor is the most substantial. However, it
is obvious that the La doping would be advantageous for the pho-
tocatalytic activity in these photocatalytic reactions that proceed
through the water activation, such as the water splitting and the
photocatalytic steam reforming of methane.
reaction. The Pt/NaTaO :La photocatalyst prepared with the low
3
hold temperature at 1073 K provided low activity, while the pho-
tocatalysts prepared with the higher hold temperature than 1173 K
gave drastically higher activity. The photocatalysts prepared with
higher hold temperatures than 1173 K showed almost similar activ-
ity, which suggests that the temperature of 1173 K was enough to
prepare the NaTaO :La photocatalysts by the flux method. The XRD
3
pattern showed that the NaTaO3 crystallites were generated when
prepared with the hold temperature higher than 1173 K (Fig. 1).
Please cite this article in press as: A. Yamamoto, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.10.031

G Model
APCATA-15607; No. of Pages8
ARTICLE IN PRESS
6
A. Yamamoto et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
1
.5
1
1.5
1
0
.5
0
.5
0
0
0
.01
0.05
0.1
0.5
1
Pt loading (wt%)
0
20
40
60
80
Fig. 7. The hydrogen production rate in the photocatalytic steam reforming of
methane over the Pt(z)/NaTaO3(70, 1273, 2) photocatalysts with various Pt load-
ing amount. The hydrogen production rate was evaluated at 4 h later from the start
of photoirradiation.
Crystallite size / nm
Fig. 9. Correlation between the hydrogen production rate in the steam reform-
ing of methane and the crystallite size of NaTaO :La in the photocatalysts. Open
3
circles: the Pt(0.03)/NaTaO3:La(x, 1273, 2) photocatalysts prepared with various
solute concentrations, open squares: the Pt(0.2)/NaTaO3:La(70, T, 1) photocatalysts
prepared with various high hold temperatures (T = 1173–1473 K), open triangles:
the Pt(0.03)/NaTaO3:La(30, 1273, y) photocatalysts with La contents (y =1 and 5),
and a closed diamond: the Pt(0.03)/NaTaO3:La(100, 1273,2) photocatalysts pre-
pared without the flux. The linear fitting was calculated from the plots of the
Pt(0.03)/NaTaO3:La(x, 1273, 2) photocatalysts prepared with the various solute con-
centrations. The hydrogen production rate was evaluated at 4 h later from the start
of photoirradiation.
1
flux, i.e., the Pt(0.03)/NaTaO :La(100, 1273, 2) photocatalyst. How-
3
0
.5
0
ever, the photocatalysts prepared with the solute concentration
less than 50 mol% exhibited lower activity than the photocata-
lyst prepared without the flux. In the low solute concentrations
such as 5–30 mol%, the activity did not depend on the solute con-
centration. Thus, it was elucidated that the flux method could be
effective for the preparation of the highly active photocatalysts if
the preparation parameter was optimized. As for the preparation
0
20
40
60
80
100
Solute concentration (mol%)
of the NaTaO :La photocatalyst in the present condition, the solute
3
concentration around 70 mol% should be desirable.
Fig. 8. The hydrogen production rate in the photocatalytic steam reforming of
methane over the Pt(0.03)/NaTaO3:La(x, 1273, 2) photocatalysts prepared by the
flux method at various solute concentrations. The plot at 100% for the solute con-
centration shows the value for the sample prepared without the flux. The hydrogen
production rate was evaluated at 4 h later from the start of photoirradiation.
3.3. Correlation between the particle size and the activity
To investigate the key parameter for the photocatalytic activity,
the hydrogen production rate listed in Table 1 was plotted against
the particle size evaluated from the SEM images (Fig. S4), and the
band gap determined by DR UV–vis spectra (Fig. S5), respectively.
However, no clear correlation was observed between the hydrogen
production rate and these parameters as shown in Figs. S4, and S5.
Fig. 9 shows the relationship between the crystallite size of
the photocatalysts and the hydrogen production rate in the pho-
tocatalytic steam reforming of methane over them among the
photocatalysts prepared in the flux method with various param-
3
.2.3. Effect of the Pt cocatalyst
Loading of Pt cocatalyst also significantly enhanced the activity
as mentioned above. Fig. 7 shows the effect of the loading amount
of the Pt cocatalyst. Even at the low Pt loading such as 0.03 wt%,
the hydrogen production rate on the Pt(0.03)/NaTaO :La(70,
3
−
1
1
273, 2) photocatalyst was high (1.2 ␮mol min ), which is 4.6
times higher than that of non-modified NaTaO :La(70, 1273, 2)
3
−1
(
0.26 ␮mol min ). The increase of the Pt loading to 0.2 wt% slightly
increased the activity, and the maximum rate for the hydrogen
eters, i.e., the Pt(0.03)/NaTaO :La(x, 1273, 2) photocatalysts (open
3
−
1
production was obtained at 0.2 wt% (1.3 ␮mol min ). The further
increase of the Pt loading to 1.0 wt% drastically suppressed the
hydrogen production rate.
circles), the Pt(0.2)/NaTaO :La(70, T, 1) photocatalysts, where
3
T = 1173–1473 K (open squares), and the Pt(0.03)/NaTaO :La(30,
3
1273, y) photocatalysts (open triangles), mentioned above (Table 1,
entries 2–11, 16, and 17). Among the Pt(0.03)/NaTaO :La(x, 1273, 2)
3
3
.2.4. Effect of the solute concentration
photocatalysts prepared with various solute concentrations (open
circles), a positive linear correlation was clearly observed between
the crystallite size and the hydrogen production rate. The line in
Fig. 9 was obtained by a least square fitting of the plots for the
Fig. shows the hydrogen production rate with the
8
Pt/NaTaO :La photocatalysts prepared with various solute concen-
trations. Among the Pt(0.03)/NaTaO :La(x, 1273, 2) photocatalysts,
the highest activity was 1.2 ␮mol min in the present condition,
which was obtained with the photocatalyst prepared with the
solute concentration of 70%. The hydrogen production rate was 1.4
times higher than that on the photocatalyst prepared without the
3
3
−
1
Pt(0.03)/NaTaO :La(x, 1273, 2) photocatalysts. In addition, the plots
3
for the Pt(0.2)/NaTaO :La(70, T, 1) photocatalysts (open squares)
3
and the Pt(0.03)/NaTaO :La(30, 1273, y) photocatalysts (open tri-
3
angles) were in alignment with the line. Thus, it was confirmed
Please cite this article in press as: A. Yamamoto, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.10.031

G Model
APCATA-15607; No. of Pages8
ARTICLE IN PRESS
A. Yamamoto et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
7
that the photocatalytic activity of the Pt/NaTaO :La photocatalysts
photo-related process such as the charge separation and migration
to the surface.
3
prepared in the flux method strongly depended on the size of the
NaTaO3 crystallites. These samples showing the good correlation
actually included various samples with the composition, the struc-
ture and the electrical properties such as the La content, the particle
size, and the band gap as shown in Table 1. This means that these
parameters could not give a dominant effect on the photocatalytic
activity, although these parameters would be related to the crys-
tallite size and indirectly or slightly influence the photocatalytic
activity.
4. Conclusions
The present study demonstrated the effectiveness of the flux
method using a sodium chloride flux to synthesize a high-
performance Pt/NaTaO :La photocatalyst for the photocatalytic
3
steam reforming of methane. The flux method provided cubic or
rectangular morphology of the NaTaO :La particles. La doping dras-
3
The positive correlation between the photocatalytic activity
and the crystallite size (or particle size) was reported by several
groups in several multi-electron reactions; e.g., photocatalytic oxy-
gen evolution from silver nitrate aqueous solutions [41,42], and
photocatalytic CO2 reduction [23]. Amano et al., reported that the
tically improved the photocatalytic activity even when the samples
were prepared in the flux method. The optimized catalyst prepared
by the flux method shows 1.6 times higher activity than the sample
prepared without the flux when compared the catalysts having the
same crystallite size.
recombination of the photogenerated carriers in the large WO par-
3
Various Pt/NaTaO :La particles were prepared by changing the
3
ticles was slower than that in the small particles, and it would
be due to the fast surface recombination compared to the bulk
preparation condition and the composition, such as the pres-
ence or absence of the flux, the solute concentration, the hold
temperature, and the amount of the lanthanum doping and the
platinum cocatalyst, and the relationship between the structural
and physical properties and the photocatalytic activity was inves-
tigated. Among the preparation parameters in the flux method,
the solute concentration much influenced the crystallite size of
[
41]. Thus, the surface recombination is one possibility for the low
activity of the present small NaTaO :La crystallites because the
3
surface-to-volume ratio would be high for small particles. Another
possibility is that a large crystallite can absorb a larger number
of photons than a small one due to its large irradiated area per
one crystallite. In the case of the multi-electron reactions, the large
number of the absorbed photons might effectively result in the high
activity.
the NaTaO :La samples; i.e., the crystallite size increased with
3
increasing the solute concentration. It is notable that a positive
correlation was observed between the NaTaO :La crystallite size
3
On the other hands, when we compared the photocatalytic
and the hydrogen production rate; in other words, the larger crys-
tallite size gave the higher photocatalytic activity. Besides this
correlation, we proposed the cubic or rectangular morphology of
activity over the photocatalysts having the same NaTaO :La crys-
3
tallite size, the activity of the Pt(0.03)/NaTaO :La(100, 1273, 2)
3
photocatalyst prepared without the flux (closed diamonds) was
obviously lower than the line. In other words, the photocatalyst
prepared with the flux exhibited higher activity than the photocat-
alyst prepared without the flux if the samples had the same size
of the crystallites, which corresponded to 1.6 times higher activ-
ity. This suggests that the photocatalyst prepared without the flux
would have negative parameters other than the crystallite size; i.e.,
the sample prepared with flux would have one or more positive fac-
tors for the efficient activity compared to that prepared without the
flux.
As mentioned above, the photocatalysts prepared with the flux
consisted of the cubic or rectangular particles, different from the
roundish particles of the sample prepared without the flux. Thus,
the merits of the photocatalyst prepared in the flux method would
possibly originate from the morphology. One possibility is that the
cubic or rectangular particles covered with the clear facets might
be more beneficial for the efficient charge separation of the photo-
excited carriers in the photocatalysts [16,20]. Another is that the
clear shape particles might have high crystallinity with less crystal
defects, which might result in the less recombination of the electron
and hole pairs. The other is that the photocatalytic activity might
depend on the population of the surface sites, such as edges and cor-
ners, that would be determined by the morphology [17]. Thus, the
preferable morphology of the present photocatalyst prepared in the
flux method would well promote the charge separation or reduce
the recombination, or the cubic or rectangular shape with the regu-
lar surface structure might be desirable for the photocatalytic steam
reforming of methane.
the NaTaO :La photocatalysts would be effective for the steam
reforming of methane by comparing the roundish shape of the
photocatalyst prepared without the flux.
3
Acknowledgments
This work was partially supported by a Grant-in-Aid for Scien-
tific Research (B), (No. 25289285), and a Grant-in-Aid for Scientific
Research on Innovative Areas “Artificial photosynthesis (AnApple)”
(
(
No. 25107515) from the Japan Society for the Promotion of Science
JSPS).
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.apcata.2015.10.
0
31.
References
[
[
[
1] L. Yuliati, H. Yoshida, Chem. Soc. Rev. 37 (2008) 1592–1602.
2] K. Shimura, H. Yoshida, Catal. Surv. Asia 18 (2014) 24–33.
3] J. Baltrusaitis, I. Jansen, J.D. Schuttlefield Christus, Catal. Sci. Technol. 4 (2014)
2397–2411.
[
[
[
4] C.E. Taylor, R.P. Noceti, Catal. Today 55 (2000) 259–267.
5] C.E. Taylor, Catal. Today 84 (2003) 9–15.
6] M.A. Gondal, A. Hameed, A. Suwaiyan, Appl. Catal. A 243 (2003) 165–174.
[7] H. Yoshida, S. Kato, K. Hirao, J. Nishimoto, T. Hattori, Chem. Lett. 36 (2007)
30–431.
4
[
8] H. Yoshida, K. Hirao, J. Nishimoto, K. Shimura, S. Kato, H. Itoh, T. Hattori, J.
Phys. Chem. C 112 (2008) 5542–5551.
In several photocatalytic reactions, molecular adsorption prop-
erties could much affect the activity, where large surface area would
provide the large adsorption/reaction sites [43,44]. However, in
the present case, the increase of the specific surface area of the
[9] K. Shimura, S. Kato, T. Yoshida, H. Itoh, T. Hattori, H. Yoshida, J. Phys. Chem. C
14 (2010) 3493–3503.
1
[
[
10] K. Shimura, H. Miyanaga, H. Yoshida, Stud. Surf. Sci. Catal. 175 (2010) 85–92.
11] K. Shimura, H. Yoshida, Energy Environ. Sci. 3 (2010) 615–617.
NaTaO :La photocatalysts resulted in the decrease of the activ-
[12] K. Shimura, H. Kawai, T. Yoshida, H. Yoshida, Chem. Commun. 47 (2011)
958–8960.
3
8
ity (see Table 1). This means that the contribution of the rate of
these surface events (e.g., adsorption, surface reaction, and desorp-
tion) to the overall reaction rate would be smaller than that of the
[
[
13] K. Shimura, H. Kawai, T. Yoshida, H. Yoshida, ACS Catal. 2 (2012) 2126–2134.
14] K. Shimura, T. Yoshida, H. Yoshida, J. Phys. Chem. C 114 (2010) 11466–11474.
[15] K. Shimura, K. Maeda, H. Yoshida, J. Phys. Chem. C 115 (2011) 9041–9047.
16] H. Kato, K. Asakura, A. Kudo, J. Am. Chem. Soc. 125 (2003) 3082–3089.
[
Please cite this article in press as: A. Yamamoto, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.10.031

G Model
APCATA-15607; No. of Pages8
ARTICLE IN PRESS
A. Yamamoto et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
8
[
[
17] M. Farhadian, P. Sangpout, G. Hosseinzadeh, J. Energy Chem. 24 (2015)
71–177.
18] Z. Wei, E. Kowalska, J. Verrett, C. Colbeau-Justin, H. Remita, B. Ohtani,
Nanoscale 7 (2015) 12392–12404.
[30] K. Teshima, S. Lee, A. Yamaguchi, S. Suzuki, K. Yubuta, T. Ishizaki, T. Shishido,
S. Oishi, CrystEngComm 13 (2011) 1190–1196.
[31] J. Sun, G. Chen, Y. Li, R. Jin, Q. Wang, J. Pei, Energy Environ. Sci. 4 (2011)
4052–4060.
1
[
[
19] G. Liu, J.C. Yu, G.Q. Lu, H.-M. Cheng, Chem. Commun. 47 (2011) 6763–6783.
20] R. Li, F. Zhang, D. Wang, J. Yang, M. Li, J. Zhu, X. Zhou, H. Han, C. Li, Nat.
Commun. 4 (2013) 1432.
[32] D. Arney, T. Watkins, P.A. Maggard, J. Am. Ceram. Soc. 94 (2011) 1483–1489.
[33] H. Kato, M. Kobayashi, M. Hara, M. Kakihana, Catal. Sci. Technol. 3 (2013)
1733–1738.
[
[
21] T. Ohno, K. Sarukawa, M. Matsumura, New J. Chem. 26 (2002) 1167–1170.
22] K. Iizuka, T. Wato, Y. Miseki, K. Saito, A. Kudo, J. Am. Chem. Soc. 133 (2011)
[34] H. Yoshida, M. Takeuchi, M. Sato, L. Zhang, T. Teshima, M.G. Chaskar, Catal.
Today 232 (2014) 158–164.
2
0863–20868.
[35] H. Kato, A. Kudo, J. Phys. Chem. B 105 (2001) 4285–4292.
[36] J. Tauc, R. Grigorovici, A. Vancu, Phys. Status Solidi B 15 (1966) 627–637.
[37] N.C. Stephenson, R.S. Roth, Acta Crystallogr. B: Struct. Sci. Cryst. Eng. Mater.
27 (1971) 1037–1044.
[
[
[
[
[
[
[
23] H. Yoshida, L. Zhang, M. Sato, T. Morikawa, T. Kajino, T. Sekito, S. Matsumoto,
H. Hirata, Catal. Today 251 (2015) 132–139.
24] K. Teshima, K. Horita, T. Suzuki, N. Ishizawa, S. Oishi, Chem. Mater. 18 (2006)
3
693–3697.
[38] J. Akella, S.N. Vaidya, G.C. Kennedy, Phys. Rev. 185 (1969) 1135–1140.
[39] M. Ahtee, L. Unonius, Acta Crystallogr. A: Found. Adv. 33 (1977) 150–154.
[40] A. Yamakata, T. Ishibashi, H. Kato, A. Kudo, H. Onishi, J. Phys. Chem. B 107
(2003) 14383–14387.
[41] F. Amano, E. Ishinaga, A. Yamakata, J. Phys. Chem. C 117 (2013) 22584–22590.
[42] F. Amano, M. Nakata, Appl. Catal. B 158–159 (2014) 202–208.
[43] J.M. Herrmann, Top. Catal. 34 (2005) 49–65.
25] K. Teshima, Y. Niina, K. Yubuta, T. Suzuki, N. Ishizawa, T. Shishido, S. Oishi,
Eur. J. Inorg. Chem. 2007 (2007) 4687–4692.
26] K. Teshima, K. Yubuta, T. Shimodaira, T. Suzuki, M. Endo, T. Shishido, S. Oishi,
Cryst. Growth Des. 8 (2008) 465–469.
27] L. Zhen, C.Y. Xu, W.S. Wang, C.S. Lao, Q. Kuang, Appl. Surf. Sci. 255 (2009)
4
149–4152.
28] S. Lee, K. Teshima, Y. Niina, S. Suzuki, K. Yubuta, T. Shishido, M. Endo, S. Oishi,
CrystEngComm 11 (2009) 2326–2331.
29] S. Lee, K. Teshima, Y. Mizuno, K. Yubuta, T. Shishido, M. Endo, S. Oishi,
CrystEngComm 12 (2010) 2871–2877.
[44] H. Kominami, S. Murakami, J.-i. Kato, Y. Kera, B. Ohtani, J. Phys. Chem. B 106
(2002) 10501–10507.
Please cite this article in press as: A. Yamamoto, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.10.031