Photoreduction of carbon dioxide to methane over Sb1.5Sn8.5-xTixO19.0 with high conductivity

Article abstract of DOI:10.1166/jnn.2018.15662

In order to enhance the photoreduction of CO₂ to CH₄, a new type of photocatalyst, Sb_1.5Sn_8.5-xTi_xO_19.0, with high conductivity and low bandgap was developed by partially incorporating Ti into the framework of Sb_1.5Sn_8.5O_19.0 (antimony-doped tin oxide, ATO) using a controlled hydrothermal method. XRD and TEM analyses indicated that the Sb_1.5Sn_8.5-xTi_xO_19.0 particles exhibited a tetragonal crystal structure and were approximately 20 nm in size. Furthermore, the bandgap and conductivity of these materials increased with increasing Ti content. A study of the photoreduction of CO₂ with H₂O revealed a remarkable increase in the generation of CH₄ over the Sb_1.5Sn_8.5-xTi_xO_19.0 catalysts. In particular, CH₄ generation was the highest when Sb_1.5Sn_8.5Ti_1.0O_19.0 was used as the photocatalyst, and was three-fold higher than that achieved by using anatase TiO₂. Photoluminescence studies showed that the enhanced photocatalytic activity of the Sb_1.5Sn_8.5-xTi_xO_19.0 materials could be attributed to the interfacial transfer of photogenerated charges, which led to an effective charge separation and inhibition of the recombination of photogenerated electron-hole (e^-/h⁺) pairs.

Full text of DOI:10.1166/jnn.2018.15662

Article
Journal of
Nanoscience and Nanotechnology
Vol. 18, 6369–6377, 2018
Copyright © 2018 American Scientific Publishers
All rights reserved
www.aspbs.com/jnn
Printed in the United States of America
Photoreduction of Carbon Dioxide to Methane Over
Sb1ꢀ5Sn8ꢀ5−xTixO19ꢀ0 with High Conductivity
Jeong Yeon Do, Byeong Sub Kwak, and Misook Kang∗
Department of Chemistry, College of Natural Sciences, Yeungnam University, Gyeongsan, Gyeongbuk 38541, Republic of Korea
In order to enhance the photoreduction of CO2 to CH4,
a new type of photocatalyst,
Sb1ꢀ5Sn8ꢀ5−xTixO19ꢀ0, with high conductivity and low bandgap was developed by partially incorpo-
rating Ti into the framework of Sb1ꢀ5Sn8ꢀ5O19ꢀ0 (antimony-doped tin oxide, ATO) using a controlled
hydrothermal method. XRD and TEM analyses indicated that the Sb1ꢀ5Sn8ꢀ5−xTixO19ꢀ0 particles
exhibited a tetragonal crystal structure and were approximately 20 nm in size. Furthermore,
the bandgap and conductivity of these materials increased with increasing Ti content. A study
of the photoreduction of CO2 with H2O revealed a remarkable increase in the generation of
CH4 over the Sb1ꢀ5Sn8ꢀ5−xTixO19ꢀ0 catalysts. In particular, CH4 generation was the highest when
Sb1ꢀ5Sn8ꢀ5Ti1ꢀ0O19ꢀ0 was used as the photocatalyst, and was three-fold higher than that achieved by
using anatase TiO2. Photoluminescence studies showed that the enhanced photocatalytic activity
of the Sb1ꢀ5Sn8ꢀ5−xTixO19ꢀ0 materials could be attributed to the interfacial transfer of photogener-
ated charges, which led to an effective charge separation and inhibition of the recombination of
IP: 62.4.−55.+113 On: Tue, 19 Jun 2018 19:21:13
photogenerated electron–hole (e /h ꢁ pairs.
Copyright: American Scientific Publishers
Keywords: Photoreduction, Carbon Dioxide, Methane, Sb1ꢀ5Sn8ꢀ5−xTixO19ꢀ0, Conductivity.
Delivered by Ingenta
1. INTRODUCTION
Sb2O3 (bandgap = 3.0 eV)5 is another promising mate-
rial for forming composite semiconductors with TiO2.
While several composite materials containing Sb2O3 have
been reported, such as Sb2O3 doped SnO2 particles
(antimony-doped tin oxide, ATO),6 the photocatalytic
properties of coupled Sb2O3/TiO2 have not been stud-
ied extensively. On the other hand, tin oxide (SnO2,
3.8–4.1 eV)7 is an n-type, wide-band gap semiconductor,
which is transparent in visible light. Its electrical prop-
erties are essentially dependent on its stoichiometry, the
nature and amount of dopants, and its nanostructure. The
dopants generally used to increase the n-type conductivity
of SnO2 are Sb (ATO) and F (F-doped tin oxide, FTO).
In particular, ATO with a tetragonal crystal lattice is an
important transparent conducting oxide (TCO)8 that has
been studied extensively for its chemical, mechanical, and
environmental stability. ATO films exhibit high conduc-
tivity and the electrons can tunnel easily in either direc-
tion between the adsorbed electroactive species and the
SnO2 conduction band. The introduction of Sb into the tin
oxide lattice increases the electron conductivity consider-
ably, which highlights the potential use of this material as
an excellent conducting agent.9 In addition, ATO is trans-
parent throughout the visible region but it reflects infrared
In recent years, the demand for technologies pertaining to
carbon management, which not only mitigate global tem-
perature, but also help partially fulfill the energy needs,
has steadily increased. One of the leading methods for the
production of methane as an energy source is the pho-
tocatalytic reduction of CO2 and H2O. This method is
also environmentally friendly: CO2 reduction with H2O
via photocatalysis is vital for the development of the solar
energy-based carbon neutral cycle. Among the various
semiconductor materials known to act as efficient pho-
tocatalysts, the main focus has been on titanium diox-
ide (TiO2)1–3 because it possesses several advantageous
attributes including a higher oxidation potential, relatively
low cost, abundance, non-toxicity, and chemical and ther-
mal stability. However, the immediate recombination of
photogenerated charges (e−/h+ꢀ leads to lower CO2 reduc-
tion efficiency and selectivity over TiO2.4 Therefore, it is
imperative to find new materials or modify the structure of
TiO2 to prevent the recombination of electrons and holes
for effective CO2 photoreduction and conversion to value-
added chemicals and fuels.
∗Author to whom correspondence should be addressed.
J. Nanosci. Nanotechnol. 2018, Vol. 18, No. 9
1533-4880/2018/18/6369/009
doi:10.1166/jnn.2018.15662
6369

Photoreduction of Carbon Dioxide to Methane Over Sb1.5Sn8.5−xTixO19.0 with High Conductivity
Do et al.
light. These features allow ATO to be used as transparent
electrodes,10 energy storage devices,11 and in photovoltaic
and optoelectronic devices.12ꢁ13 In this study, the excel-
lent conducting properties and lower bandgap attribute of
ATO were applied to photocatalysis. Generally, ATO par-
ticles were composited at an Sb:Sn molar ratio of 8.5:1.5
with a high concentration of Sb ions. ATO has been pre-
viously synthesized successfully with Sb contents below
1.5 mol%.14 Therefore, an attempt was made to prepare
ATO (Sb:Sn = 1.5:8.5 atomic ratio) with low Sb contents
for high transparency in the visible region and at low cost.
Therefore, Sb1ꢂ5Sn8ꢂ5O19ꢂ0 was introduced as a base
material and the resulting Sb1ꢂ5Sn8ꢂ5−xTixO19 (x = 0,
0.5, 1.0, 1.5) semiconductor composites, which partially
replaced Ti ions in the Sn sites in the ATO framework,
were synthesized by conventional hydrothermal treatment.
These materials were applied to CO2 photoreduction to
obtain CH4. Experiments were conducted to identify the
most appropriate Sn/Ti ratios for efficient methane pro-
duction. The synthesized Sb1ꢂ5Sn8ꢂ5−xTixO19 materials were
characterized by X-ray diffraction (XRD), transmission
electron microscopy (TEM), cyclic voltammetry (CV),
conductivity, ultraviolet-visible (UV-Vis) spectroscopy,
and photoluminescence (PL) spectroscopy.
powders was performed at room temperature and a scan
rate of 100 mV s−1 in 0.1 M KCl solution as the support-
ing electrolyte, platinum wire as the working and counter
electrodes, and Ag/AgCl as the reference electrode. The
UV-vis spectra were obtained using a Cary 500 spectrom-
eter with a reflectance sphere in the range, 200–800 nm.
Finally, the recombination tendency of the photogener-
ated electron–hole pairs (e−/h+ꢀ was estimated by using a
Perkin Elmer instrument with a He–Cd laser source at a
wavelength of 325 nm.
To determine the conductivity (ohm/square) of the
prepared materials, the crystallized Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0
powders were fixed on polyethylene terephthalate (PET)
films using a binder prepared in our laboratory. The
reagents used for the colloidal solutions, which were pre-
pared by mixing the Sb1ꢂ5Sn8ꢂ5−xTixOy powders and silane
binders, were as follows: tetraethylorthosilicate (TEOS,
99.95%, Junsei Chemical, Japan), which was used as a
silane binder precursor, hydrochloric acid (35%, Junsei
Chemical, Japan), and distilled water, used for hydroly-
sis. Tetraethylorthosilicate had four OH groups, divided
according to substitution reactions from the OR group into
the OH group. The hydroxyl group combines with the
oxygen group in the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders, elim-
inating H2O via condensation reaction. Finally, another
condensation reaction between the hydroxyl groups of the
binder and the PET film led to the combination of the
2. EXPERIMENTAL DETAILS
Sb1ꢂ5Sn8ꢂ5O19ꢂ0 as the basic material and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0
IP: 62.4.55.113 On: Tue, 19 Jun 2018 19:21:13
silane binder with the substrate. The fixation was con-
nanomaterials, in which the Ti ions partially replaced the
Copyright: American Scientific Publishers
ducted through the elimination of one H2O molecule dur-
Delivered by Ingenta
Sn sites in the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 framework, were prepared
by conventional hydrothermal treatment. Tin chloride
(SnCl2 · xH2O, 99.9%, Junsei Chemical, Japan) and anti-
mony chloride (SbCl3, 99.9%, Junsei Chemical, Japan)
were used as the Sn and Sb precursors, respectively. They
were dissolved gradually in distilled water and the mixture
solution was aged for 2 h. TTIP (TTIP, 98.0%, Junsei
Chemical, Japan), used as a Ti source, was added slowly
to the solution with stirring. The amount of Ti added was
varied according to the Ti/Sn molar ratios 0, 0.06, 0.13,
and 0.21 mol% (x = 0, 0.5, 1.0, 1.5). The pH of the solu-
tion was maintained at 10 using ammonium hydroxide.
After stirring homogeneously for 2 h, the final solution
was traꢀnsferred to an autoclave for thermal treatment
at 200 C under a nitrogen environment for 12 h. The
resulting precipitꢀate was washed with distilled water and
then dried at 60 C for 24 h. The four materials obtained
ꢀ
ing the drying treatment at 160 C for 5 min.
The photoreactor consisted of a rectangular quartz cell
with a total volume of 7.0 mL. The photocatalytic activ-
ity was studied by using 0.20 g of the powdered cata-
lyst distributed uniformly at the bottom of the reaction
chamber. A 1.0 mm-thick quartz glass window cover was
placed on top of the reactor to enable an effective trans-
fer of irradiation from a 6.0 W cm−2 mercury lamp at
ꢄ = 365 nm. The reactor chamber and lamp were covered
with aluminum foil to ensure that all the irradiation that
participated in the reaction passed only through the quartz
window. The reactor, which was checked for leakage at
atmospheric pressure for several hours, was purged with
helium carrier gas. The concentration of CO2 was con-
trolled by using helium (99.99%) as the diluent. The reac-
tion temperature and pressure were maintained at 303 K
and 1.0 atm, respectively. The reactor was purged with
a mixture of CO2 and helium for 1 h prior to starting
the experiment. The CO2:H2O ratio introduced was fixed
to 1:2. During photocatalysis, the product mixture was
sampled off-line by using a gas tight syringe with the
same volume, and analyzed by gas chromatography (GC,
iGC7200, Donam Co., Korea) equipped with a thermal
conductivity detector (TCD) and a flame-ionized detector
(FID). First, the gaseous products were in situ flowed into
the TCD detector, which was connected to the Carboxen
in this manner were Sb1ꢂ5Sn8ꢂ5O19ꢂ0, Sb1ꢂ5Sn8ꢂ0Ti0ꢂ5O19ꢂ0
Sb1ꢂ5Sn7ꢂ5Ti1ꢂ0O19ꢂ0, and Sb1ꢂ5Sn7ꢂ0Ti1ꢂ5O19ꢂ0
,
.
The synthesized Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 (x = 0, 0.5, 1.0,
1.5) powders were examined by XRD (MPD, PANalyti-
cal) using nickel-filtered CuKꢃ radiation (30 kV, 30 mA).
The morphology of the particles was studied by TEM
(JEOL 2000EX) and the atomic compositions of the films
were determined by energy dispersive X-ray spectroscopy
(EDAX, EX-250, Horiba) operated at 120 kV. CV (BAS
100B) of the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 (x = 0, 0.5, 1.0, 1.5)
6370
J. Nanosci. Nanotechnol. 18, 6369–6377, 2018

Do et al.
Photoreduction of Carbon Dioxide to Methane Over Sb1.5Sn8.5−xTixO19.0 with High Conductivity
1000 (Young Lin Instrumentals Co., Korea) column to ana-
lyze the light gases (H2, O2, and CO). The extracted gases
were again inserted into the FID detector to separate the
C1 (methane)–C3 light hydrocarbons and oxygenated com-
pounds such as methanol, acetaldehyde, and formic acid.
The product selectivity was calculated by using the fol-
lowing equation, Ci (%) = Ci moles in product/total moles
of C produced ×100.
Figure 2 shows the TEM images of the Sb1ꢂ5Sn8ꢂ5O19ꢂ0
and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 particles. It was found that the
morphologies of the particles were almost independent of
the amount of Ti. Irregular particles of sizes >100 nm
were obtained and aggregated into nanosheets of 20–40 nm
thickness. The particle sizes decreased with the increas-
ing amount of inserted Ti, and were closer to the crystal-
lite sizes estimated by the Scherrer equation in Figure 1.
The nanoparticles contained bulk pores between the inter-
particles, possibly because of the growth and aggregation
of small particles, and the aggregations decreased with
increasing Ti insertion.
EDAX confirmed the presence of metals on the surface
of the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 particles, as
illustrated in Figure 3. The EDAX spectra revealed the
O, Sn, Sb, and Ti peaks, indicating that the other com-
ponents of the precursors were not present in the sam-
ples. The table located on the below lists the actual atomic
compositions on the surface of all the samples deter-
mined by EDAX. The actual Ti/Sn atomic molar ratios in
3. RESULTS AND DISCUSSION
Figure 1 shows the XRD patterns of the Sb1ꢂ5Sn8ꢂ5O19ꢂ0
and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders. All XRD peaks of the
synthesized samples were indexed to the tetragonal crys-
tal system with the P42/mnm space group, which was in
good agreement with the literature data (JCPDS card no.
01-088-2348). Compared to the standard card, the relative
intensities of the peaks corresponding to the 2ꢅ = 26.59
(110), 33.91 (101), 37.96 (200), 39.01 (111), 60.3 (211),
54.76 (220), 57.92 (002), 61.89 (310), 64.83 (112), 65.99
(301), 71.37 (202), and 78.74ꢀ (321) reflections were sim-
ilar. A small XRD peak at 28.5ꢀ (2ꢅꢀ, which was assigned
to Sb2O3, was observed in the Ti-inserted samples. This
indicated that the unstable Sb ions were also removed
from the framework when Sn was substituted with Ti
ions. The Ti ions appeared to have been inserted success-
fully without the collapse of the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 tetragonal
framework. On the other hand, the mean crystallite sizes
the Sb1ꢂ5Sn8ꢂ5O19ꢂ0, Sb1ꢂ5Sn8ꢂ0Ti0ꢂ5O19ꢂ0, Sb1ꢂ5Sn7ꢂ5Ti1ꢂ0O19ꢂ0
,
and Sb1ꢂ5Sn7ꢂ0Ti1ꢂ5O19ꢂ0 samples were calculated to be
0.00, 0.06, 0.13, and 0.21, respectively, and the values
were the same as the amount of Ti added in the synthe-
sis steps except for the O-atomic content. This suggests
that stoichiometric Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 particles had been
synthesized.
Figures 4(a) and (b) show the UV-vis absorption spec-
IP: 62.4.55.113 On: Tue, 19 Jun 2018 19:21:13
16
of Sb1ꢂ5Sn8ꢂ5O19ꢂ0, Sb1ꢂ5Sn8ꢂ0Ti0ꢂ5O19ꢂ0, Sb Sn Ti
O
,
Copyright: American Scientific Publishers
tra and the Tauc plots of the synthesized Sb1ꢂ5Sn8ꢂ5O19ꢂ0
and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders, respectively. Generally,
absorption spectroscopy is a good way of examining the
changes in particle size with small changes in the chem-
ical nature of the particle surface.17 This is a practical
method to estimate the energy gap from the wavelength
at which the extrapolated absorption edge spectrum inter-
cepts on the x-axis or of that at the excitonic (shoul-
der) peak, called ꢄ1/2. The absorption spectra showed a
steeper absorption edge, which was similar to the results
obtained by the Tauc equation. A broad absorption band
corresponding to the tetragonal Sb1ꢂ5Sn8ꢂ5O19ꢂ0 material
appeared at approximately 250–500 nm. The two curves
in the absorption spectra for isolated SnO2 (first curve at
∼300 nm) and Sb2O3 (second curve at ∼420 nm) parti-
cles showed band gaps of 4.1 and 3.0 eV, respectively.
This result is consistent with the XRD patterns shown
in Figure 1. The band gap of the semiconducting mate-
rials is closely related to the range of absorbed wave-
lengths, where the band gap decreases with increasing
absorption wavelength. The band gap can be calculated
using the Tauc equation: (ꢃhꢆꢀn = Bꢇhꢆ − Egꢀ, where
hꢆ is the photon energy, ꢃ is the absorption coefficient,
B is a constant relative to the material, and the exponent
n is a value that depends on the nature of the transition
(n = 2 for a direct allowed transition, 2/3 for a direct for-
bidden transition, and 1/2 for an indirect allowed transi-
tion). The band gap energies were determined by simply
1ꢂ5
7ꢂ5 1ꢂ0 19ꢂ0
Delivered by Ingenta
and Sb1ꢂ5Sn7ꢂ0Ti1ꢂ5O19ꢂ0 particles calculated using the
Scherrer equation15 were approximately 40, 35, 28, and
20 nm, respectively. On the other hand, the conductivity
increased from 1.0×106 to 1ꢂ0×105 ohm/square depend-
ing on the extent of Ti-insertion. This was probably the
consequence of an increase in the crystal defects because
of the insertion of Ti.
Figure 1. XRD patterns and conductivities of the prepared
Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders.
J. Nanosci. Nanotechnol. 18, 6369–6377, 2018
6371

Photoreduction of Carbon Dioxide to Methane Over Sb1.5Sn8.5−xTixO19.0 with High Conductivity
Do et al.
IP: 62.4.55.113 On: Tue, 19 Jun 2018 19:21:13
Figure 2. TEM images of the synthesized Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO
particles.
Copyright: American Sci1e9ꢂn0 tific Publishers
Delivered by Ingenta
extrapolating the end of the curve under an assumption
that the Sn and Sb ions were completely included into the
tetragonal crystal structure. The band gap thus obtained
for Sb1ꢂ5Sn8ꢂ5O19ꢂ0 was approximately 1.04 eV. The band
gaps of the Sb1ꢂ5Sn8ꢂ0Ti0ꢂ5O19ꢂ0, Sb1ꢂ5Sn7ꢂ5Ti1ꢂ0O19ꢂ0, and
Sb1ꢂ5Sn7ꢂ0Ti1ꢂ5O19ꢂ0 particles were 1.03, 1.10, and 1.27 eV,
respectively; the band gaps increased with the increasing
Ti-insertion. The different band heights were also affected
by the amount of Ti inserted in the lattice. Generally, a
lower band gap is better for enhancing the photocatalytic
performance under weak solar radiation.18 However, this
principle is not always true because there is a smaller band
gap to facilitate the recombination between the electrons
and holes. This result and the XRD patterns suggest that
the synthesized Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0
particles co-exist with their perfect composites and a few
isolated SnO2, Sb2O3 and TiO2 particles.
in the form of the half-cell potential, E01/2, can also be
determined.19 Recently, some researchers have devised a
useful equation based on the CV data for determining
the HOMO and LUMO energy levels. Here, Eonset is the
starting point of the redox potential, which is used more
often than the peak potential value. The potential loca-
tion of the valence and conduction bands are as impor-
tant as the band gaps of the photocatalysts because the
potentials should always be included in the CO2 oxidation-
reduction potential, CO2/CH4 = −0ꢂ24 eV. In the case
of Sb1ꢂ5Sn8ꢂ5O19ꢂ0, there are two redox reversible reac-
tions, namely, Sb(III) → Sb(0) and Sn(IV) → Sn(0), whose
absolute reduction potentials (Epcꢀ are at −0.2667 and
+0.265.0 V, respectively. Here, the potential was only
considered for the Sn(IV) → Sn(0) reaction in order to
focus on the change in Sn upon Ti-insertion because
of the amount of Sb. The onset potentials for reduc-
Figure 5 shows the cyclic voltammograms of the
Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders. In the
present study, the oxidation potentials of these materials
were measured by CV in distilled water solutions. The
pelletized samples were used as the working electrodes,
Ag/AgCl was used as the reference electrode, and 0.1 M
KCl was used as the supporting electrolyte. A reversible
wave for the absolute potentials between the reduction
(Epcꢀ and oxidation peaks (Epaꢀ is desirable. When such
reversible peaks are observed, thermodynamic information
tion (LUMO) for Sb1ꢂ5Sn8ꢂ0Ti0ꢂ5O19ꢂ0, Sb1ꢂ5Sn7ꢂ5Ti1ꢂ0O19ꢂ0,
and Sb1ꢂ5Sn7ꢂ0Ti1ꢂ5O19ꢂ0 powders were determined to be
+0.25, +0.248, and +0.255 V, respectively. The HOMO
potentials for the Sb1ꢂ5Sn8ꢂ5O19ꢂ0
, Sb1ꢂ5Sn8ꢂ0Ti0ꢂ5O19ꢂ0,
Sb1ꢂ5Sn7ꢂ5Ti1ꢂ0O19ꢂ0, and Sb1ꢂ5Sn7ꢂ0Ti1ꢂ5O19ꢂ0 samples were
calculated to be −0.775, −0.78, −0.852, and −1.015 V,
respectively, by subtracting the calculated band gap from
the LUMO values, assuming that the Sn and Sb ions were
completely included into the tetragonal crystal structure in
Figure 4.
6372
J. Nanosci. Nanotechnol. 18, 6369–6377, 2018

Do et al.
Photoreduction of Carbon Dioxide to Methane Over Sb1.5Sn8.5−xTixO19.0 with High Conductivity
IP: 62.4.55.113 On: Tue, 19 Jun 2018 19:21:13
Copyright: American Scientific Publishers
Delivered by Ingenta
Figure 3. Atomic compositions on the surface of the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 particles determined by EDAX analysis.
Figure 6 compares the PL spectra of the Sb1ꢂ5Sn8ꢂ5O19ꢂ0
and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 nanoparticles with different Ti-
content. The Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 nanoparticles exhibited a
PL signal with a similar curve shape, showing that the
presence of Ti did not result in a new PL phenomenon. The
Sb1ꢂ5Sn8ꢂ5O19ꢂ0 nanoparticles exhibited wide and strong PL
signals in the range of 380–550 nm, with the maximum
excitation wavelength of 450 nm. The Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0
curve intensities weakened gradually depending on the
extent of Ti-insertion because of a transition from the
trapped electrons to the Ti ions. Generally, owing to
the formation of new oxygen vacancies by metal-doping,
the photogenerated electrons in the conduction band ini-
tially reach the vacant space and then recombine with
the photogenerated holes in the valance band to pro-
duce fluorescence emission.20 The reduced PL intensity
for the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 catalysts was possibly a result
of the defects (regardless of the chemical stoichiomet-
ric or non-stoichiometric) generated by the incorporation
of Ti ions into the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 structure, accelerating
the electron transfer and hindering the electron–hole pair
recombination on the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 surface. In other
words, the recombination of the photogenerated carriers
via a transition from the structural defects or oxygen
vacancies to the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 conduction band was sup-
pressed by doping with Ti-ions. This result was also con-
sistent with the results of the conductivity tests.
Photocatalytic reduction processes require a catalyst,
feed (CO2 and H2O), and a light source. All photocatalytic
CO2 reduction experiments were repeated at least thrice
and the aggregated results have been reported. Figure 7
shows the effects of Ti-insertion into Sb1ꢂ5Sn8ꢂ5O19ꢂ0 on the
photocatalytic reduction of CO2 with H2O vapor to CH4.
The mechanism for methane production from CO2 reduc-
tion is divided into three sub-processes.21
1. Proton production from H2O photodecomposition
H2Oads +h+ → ꢁOH+H+ꢁ
ꢁOH+3h+ → 3H+ +O2
J. Nanosci. Nanotechnol. 18, 6369–6377, 2018
6373

Photoreduction of Carbon Dioxide to Methane Over Sb1.5Sn8.5−xTixO19.0 with High Conductivity
Do et al.
Figure 4. (A) UV-Vis absorption spectra and (B) Tauc plots of the synthesized Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders.
2. CO radical production from CO2 photo cleavage
3. Methane production from photosynthesis between the
CO radical and proton
CO2ads +e− → ꢁCO−2 ꢁ
ꢁCO+H+ +e− → ꢁC+OH−ꢁ
ꢁCO−2 → CO+OH−ꢁ
−
ꢁ
ꢁ
+
−
ꢁ
IP: 62.4.55.113 On: Tue, 19 Jun 2018 19:21:13
CO+e → CO
C+H +e → CHꢁ
Copyright: American Scientific Publishers
Delivered by Ingenta
Figure 5. Cyclic voltammetry measurements of the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders.
J. Nanosci. Nanotechnol. 18, 6369–6377, 2018
6374

Do et al.
Photoreduction of Carbon Dioxide to Methane Over Sb1.5Sn8.5−xTixO19.0 with High Conductivity
dependent on the production of the intermediate CO.
Based on the reaction summary and experimental obser-
vations, it can be concluded that it is possible to pro-
duce hydrocarbons using photocatalysts with appropriate
reduction potentials. With Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0, Ti-insertion
into the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 framework led to a higher pho-
toactivity for CO2 photoreduction. The yield of CH4
was low over pure ATO, but increased gradually with
increasing Ti-content in the photocatalyst. The maximum
yield of CH4 obtained over the Sb1ꢂ5Sn7ꢂ5Ti1ꢂ0O19ꢂ0 cat-
alyst after 8 h was 30 ꢈmol/g-catal · L−1, as compared
to 5.0 ꢈmol/g-catal· L−1 using Sb1ꢂ5Sn8ꢂ5O19ꢂ0. However,
the yield was lower when the photocatalyst with more Ti
content, i.e., Sb1ꢂ5Sn7ꢂ0Ti1ꢂ5O19ꢂ0, was used. On the other
hand, the yield of CH4 was 12 ꢈmol/g-catal · L−1 after
irradiation for 8 h when pure TiO2-anatase powder, which
had been synthesized by a hydrothermal method in a pre-
vious study,22 was used for comparison. Therefore, the
optimal Ti-content in Sb1ꢂ5Sn8ꢂ5O19ꢂ0 was 13 mol% as
compared to the amount of Sn, indicating that the effect
of Ti-insertion into Sb1ꢂ5Sn8ꢂ5O19ꢂ0 framework was more
extensive.
Based on the results of CV, and UV-Vis and PL spec-
troscopic analysis, two mechanistic models have been sug-
gested (Scheme 1). The CO2/CH4 redox potentials for all
samples were included to −0.24 V.23 If the metal dopants
were substituted into the metal sites of the tetragonal
Figure 6. PL spectra of the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 and Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0
nanoparticles with different Ti content.
ꢁCH +H+ +e− → ꢁCH2ꢁ
ꢁCH2 +H+ +e− → ꢁCH3ꢁ
ꢁCH3 +H+ +e− → CH4
The photogenerated electrons on the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0
catalysts by UV-radiation induce CO2 reduction to produce
CO2 radicals, whereas the holes react with the adsorbed
H2O molecules to produce hydroxyl radicals. The inter-
mediate photogenerated species underwent several differ-
ent reactions, producing CO and CH4. CH4 was confirmed
to be produced from methyl radicals (ꢂCH3ꢀ because all
of the abovementioned products were detected experi-
mentally. The formation of these radicals was directly
framework, the operative mechanism would be close to
IP: 62.4.55.113 On: Tue, 19 Jun 2018 19:21:13
Copyright: American Scientific Publishers
that shown in Model (a). As Sb1ꢂ5Sn8ꢂ5O19ꢂ0 has a very
Delivered by Ingenta
short band gap, it is expected that its photocatalytic activ-
ity would be lesser because of the rapid recombination of
the electron and holes. On the other hand, the LUMO ener-
gies of the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 samples shifted to higher
positions with increasing Ti-insertion, leading to larger
band gaps. If Model (b) is accepted under the assumption
that the metal ingredients are partially isolated from the
tetragonal framework of Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0, then because
of the ability of Ti to trap and release electrons into
the conduction band of SnO2, there may be a signifi-
cant increase in the CH4 yield and a decrease in their
recombination with holes as also seen from the PL spec-
tra. Therefore, Ti-insertion can promote the separation of
the charge pairs in the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 photocatalyst for
higher CO2 reduction efficiency. The photoexcited elec-
trons in the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 conductance band are accepted
and transferred efficiently to CO2, functioning as an effi-
cient reducing agent. The electrons can also be scav-
enged by O2, which in turn produces the superoxide ion
(O−2 ꢀ. The positive holes in the Sb1ꢂ5Sn8ꢂ5O19ꢂ0 valance
band can be trapped by the OH− or H2O species. These
processes promote the separation of the photogenerated
electron–hole pairs (e−/h+ꢀ and accelerate the reduction
of CO2. This explains the enhanced catalytic performance
of the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders. In general, inserting
metals as defects can help trap the photogenerated elec-
trons and hinder the charge recombination, which is one
Figure 7. Photocatalytic CO2 reduction with H2O vapor for CH4 pro-
duction according to the level of Ti insertion over the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0
catalysts.
J. Nanosci. Nanotechnol. 18, 6369–6377, 2018
6375

Photoreduction of Carbon Dioxide to Methane Over Sb1.5Sn8.5−xTixO19.0 with High Conductivity
Do et al.
Scheme 1. Proposed models based on the CV, UV-Vis and PL results.
24
IP: 62.4.55.113 On: Tue, 19 Jun 2018 19:21:13
of the limiting factors in general photoactivity. How-
References and Notes
Copyright: American Sc1ie. nLt.ifCichePn,uMbl.isEh. Gerrasham, G. Li, D. R. Gentner, N. M. Dimitrijevic,
ever, an excess of the inserted metal content reduces
Delivered by Ingenta
and K. A. Gray, Thin Solid Films 517, 5641 (2009).
2. Q. Li, L. Zong, C. Li, and J. Yang, Appl. Surf. Sci. 314, 458
(2014).
3. J. H. Lee, S.-I. Kim, S.-M. Park, and M. Kang, Ceram. Inter.
43, 1768 (2017).
the photoactivity, possibly because of the establishment
of charge recombination centers inside the semiconductor
structure.25
4. O. Ola and M. M. Maroto-Valer, J. Catal. 309, 300 (2014).
5. J. Tan, L. Shen, X. Fu, W. Hou, and X. Chen, Dye Pigment. 61, 31
(2004).
6. J. Montero, J. Herrero, and C. Guillen, Sol. Energ Mat. Sol. C.
94, 612 (2010).
7. J. Montero, C. Guillén, and J. Herrero, Surf. Coat. Technol. 211, 37
(2012).
8. Y. S. Kim, B. K. Yu, D. Y. Kim, and W. B. Kim, Sol. Energ Mat.
Sol. C. 95, 2874 (2011).
9. G. Guzman, B. Dahmani, J. Puetz, and M. A. Aegerter, Thin Solid
Films 502, 281 (2006).
10. Z. Zhang, C. Ma, L. He, M. Huang, L. Yu, and Y. Zhang, Ceram.
Inter. 40, 10309 (2014).
11. E. K. Shokr, M. M. Wakkad, H. A. A. El-Ghanny, and H. M. Ali,
J. Phys. Chem. Solid 61, 75 (2000).
12. J. Montero, C. Guillén, and J. Herrero, Thin Solid Film 519.0, 7564
(2011).
13. H. J. Jeon, M. K. Jeon, M. Kang, S. G. Lee, Y. L. Lee, Y. K. Hong,
and B. H. Choi, Mat. Lett. 59, 1801 (2005).
14. F. J. Berry and C. Greaves, J. Chem. Soc., Dalton Trans. 12, 2447
(1981).
15. V. K. Puthiyapura, M. Mamlouk, S. Pasupathi, B. G. Pollet, and
K. Scott, J. Power Sources 269, 451 (2014).
16. K. Ravichandra and K. Thirumurugan, J. Mat. Sci. Technol. 30, 97
(2014).
4. CONCLUSION
Novel
photocatalysts
with
high
conductivity,
Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0, comprised of three metal oxide
semiconductors, in which Ti was partially incorporated
into the tetragonal framework of Sb1ꢂ5Sn8ꢂ5O19ꢂ0 (ATO),
were prepared. These materials are very different from
the previously reported catalysts based on the TiO2
anatase structure. The band gaps and PL intensities of the
Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders increased and decreased dra-
matically with the increasing amount of Ti, respectively.
The photoreduction of CO2 to CH4 increased remarkably
over the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 catalysts. Two models were
proposed to explain the enhanced photocatalytic activity
of the Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders. It was concluded that
the effective charge separation and inhibited recombi-
nation of the photogenerated electron–hole pairs in the
Sb1ꢂ5Sn8ꢂ5−xTixO19ꢂ0 powders were responsible for the
improved catalytic activity.
Acknowledgments: This research was supported by the
216A061008 Yeungnam University Research Grant, for
which the authors are very grateful.
17. E. G. Goh, X. Xu, and P. G. McCormick, Script Materials 78–79, 49
(2014).
6376
J. Nanosci. Nanotechnol. 18, 6369–6377, 2018

Do et al.
Photoreduction of Carbon Dioxide to Methane Over Sb1.5Sn8.5−xTixO19.0 with High Conductivity
18. V. Trevisan, A. Olivo, F. Pinna, M. Signoretto, F. Vindigni,
G. Cerrato, and C. L. Bianchi, Appl. Catal. B 160–161, 152
(2014).
21. M. Tahir and N. S. Amin, Appl. Catal. B 142–143, 512 (2013).
22. B. S. Kwak, J. Chae, J. Kim, and M. Kang, Bull. Kor. Chem. Soc.
30, 1047 (2009).
19. I. Trapp, S. González-Gallardo, S. Hohnstein, D. Garnier,
P. Oña-Burgos, and F. A. Breher, Dalton Trans. 43, 4313
(2014).
20. B. Neppolian, Y. Kim, M. Ashokkumar, H. Yamashita, and H. Choi,
J. Hazardous Mater. 182, 557 (2010).
23. M. Tahir and N. S. Amin, Energy Conver. Manage. 76, 19.04 (2013).
24. I. N. Martyanov, T. Berger, O. Diwald, S. Rodrigues, and K. J.
Klabunde, J. Photochem. Photobiol. A 212, 135 (2012).
25. L. L. Tan, W. J. Ong, S. P. Chai, and A. R. Mohamed, Chem.
Commun. 50, 6923 (2014).
Received: 25 July 2017. Accepted: 21 November 2017.
IP: 62.4.55.113 On: Tue, 19 Jun 2018 19:21:13
Copyright: American Scientific Publishers
Delivered by Ingenta
J. Nanosci. Nanotechnol. 18, 6369–6377, 2018
6377