F. Fresno, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
highly reducing conduction band in order for excited electrons to be
transferred to the highly stable and little reactive CO2 molecule. This
requirement challenges the utilization of classical photocatalysts, which
are usually characterized by a strongly oxidizing valence band holes,
while conduction band electrons lie at energy levels very close to or
even below the ones corresponding to the reduction potentials involved
in CO2 conversion [7]. In this respect, an interesting class of semi-
conductors is that composed of ternary oxides with perovskite structure
like alkali-metal niobates and tantalates, which show conduction bands
with high enough energy for transferring electrons to CO2 while, at the
same time, they posess valence bands energy levels such that, upon
irradiation, photo-produced holes have sufficiently positive reduction
potential as to oxidize the water molecule [13,14]. In addition, the
perovskite layered structure has been demonstrated in the last years to
be particularly suitable for photocatalytic applications due to its charge
separation properties and the possibly separated reduction and oxida-
tion sites, which minimizes the extent of back reactions [15]. Thus,
several works have reported remarkable results with niobates and
tantalates for photocatalytic production of hydrogen via water splitting
or alcohol photoreforming [16–19] and, although scarcer, some works
have revealed the potential of these structures for CO2 photocatalytic
conversion [20–24]. Recently, we reported the performance of sodium
niobate and tantalate for the production of (mainly) carbon monoxide
from CO2 [25]. Both catalysts showed promising results in terms of
selectivity towards CO2 reduction against water reduction, but suffered
from low surface areas related to the solid-sate reaction synthesis,
which limited the overall gas-phase CO2 conversion. Therefore, a
modification of the synthesis method towards higher surface areas is
reported here in order to overcome such limitation in sodium niobate
and tantalate perovskites. As mentioned above, these materials show
particular interest in terms of a very reducing conduction band, in the
case of the latter, and a good compromise between light absorption and
reducing power, in the case of the former. In addition, an intermediate
composition has been synthesized in order to tune the band gap and the
conduction band level.
2. Experimental
2.1. Synthesis of photocatalysts
All the reagents were analytically pure, commercially available, and
used without further purification. NaNbO3 was prepared following a
hydrothermal route [19]. 50 mL of a 10 M NaOH solution were placed
in a 100 mL Teflon vessel. Nb2O5 powder (5.87 mmol) was added, after
which the mixture was magnetically stirred for 2 h. Then the vessel was
fitted into a stainless steel autoclave and submitted to hydrothermal
treatment at 230 °C for 24 h. The resulting white powder was separated
by centrifugation, washed with deionized water until the washate
reached neutral pH, dried overnight at 100 °C and homogenized in an
agate mortar.
Silver was deposited on the surface of NaNbO3 through wet im-
pregnation. For that purpose, the appropriate amount of silver nitrate
for a Ag load of 0.1 or 0.5 wt.% with respect to the oxide was dissolved
in 25 mL deionized water and transferred to a 50 mL round-bottomed
flask together with 0.5 g NaNbO3. The mixture was then submitted to a
slow evaporation in a rotatory evaporator at 60 °C, 200 mbar and
185 rpm. The obtained solid was dried at 100 °C overnight, homo-
genized in an agate mortar and calcined in static air at 400 °C for 4 h.
2.2. Characterization techniques
The chemical composition of the photocatalysts was quantified
using an Induced Coupled Plasma Atomic Emission Spectrometer (ICP-
AES, Perkin Elmer 2300 DV) after digesting the samples in a mixture of
HNO3 and HF. Powder X-ray diffraction (XRD) patterns were recorded
in a Panalytical EMPYREAN diffractometer using CuKα radiation (λ
=1.54178 Å) at a scanning rate of 0.01°s−1. Structural analysis was
carried out by Rietveld refinement using the X’Pert Highscore Plus
software. Crystallite size was estimated from the width diffraction
peaks by comparing to a LaB6 standard sample for instrumental peak
broadening. N2 adsorption–desorption isotherms were recorded at
−196 °C in a QUADRASORB instrument after degassing at 200 °C under
vacuum. Diffuse reflectance UV–vis spectra were recorded in a UV/Vis/
NIR Perkin Elmer Lambda 1050 spectrometer. Photoluminescence ex-
periments were carried out with a Fluorescence Spectrometer Perkin
Elmer LS55, using an excitation wavelength of 300 nm and a cut-off
filter at 350 nm. TEM images were taken using a PHILIPS TECNAI 20 T
instrument, working with a tungsten filament at 200 kV and equipped
with an EDX spectrometer.
On the other hand, as mentioned above, versatility in terms of
controllable selectivity towards different products can be one of the
characteristics driving the potential of photocatalysis for CO2 utiliza-
tion. In this respect, the concourse of metal co-catalysts has been shown
to strongly influence this selectivity, which can be then tuned not only
by introducing metal nanoparticles into the photocatalytic system, but
also by modifying the nature, size or shape of those nanoparticles [3].
Thus, some metals such as platinum, gold, silver or copper, when
supported on the semiconductor photocatalyst, have been reported to
modify the reaction selectivity from the preferential production of CO
observed with the bare semiconductor to the selective evolution of
highly reduced products such as methanol [26–28] or methane
relatively low price and its catalytic and photophysical properties,
which influence not only the surface properties of the photocatalytic
system, by acting as catalytic active site enabling electron withdrawal
and transfer [32,36], but also by displaying surface plasmon resonance
effects that modify the light management capabilities of the semi-
Electrochemical measurements were performed in a three-electrode
glass cell with a quartz window containing an aqueous solution of 0.5 M
Na2SO3, being the studied catalysts the working electrodes. A suspen-
sion of the catalyst powder (1 mg in 10 ml ethanol) was deposited by
drop-casting on an ITO-covered glass, and then sintered at 200 °C.
Platinum wire and Ag/AgCl were used as counter and reference elec-
trodes, respectively. Voltage, current density and electrochemical im-
pedance spectroscopy were measured in dark and under illumination
with a potentiostat-galvanostat PGSTAT204 provided with an in-
tegrated impedance module FRAII. A Hamamatsu high power UV LED
centred at 365 nm was used as the light source.
Ambient-pressure X-ray photoelectron spectroscopy (APXPS) stu-
dies were carried out at the NAPP branch from CIRCE, an undulator
beamline with 100−2000 eV photon energy range at ALBA
Synchrotron Light Source [37]. Acquisition was performed using a
PHOIBOS 150 NAP electron energy analyzer (SPECS GmbH) equipped
with four differential pumping stages and a set of electrostatic lenses
which enable XPS measurements with the sample at pressures from
ultra-high vacuum (UHV, with a base pressure of 2 − 5 × 10−10 mbar)
and up to 20 mbar. The take-off angle was approximately 50° and the
angle between the analyzer axis and the incoming synchrotron radia-
tion horizontal linear polarization vector was 54.7°, the magic angle.
The spectra were acquired with 20 μm exit slit and the diameter of the
tion of hydrothermally synthesized niobium, tantalum and mixed-metal
perovskites by decoration with silver nanoparticles for their study as
CO2 conversion photocatalysts, in order to shed light on the mechan-
isms that rule their behaviour in this application. For that purpose,
photoreaction studies have been combined with extensive character-
ization by different techniques, including in-situ studies by ambient-
pressure x-ray photoelectron spectroscopy, which, to the best of the
authors’ knowledge, is reported for the first time for such photocatalytic
materials.
2