Partial oxidation of methane to methanol using bismuth-based photocatalysts

Article abstract of DOI:10.1021/cs500821r

Bismuth-based photocatalysts, Bi₂WO₆, BiVO₄, and coupled Bi₂WO₆/TiO₂-P25, have been synthesized by a facile hydrothermal method, characterized, and evaluated for the first time for the selective photooxidation of methane to methanol. Several conditions were used in order to better comprehend the reaction mechanism. The obtained BiVO₄is, among the others, the most promising photocatalyst for this reaction, displaying higher CH₃OH selectivity and being more stable than the others. When Bi₂WO₆was coupled with TiO₂, the methane conversion increased; however, overoxidation of CH₄to CO₂predominates. A similar effect is observed when electron scavengers such as O₂or Fe³⁺were introduced in the photoreactor as a result of the formation of highly oxidant radicals. (Chemical Equation Presented).

Full text of DOI:10.1021/cs500821r

Research Article
pubs.acs.org/acscatalysis
Partial Oxidation of Methane to Methanol Using Bismuth-Based
Photocatalysts
Sebastian Murcia-Lopez,*,† Katherine Villa,† Teresa Andreu,*,† and Joan R. Morante†,‡
́
́
†Catalonia Institute for Energy Research (IREC), Jardins de les Dones de Negre 1, 08930 Sant Adria
‡Department of Electronics, University of Barcelona (UB), Marti i Franques
1, 08028 Barcelona, Spain
̀ ̀
de Besos, Spain
̀
ABSTRACT: Bismuth-based photocatalysts, Bi2WO6, BiVO4, and
coupled Bi2WO6/TiO2−P25, have been synthesized by a facile
hydrothermal method, characterized, and evaluated for the first time
for the selective photooxidation of methane to methanol. Several
conditions were used in order to better comprehend the reaction
mechanism. The obtained BiVO4 is, among the others, the most
promising photocatalyst for this reaction, displaying higher CH3OH
selectivity and being more stable than the others. When Bi2WO6 was
coupled with TiO2, the methane conversion increased; however,
overoxidation of CH4 to CO2 predominates. A similar effect is
observed when electron scavengers such as O2 or Fe3+ were introduced in the photoreactor as a result of the formation of highly
oxidant radicals.
KEYWORDS: photocatalysis, selective oxidation, Bi2WO6, BiVO4, photo-Fenton
lower band gap value, which makes it a good alternative in
visible light processes. Initially, Taylor et al.6,7 proposed the
photocatalytic conversion of methane into methanol in a
continuous process with La-WO3 and UV light, at ∼90 °C. An
electron scavenger (methyl viologen dichloride hydrate) and
hydrogen peroxide as hydroxyl radical source were used in their
studies. Later, Gondal et al.8 also carried out the photo-
conversion of methane into methanol with WO3, in a batch
setup with a laser irradiation (visible light). Hydrogen peroxide
and Fe3+ as electron scavenger were evaluated.
Besides WO3, other kinds of materials as mesoporous silicas
(V-containing MCM 41) and zeolites9−12 have also been
reported as photocatalysts for this reaction, by using NO and
molecular O2 as oxidizers. On the other hand, TiO2, the most
used semiconductor in photocatalytic applications, has attracted
limited attention for this reaction. For instance Gondal et al.13
have reported WO3 as a better photocatalyst for methanol
generation than rutile-TiO2 and NiO. Other studies with Pt-
TiO2 materials, though, have been focused in steam reforming
of methane (and methanol) for the generation of H2 and
CO2.14,15
1. INTRODUCTION
Nowadays, the development of new technologies for the direct
conversion of methane into other products is an important
challenge that has attracted much attention. As the main
component of natural gas, methane is very abundant and
possesses a high calorific power, so it is commonly used as
direct fuel.1 Besides this, the possibility of directly transforming
methane into oxygenated products, such as methanol and
formaldehyde, and even into longer chain compounds (e.g.,
ethane) also becomes an interesting alternative for its
revalorization. Methanol, for instance, is a desirable product
because it is an important chemical intermediate, used for
obtaining other oxygenated compounds such as formaldehyde
and acetic acid.2 For these reasons, several efforts have been
made in heterogeneous and homogeneous catalysis with diverse
types of materials.3,4 These studies, however, usually imply high
temperatures and pressures or more complex systems, besides
displaying poor selectivity associated with the continuous
oxidation of methanol into CO and CO2.
The photocatalytic processes allow promoting difficult
reactions (even thermodynamically limited1) at mild con-
ditions, by taking advantage of photoenergy instead of thermal
activation. In this sense, photocatalysis arises as a promising
alternative for the selective oxidation of methane into
oxygenated products. Some approximations have been
evaluated in literature, using several types of materials and
oxidizers. From the evidence of the photochemical conversion
of methane with water vapor at low temperature under certain
irradiation conditions,5 some authors have reported the
photocatalytic oxidation of methane, mediated by semi-
conductors in the presence of H2O. WO3 has been one of
the most commonly studied materials, mainly because of its
In recent years, bismuth-based materials have emerged as
alternative photocatalysts to binary oxides as WO3 and TiO2.
Specially considered for visible-light processes, it has been
proposed that new hybridized band structures with Bi 6s and O
2p orbitals can be formed. Thus, besides a decrease in the band
gap, the charge carrier mobility could be enhanced due to a less
localized character of the s orbitals.16 Among them, Bi2WO6
Received: June 12, 2014
Revised: July 18, 2014
Published: July 28, 2014
© 2014 American Chemical Society
3013
dx.doi.org/10.1021/cs500821r | ACS Catal. 2014, 4, 3013−3019

ACS Catalysis
Research Article
and BiVO4, with narrow band gaps (2.8 and 2.4 eV,
respectively), have been two of the most studied oxides for
different photocatalytic applications, like water splitting and
pollutant degradation under visible light. For the latter, many
efforts have been made for developing materials with controlled
morphologies.17−20 Despite being promising, only few reports
deal with their application as photocatalysts in selective
reactions,21−23 besides some reports dealing with the CO2
photoreduction into hydrocarbons with Bi2WO6 and
BiVO4.24−26
Moreover, to the best of our knowledge, these materials have
not been yet evaluated in the selective photooxidation of
methane to methanol. For this reason, in the present work,
several bismuth-based materials have been studied for this
reaction. At first, Bi2WO6 has been evaluated by changing some
reaction conditions, including the addition of Fe3+ as electron
scavenger. After this, given the possible drawbacks associated
with the band configuration in this oxide, two additional
materials have been synthesized: a coupled Bi2WO6/TiO2−P25
system and BiVO4, with expected more negative redox
potentials in comparison to Bi2WO6. These materials have
been characterized and evaluated under the best reaction
conditions.
continuously recirculated through the inner jacket of the
reactor, in order to control both the temperatures of the lamp
and of the reaction system (∼55 °C). The outlet gas tube was
connected to a six port valve with a loop, used to allow the
sample injection to a gas chromatograph (Shimadzu GC-2010,
equipped with an Agilent J&W HP-PLOT Q column and TCD
and FID detectors), in which the gas composition was analyzed.
During each test, a suspension of Milli-Q water (300 mL)
with the corresponding amount of catalyst (1g/L) was added to
the reactor and kept in suspension by mechanical stirring. A
calibrated gas mixture of CH4 in He (20%) was continuously
sparged (∼22.4 mL/min) through the reactor in the dark for 30
min, after which the illumination was turned on and a
periodical injection of the outlet gas was made for the analysis
of products.
In the additional tests with Fe3+, FeCl3 was used as salt
precursor (1 mM), and 1 M H2SO4 was used to adjust the pH
value to ∼2.5. Then, the same amount of catalysts was added to
the system, and the reaction was carried out following the
procedure described above.
3. RESULTS AND DISCUSSION
3.1. Characterization. The XRD measurements of the
different materials can be observed in Figure 1. In single
2. EXPERIMENTAL SECTION
2.1. Preparation of the Materials. Bi2WO6 was
synthesized by following a procedure reported elsewhere.27
Briefly, two solutions with the corresponding amounts of
Bi(NO3)3·5H2O and Na2WO4·2H2O were prepared. The Bi
precursor (0.005 mol) was dissolved in 5 mL of glacial acetic
acid and the W solution (0.0025 mol) in 45 mL of Milli-Q
water. These solutions were mixed to form a white suspension
which was kept under stirring for 1 h. Then, this mixture was
transferred to a Teflon liner inside a stainless steel autoclave
and submitted to hydrothermal treatment at 140 °C for 20h.
Finally, the precipitate was filtered, repeatedly washed, dried
overnight, and calcined at 300 °C for 4h.
An analogous procedure was used for the preparation of
BiVO4. In this case, the vanadium precursor was anhydrous
NaVO3, which was added in a stoichiometric proportion (0.005
mol) to Bi.
The coupled Bi2WO6/TiO2−P25 sample was synthesized
with the same method. The corresponding amount of
commercial TiO2 (Evonik, P25) was added to the white
suspension of Bi−W prior to the hydrothermal treatment, with
a TiO2/Bi2WO6 ratio of 50% wt. This material was named
BW−Ti.
2.2. Characterization. The crystalline phase composition
was estimated by X-ray diffraction (XRD), using a Bruker D8
Advance diffractometer equipped with a Cu Kα radiation
(0.15417 nm) source, a LYNXEYE super speed detector and a
Ni filter. Crystallite sizes of the different phases were
determined from the line broadening of the corresponding
XRD peaks by using the Scherrer equation.
BET surface area and porosity measurements were carried
out by N2 adsorption in a Micromeritics TriStar II instrument.
The morphology of the samples was observed in a Zeiss
Auriga FESEM microscope.
2.3. Photocatalytic Oxidation Tests. For the photo-
catalytic oxidation of methane to methanol, a commercial
photochemical reactor (Ace Glass) was used. This setup was
equipped with an immersion medium-pressure mercury lamp
(450W, Ace Glass) with UVC-visible radiation. Cold water was
Figure 1. XRD patterns of: (A) Bi2WO6, (B) BW−Ti, (C) TiO2−P25,
and (D) BiVO4.
Bi2WO6, only peaks corresponding to the russellite ortho-
rhombic phase (JCPDS 39-0256) are seen. Some small peaks
related to the anatase phase (JCPDS 21-1272) are also present,
when this material is coupled to TiO2−P25 as in the case of the
BW−Ti sample. In this sense, the contribution of both single
structures is evident.
In the case of the BiVO4 sample, the pattern in Figure 1D
corresponds to the scheelite monoclinic phase (JCPDS 14-
0688), which has been usually described as the most
photoactive in comparison to the tetragonal phases.20,28,29
The crystallite sizes calculated by using the Scherrer equation
are collected in Table 1. Bi2WO6 and BW−Ti display similar
sizes for the main russellite peak, with a small increment
associated with the presence of TiO2−P25. On the other hand,
BiVO4 presents the highest crystallite size. This fast growth of
BiVO4 particles when prepared by hydrothermal synthesis has
been already reported in literature, having a particular influence
on the surface area.20
3014
dx.doi.org/10.1021/cs500821r | ACS Catal. 2014, 4, 3013−3019

ACS Catalysis
Research Article
has the lowest one. From the point of view of the particle sizes,
the values correspond well to the expected behavior among the
four materials.
Table 1. SBET and Crystallite Size Values (Calculated from
Scherrer Equation) for the Different Samples
sample
SBET (m2g−1
)
crystallite size (nm)
3.2. Photocatalytic Activity. 3.2.1. Blank Tests. Prior to
the photocatalytic tests, several blank tests without catalyst have
been performed. In the test with only CH4 under irradiation, no
product was formed. Though, in the presence of water (300
mL), certain amounts of CH3OH, CO2, and C2H6 were
observed under illumination. The corresponding conversion
and selectivity results are presented in Table 2. Close to 2
μmol·h−1 of CH3OH can be obtained after 120 min of
irradiation. This process through a water photolysis mechanism
in gas phase has been previously proposed by some authors,5,7
as indicated in Scheme 1. In this case, under VUV light (at λ ≥
185 nm), hydroxyl radicals can be produced from water
photolysis and oxidize CH4 into CH3OH, through the
formation of methyl radicals and their reaction with H2O, as
shown in eqs 2 and 3. Though, besides methanol, C2H6 can
also be produced from the reaction between two methyl
radicals as in eq 4, which differs from those results of Ogura and
Kataoka, in which almost negligible combination of methyl
radicals was observed.5 This situation can be explained from the
different used configurations, including the use of a different
irradiation source. Finally, in addition to these two products, a
relatively high amount of CO2 was produced, indicating that
some overoxidation process has taken place, through any of
some potential oxidizing paths: direct oxidation with the
nonselective ·OH radicals or through the formation of
hydrogen peroxide or oxygen, as in eq 5.
Bi2WO6
BW−Ti
BiVO4
30
41
2
8
10
28
The morphologies of the samples observed by SEM are
presented in Figure 2. In this sense, there are important
differences in the shapes and sizes of the different materials.
Bi2WO6 (Figure 2A,B) displays an anisotropic growth in order
to form nanoplates that self-assemble into 3D hierarchical
structures through an Ostwald ripening process, as reported in
previous works.17 In the coupled BW−Ti system (Figure 2C),
the presence of TiO2−P25 during the hydrothermal treatment
tends to inhibit the self-assembling mechanism. Thus, a
heterogeneous morphology is obtained, with roundish particles
that can be associated with TiO2−P25 mixed with small
nanoplates related to Bi2WO6. EDX analysis of this material
indicates a final TiO2/Bi2WO6 ratio of ca. 42%, more or less
close to the nominal one. Finally, BiVO4 (Figure 2D) consists
of large irregular polyhedron of several sizes (1−2 μm), some
of them with preferential exposure of (001) planes, suggesting
also an anisotropic growth confirmed from the higher ratio of
intensities between the (004) and the (121) peaks from XRD
results.
In the case of single Bi2WO6, the difference between
crystallite (estimated by the Scherrer equation) and particle
sizes (seen by SEM) can be explained from the hierarchical
growth, in which the large 3D structures are formed by smaller
nanoplates.
The measurements of the BET area are displayed in Table 1.
It is possible to observe that the coupled BW−Ti system
exhibits the highest specific surface with an intermediate value
between Bi2WO6 and TiO2−P25 (∼51 m2 g−1), whereas BiVO4
3.2.2. Tests with Bi2WO6. Afterward, the photocatalytic
process, which is represented in Scheme 2, was evaluated with
Bi2WO6 under several reaction conditions. In this case, the
irradiation of the semiconductor leads to the formation of
hole−electron pairs in the valence and conduction bands,
respectively. These holes (h+) are able to initiate the hydroxyl
radical formation by oxidizing adsorbed water onto the
Figure 2. SEM micrographs of: (A, B) Bi2WO6, (C) BW−Ti, and (D) BiVO4.
3015
dx.doi.org/10.1021/cs500821r | ACS Catal. 2014, 4, 3013−3019

ACS Catalysis
Research Article
Table 2. Selectivity and Total Conversion Results after 120 min of Illumination during the Different Tests with and without
a
Photocatalyst, at 55°C
selectivity (%)
sample
CH3OH
CO2
C2H6
total conversion (%)
CH3OH rate (μmol·h−1
)
without catalyst (H2O)
Bi2WO6
22.3
27.6
59.7
63.4
18.1
9.0
0.8 (0.8)
1.4 (1.6)
2.1
4.6
a
In parentheses, the conversion values at 30 min.
conduction band has also to be taken into account. For
instance, if the semiconductor displays the proper band
structure, it would be able to reduce water into H2 (eqs 8−
9). However, from the theoretical band structure proposed in
literature for Bi2WO6, these processes are not expected to take
place, as the electrons in its conduction band might not have
the necessary redox potentials. The positions of the valence and
conduction bands at pH = 0 for several materials are illustrated
in Figure 3 (values extracted from literature30−33). It is
Scheme 1. Photolytic Path
Scheme 2. Photocatalytic Path
photocatalyst. The subsequent reactions for obtaining methanol
and other products would be the same already represented in
eqs 2−5 in Scheme 1.
Figure 3. Conduction and valence bands positions for several materials
at pH = 0, compared to some redox potentials. Some values were
extracted from references.30−33
At 55 °C, a CH3OH rate of 4.6 μmol·h−1 was obtained with
Bi2WO6 after 120 min, higher than the one obtained in the
blank test without catalyst. Moreover, as seen in Table 2, the
methanol selectivity increases at the expense of a decrease in
that of ethane. This result suggests that the presence of the
photocatalyst might be important in order to favor the reaction
between methyl radicals and water, thus inhibiting their
combination to produce ethane. From another point of view,
the presence of the photocatalyst also enhances the methane
conversion since the first 30 min of reaction, in comparison to
the blank test.
important to keep in mind that as many aspects influence the
solid−liquid interaction, different reported values can be found
at varied conditions. In many cases, some variations might
come, for instance, from differences in the estimated band gap
values. Usually, calculations at several pH values are done by
using the Nernst equation. Thus, more negative redox
potentials can be expected at higher pH values (though the
H2O/H2 potential would also be affected). Among the
materials presented in Figure 3, BiVO4 and TiO2 have
conduction bands closer to the potential for water reduction;
therefore, from the thermodynamic point of view, they are
more likely to evolve H2 from H2O, than Bi2WO6 and WO3.
On the basis of the above considerations, the second
additional test was performed with Bi2WO6 by continuously
supplying O2 into the system in order to elucidate the effect
that this species may have as electron scavenger. In this case,
the CO2 production was greatly increased (almost 20-fold),
being an important indicator of highly oxidizing conditions. On
the opposite, C2H6 formation was completely inhibited, while
CH3OH rate was more or less similar to that obtained without
O2. Besides these results, it is important to point out that no
change in the color of the suspension was observed at the end
of the test. This is a qualitative indicator of the fact that some
electron trapping must take place in the presence of O2.
Even when no improvement in terms of methanol
production is observed under these conditions, some
Two additional tests were carried out with Bi2WO6 in order
to determine the effect of different conditions. In the first one,
the reaction temperature was increased to 90 °C, leading to a
significant increase in the C2H6 formation (almost 3-fold),
although CH3OH was almost inhibited. This effect might be
explained because of the expected higher amount of water
vapor inside the reactor at this temperature, thus enhancing the
photolytic formation of methyl radicals and their combination
through reaction in eq 4. Hence, the decrease on the methanol
rate is justified, as the ethane formation is a competitive
reaction in which methyl radicals are consumed.
In the case of Bi2WO6, an important change in the color of
the suspension was observed at the end of the tests. After a
short time of exposure to air, the suspension easily recovers its
original color, thus indicating this is a reversible process.
Besides the oxidation reactions through the holes, the
reductive path through the concomitant electrons in the
3016
dx.doi.org/10.1021/cs500821r | ACS Catal. 2014, 4, 3013−3019

ACS Catalysis
Research Article
interesting aspects may arise from this test. Some authors have
reported experimental and theoretical evidence about the
thermodynamic drawbacks for the generation of superoxide
radicals (O2·−) with Bi2WO6 (eq 10).34 However, additional
reactions involving multielectron transfer to molecular oxygen
have been proposed to occur in photodegradation processes
with Bi2WO6. In fact, Sheng et al.32 have recently found that
H2O2 and ·OH radicals can be formed through multielectron
reduction of O2 with Bi2WO6, despite the expected lower rates
associated with this kind of processes. In this sense, the
Scheme 3. Fenton and Photo-Fenton Paths
further play an important role in the overall process, as will be
shown later.
On the basis of these considerations, two final tests were
carried out by adding Fe3+ to the system: a blank and a
photocatalytic test in the presence of Bi2WO6, under the same
standardized conditions (55 °C, 1g·L−1 of catalyst, 1 mM Fe3+,
pH 2.5).
•
formation of perhydroxyl radical (HO2 ) and H2O2 are more
plausible than the direct formation of O2•−, as these reactions
•
have more positive redox potentials (E0 O2/HO2 = −0.05
VNHE) and (E0 O2/H2O2 = +0.695 VNHE), compared to the
one-electron reduction of O2 (E0 O2/O2•− = −0.33 VNHE).33 As
it is well-known, H2O2 might act as the ·OH source. For this
reason, its influence in the photooxidation of methane to
methanol for systems with WO3 has also been evaluated in
literature, with more or less contradictory results: improvement
in the CH3OH to CO2 ratio7 or very oxidizing conditions in
which CO2 predominates overtime,8 as in ours. Additionally,
given the electron trapping associated with the O2 presence
(through any of the proposed mechanisms), a better charge
separation can be expected, thus increasing the available holes
for water and/or CH3OH oxidation into other compounds and
eventually CO2.
The obtained results are depicted in Figure 4. There, it is
possible to observe an important increase in the CO2 rate
An additional alternative that has not been mentioned is the
generation of ozone via oxygen photolysis. As known, the VUV
emission of 185 nm of the mercury lamp is able to dissociate O2
into two O3(P) oxygen molecules, which could eventually
generate O3 by combining with other O2 species,35 and finally
could dissociate with the 254 nm radiation of the mercury
lamp, resulting in more O2 and, eventually, H2O2 and ·OH
radicals. However, this mechanism seems lees plausible from
the point of view of our reaction conditions and would not be
able to explain the nonreduction effect of the material in the
presence of O2.
It becomes clear in this sense that the presence of O2
•−
somehow enhances the oxidation process either by O2 or
H2O2 generation, by higher ·OH concentration or by increased
availability of holes in the valence band. So, even when finding
an alternative path in order to avoid the self-reduction of the
material can be a good way for improving the photocalytic
performance, the conditions must be carefully controlled in
order to increase the CH4 oxidation only to a certain extent.
Hence, other approximation can rely on the direct substitution
of the photocatalyst for others with different band structures.
This possibility was also studied, and the results are presented
below.
Figure 4. CH3OH, C2H6 and CO2 rates obtained during the different
tests. Single Bi2WO6 (A), Bi2WO6 and Fe3+ (1 mM) as electron
scavenger (B) and a solution of Fe3+ without photocatalyst (C).
(more than 5-fold), in comparison to the photocatalytic test
without Fe3+. These high rates indicate that very oxidizing
conditions are reached, which could be explained from two
aspects: the improved electron transfer from the conduction
band of Bi2WO6 to Fe3+, thus enhancing the charge separation,
and the higher hydroxyl radicals concentration, able to
overoxidise CH4. Even when methanol rates are also increased,
the presence of Fe3+ greatly affects Bi2WO6 selectivity to
CH3OH (below 10%), despite the significant increase in total
conversion to ∼7%. It is important to point out that the
conversion and selectivity estimation in these two tests with
Fe3+ did not take into account an additional byproduct
observed as a significant chromatographic peak that was not
possible to identify among a wide series of organic compounds.
Further analysis of this would be necessary for proper
calculations. However, the measured rates in Figure 4 are not
affected for this additional species.
3.2.3. Fe3+ Addition. Given the evidence about the influence
of the two concomitant paths in the photocatalytic process and
the thermodynamic drawback associated with the reducing
potential of materials as WO3, different additives have been
evaluated as electron scavengers in the literature. For this
process in particular, methyl viologen dichloride (MV) and Fe
have been reported to serve as electron-transfer agents in
studies involving lanthanum-doped WO3.6 In the second case,
the addition of Fe3+ seems to stabilize the methanol formation8
through the capture of the photogenerated electrons in the
conduction band in a process involving the Fe3+/Fe2+ pair, as
indicated in Scheme 3. As it can be observed, besides the
electron scavenging, additional hydroxyl radicals can be formed
via a Fenton reaction with H2O2.36,37 These radicals may
From another point of view, it is interesting to observe that
the coloration of the Bi2WO6 suspension seemed not to suffer
3017
dx.doi.org/10.1021/cs500821r | ACS Catal. 2014, 4, 3013−3019

ACS Catalysis
Research Article
any change at the end of the test, suggesting that self-reduction
is inhibited because of the Fe3+ presence.
Thus, despite showing a slightly lower methane conversion, the
best CH3OH selectivity is obtained with this material. At the
end of the test, the suspension with BiVO4 suffers a minimal
change in the coloration, compared to the radical variation
when using Bi2WO6. In this sense, the reversible reduction of
the material might be much less pronounced with BiVO4, as
expected from its band structure. This fact may be confirmed
by the differences in the total conversion between 30 and 120
min: a slight increase from 0.8 to 1.0% is found with BiVO4,
whereas the conversion seems to decrease from 1.6 to 1.4%
with Bi2WO6.
Despite these results, the significant differences on the SBET
values among the samples have to be considered. Considerable
differences would be obtained between the Bi2WO6-based
materials and BiVO4, if the product rates were normalized by
surface area (as shown in the Graphical Abstract). This way, it
is interesting to find that BiVO4 is a good alternative as
On the other hand, in the blank test with a Fe3+ solution
under the same reaction conditions revealed that solely Fe3+ in
water (under illumination) is able to oxidize methane to
CH3OH, CO2, and C2H6, in a similar manner to the
photocatalytic tests. In this photo-Fenton process expressed
through eq 14, hydroxyl radicals and Fe2+ are formed,36 and
methane is oxidized through the mechanism previously
proposed.
3.2.4. Other Photocatalysts. Two additional materials were
evaluated. First, a coupled system of Bi2WO6 and TiO2−P25
was prepared as alternative, given the more negative conduction
band of TiO2−P25 in comparison to Bi2WO6, which could be
better for water reduction to H2, leading to an improved charge
separation mechanism. This material was previously reported as
a very active catalyst in a photodegradation process,27 thanks to
the intermediate features between the two single oxides and,
probably, to the enhanced charge transfer from one material to
the other, usually associated with this kind of coupled
systems.38,39 Second, a BiVO4 sample prepared by very similar
conditions to Bi2WO6, like other bismuth-based ternary oxides,
with the advantage of exhibiting a conduction band with more
reducing potential (as shown in Figure 3).
photocatalyst for this selective process despite its low SBET
.
Designing and controlling textural and surface features during
the preparation becomes an important challenge, in order to
obtain materials with enhanced behavior.
In the particular case of methane to methanol photo-
oxidation, these results demonstrate the importance of the two
concomitant paths: oxidation and reduction, through both
charge carriers. For this reason, the selection of a photocatalyst
with the proper band configuration is crucial, unless some
additives (as electron scavengers) are used. Besides this, a
careful design and control of some features as the surface area is
also important. This aspect would imply reaching a certain
commitment degree with the crystalline growth.
CH3OH, CO2, and C2H6 rates and selectivities after 120 min
of illumination are depicted in Figure 5 for the three different
4. CONCLUSIONS
Several bismuth-based materials were synthesized, character-
ized, and evaluated in the selective photooxidation of methane,
with CH3OH, CO2, and C2H6 as the main products. Higher
methane conversion and methanol selectivity were obtained in
the photocatalytic test with Bi2WO6 than through the
photolytic process with solely water. In presence of a
continuous O2 flow, very high CO2 levels were obtained,
because of the possible formation of oxidizing radicals as a
consequence of electronic transfer from the conduction band of
the photocatalyst to adsorbed O2. When Fe3+ was added as
electron scavenger, the total conversion was increased, given an
enhanced charge separation and a major generation of hydroxyl
radicals via Fenton and photo-Fenton mechanisms. However,
despite the increased CH3OH rates, CO2 was the main product
because of the highly oxidizing conditions.
BiVO4 presented the best combination of conversion and
selectivity, even better than those of Bi2WO6, in spite of its
lower surface area. Additionally, an important aspect derived
from the results and related to the two redox processes,
indicated that although Bi2WO6 tends to self-reduce and
deactivate after certain extent due to its band structure, BiVO4,
which has a conduction band with more negative potential and
thus higher probabilities of generating H2 from H2O, is not
reduced during the test.
Figure 5. CH3OH (A), CO2 (B), and C2H6 (C) production rates with
the different materials, at 55 °C.
materials. BW−Ti presented the highest CO2 rates. This
preferential formation of CO2 at expense of the lowest CH3OH
rates is not unexpected, from the point of view of the reported
results for this photocatalyst,26 which exhibits a high oxidizing
power that clearly exceeds those of Bi2WO6. The C2H6 levels
are also higher with this sample, which could be expected if
higher CH3· concentrations are understood to come from more
hydroxyl radicals formed through the process indicated in
Scheme 2. Hence, the total conversion rises to ∼3.5%, but
methanol selectivity decreases below 8%.
Finally, a coupled Bi2WO6/TiO2−P25 system was evaluated,
showing higher photoactivity, at the expense of an important
increase in the CO2 formation.
BiVO4, on the contrary, displays an interesting behavior in
comparison to the other photocatalysts. We observe not only
an increase in the CH3OH rate from ∼15 μmol·h−1g−1 with
Bi2WO6 to 21 μmol·h−1g−1 but also lower CO2 and C2H6 levels.
These materials emerge as interesting photocatalysts for this
reaction. Additional strategies such as increasing the surface
area, enhancing the electron transfer by using other electron
scavengers unable to produce ·OH radicals, or coupling with
3018
dx.doi.org/10.1021/cs500821r | ACS Catal. 2014, 4, 3013−3019

ACS Catalysis
Research Article
(30) Ng, C.; Iwase, A.; Ng, Y. H.; Amal, R. J. Phys. Chem. Lett. 2012,
3, 913−918.
other semiconductors will need to be expplored as alternatives
in future work.
(31) Hill, J. C.; Choi, K. S. J. Mater. Chem. A 2013, 1, 5006−5014.
(32) Sheng, J.; Li, X.; Xu, Y. ACS Catal. 2014, 4, 732−737.
(33) Kim, J.; Lee, C. W.; Choi, W. Environ. Sci. Technol. 2010, 44,
6849−6854.
AUTHOR INFORMATION
Corresponding Authors
*E-mail: smurcia@irec.cat. Tel.: (34) 933 562 615.
*E-mail: tandreu@irec.cat.
■
(34) Saison, T.; Gras, P.; Chemin, N.; Chanea
Brezova, V.; Colbeau-Justin, C.; Jolivet, J. P. J. Phys. Chem. C 2013,
117, 22656−22666.
(35) Zoschke, K.; Bornick, H.; Worch, E. Water Res. 2014, 52, 131−
́
c, C.; Durupthy, O.;
́
Notes
̈
The authors declare no competing financial interest.
145.
(36) Rincon
(37) Mest’an
́
, A. G.; Pulgarín, C. Appl. Catal. B 2006, 63, 222−231.
kova, H.; Mailhot, G.; Jirkovsky, J.; Krysa, J.; Bolte, M.
ACKNOWLEDGMENTS
This work was partially supported by the European Regional
Development Funds (ERDF, FEDER Programa Competitivitat
■
́
́
Appl. Catal. B 2005, 57, 257−265.
(38) Shang, M.; Wang, W.; Zhang, L.; Sun, S.; Wang, L.; Zhou, L. J.
Phys. Chem. C 2009, 113, 14727−14731.
́
de Catalunya 2007-2013), Ministerio de Economia y
(39) Colon
́ ́
, G.; Murcia-Lopez, S.; Hidalgo, M. C.; Navío, J. A. Chem.
Competitividad (CSD2009-00050), and the Framework 7
program under the project CEOPS (FP7-NMP-2012-309984).
Commun. 2010, 46, 4809−4811.
REFERENCES
■
(1) Yuliati, L.; Yoshida, H. Chem. Soc. Rev. 2008, 37, 1592−1602.
(2) Periana, R. A.; Mironov, O.; Taube, D.; Bhalla, G.; Jones, C.
Science 2003, 301, 814−818.
(3) Otsuka, K.; Wang, Y. Appl. Catal. A 2001, 222, 145−161.
(4) Reddy, P. V. L.; Kim, K. H.; Song, H. Renew. Sust. Energy Rev.
2013, 24, 578−585.
(5) Ogura, K.; Kataoka, M. J. Mol. Catal. 1988, 43, 371−379.
(6) Taylor, C. E.; Noceti, R. P. Catal. Today 2003, 84, 9−15.
(7) Taylor, C. E.; Noceti, R. P. Catal. Today 2000, 55, 259−267.
(8) Gondal, M. A.; Hameed, A.; Suwaiyan, A. Appl. Catal. A 2003,
243, 165−174.
(9) Hu, Y.; Nagai, Y.; Rahmawaty, D.; Wei, C.; Anpo, M. Catal. Lett.
2008, 124, 80−84.
(10) Hu, Y.; Anpo, M.; Wei, C. J. Photochem. Photobiol., A 2013, 264,
48−55.
(11) Sastre, F.; Fornes
2011, 133, 17257−17261.
(12) Lop
́
, V.; Corma, A.; García, H. J. Am. Chem. Soc.
́
ez, H. H.; Martínez, A. Catal. Lett. 2002, 83, 37−41.
(13) Gondal, M. A.; Hameed, A.; Yamani, Z. H.; Arfaj, A. Chem. Phys.
Lett. 2004, 392, 372−377.
(14) Yoshida, H.; Hirao, K.; Nishimoto, J.; Shimura, K.; Kato, S.;
Itoh, H.; Hattori, T. J. Phys. Chem. C 2008, 112, 5542−5551.
(15) Nomikos, G. N.; Panagiotopoulou, P.; Kondarides, D. I.;
Verykios, X. E. Appl. Catal. B 2014, 146, 249−257.
(16) Belver, C.; Adan
143, 274−281.
́ ́
, C.; Fernandez-García, M. Catal. Today 2009,
(17) Zhang, L.; Wang, H. L.; Chen, Z. G.; Wong, P. K.; Liu, J. S.
Appl. Catal. B 2011, 106, 1−13.
(18) Shang, M.; Wang, W.; Ren, J.; Sun, S.; Zhang, L. CrystEngComm
2010, 12, 1754−1758.
(19) Xi, G.; Ye, J. Chem. Commun. 2010, 46, 1893−1895.
́ ́
(20) Obregon, S.; Colon, G. J. Mol. Catal. A 2013, 376, 40−47.
(21) Zhang, Y.; Zhang, N.; Tang, Z. R.; Xu, Y. J. Chem. Sci. 2013, 4,
1820−1824.
(22) Zhang, Y.; Xu, Y. J. RSC Adv. 2014, 4, 2904−2910.
́
(23) Murcia-Lopez, S.; Vaiano, V.; Sannino, D.; Hidalgo, M. C.;
Navío, J. A. Res. Chem. Intermed. 2014, DOI: 10.1007/s11164-013-
1523-3.
(24) Cheng, H.; Huang, B.; Liu, Y.; Wang, Z.; Qin, X.; Zhang, X.;
Dai, Y. Chem. Commun. 2012, 48, 9729−9731.
(25) Zhou, Y.; Tian, Z.; Zhao, Z.; Liu, Q.; Kou, J.; Chen, X.; Gao, J.;
Yan, S.; Zou, Z. ACS Appl. Mater. Interfaces 2011, 3, 3594−3601.
(26) Mao, J.; Peng, T.; Zhang, X.; Li, K.; Zan, L. Catal. Commun.
2012, 28, 38−41.
́
(27) Murcia-Lopez, S.; Hidalgo, M. C.; Navío, J. A. Appl. Catal. A
2012, 423−424, 34−41.
(28) Fu, Y.; Sun, X.; Wanga, X. Mater. Chem. Phys. 2011, 131, 325−
330.
(29) Zhang, A.; Zhang, J. Mater. Lett. 2009, 63, 1939−1942.
3019
dx.doi.org/10.1021/cs500821r | ACS Catal. 2014, 4, 3013−3019