Vous êtes sur la page 1sur 10

Molecular Catalysis 436 (2017) 19–28

Contents lists available at ScienceDirect

Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat

Unraveling the role of surface property in the photoreduction


performance of CO2 and H2 O catalyzed by the modified ZnO
Qiangsheng Guo a,c , Qinghong Zhang a,c,∗ , Hongzhi Wang b , Zhifu Liu c , Zhe Zhao c,∗
a
State Key Laboratory for Modification of Chemical Fibers and Polyer Materials, Donghua University, Shanghai 201620, PR China
b
Colleage of Materials and Engineering, Donghua University, Shanghai 201620, PR China
c
Department of Materials Science and Engineering, Shanghai Institute of Technology, Shanghai 201418, PR China

a r t i c l e i n f o a b s t r a c t

Article history: In the tide of investigating catalysis process and developing efficient catalysts, the surface properties of
Received 17 December 2016 catalysts is vitally important, which is lack of due attention in CO2 photoreduction. Herein, a series of Ce
Received in revised form 23 February 2017 modified and La-Ce co-modified ZnO nanorod with varied compositions were synthesized via a simple
Accepted 11 April 2017
impregnation route. The efficient rare-earth modification could obviously decrease the light absorption
Available online 22 April 2017
and increased the specific surface. In situ DRIFTS CO2 -FTIR results indicated that introduction of appro-
priate CeO2 can effectively promote CO2 adsorption and the formation of b-CO3 2− intermediate, which
Keywords:
resulted in the 4.2 times increase of the CH4 yields. While, La and Ce co-modified ZnO catalyst contributed
ZnO
La and Ce co-modified
to the formation of CH3 OH from the mixture of CO2 and H2 O. It was probably due to that electron sepa-
CO2 phototreduction rated rapidly transfered to the surface OH groups, which was directly related to the product of CH3 OH. The
CO2 adsorption combined characterizations and photoactivity studies reported in this paper have provided new insights
Surface sites to the role of surface property (surface sites/groups, reaction intermediates) in CO2 photoreduction on
semiconductor catalyts, which is of significance but rarely reported.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction preventing the photogenerated charge recombination was gener-


ally considered as one of the effective methods for the promoted
Since the first report on photocatalytic reduction of CO2 the photocatalytic efficiency. The manipulation is carried out by
into organic compounds (methanol, methane, formaldehyde, and combining two different catalysts with suitable conduction [12,14],
formic acid) over suspending TiO2 particles by Fujishima et al. self-assembly [15,17], and surface decorates by loading metal oxide
[1] recycling CO2 to fuels using sunlight as the sole energy input and quantum dot [7,8,18–20]. The g-C3 N4 /Bi2 WO6 composites
offered a brand new opportunity for a sustainable energy future. obtained the optimized photocatalytic activity with CO2 produc-
This prospective photochemical process can not only mitigate CO2 tion rate of 5.19 mmol g−1 h−1 under visible light irradiation, which
emissions but also produce energy-bearing compounds, such as was 22 and 6.4 times than that on pure g-C3 N4 and Bi2 WO6 , respec-
CO, methane, and methanol [2–4] which can be subsequently con- tively [14]. Zou and co-worker [17] successfully fabricated novel
verted to liquid transportation fuels. However, CO2 photoreduction hollow spheres consisting of molecular-scale alternating Ti0.91 O2
in engineered systems still suffers from low quantum efficiencies nanosheets and G nanosheets via the layer-by-layer assembly tech-
and yields, deriving from fast electron-hole recombination and nar- nique, delivering 9 times increase of the photocatalytic activity
row light absorption range of semiconductor catalysts. relative to commercial TiO2 . Moreover, a reduced graphene and NiO
In order to enhance the photoactivity of semiconductor cata- co-modified Ta2 O5 photocatalyst (NiOx -Ta2 O5 -rG) prepared by a
lysts, several approaches have been developed, such as synthesis one-step hydrothermal method showed good production selectiv-
of mesoporous catalysts [5–7], loading of metal [8–11], and manip- ity for an aqueous CO2 solution to CH3 OH resulting from graphene’s
ulation of semiconductor heterojunctions [12–16]. Among the excellent electron transfer capacity [19].
various methods, the manipulation of heterojunctions materials to The hybrids of chemical compounds always show several times
the photocatalytic activity compared with the single semiconduc-
tor catalysts, which was simply attributed to the charge effective
separation and the photoexcited electrons fast transfer. How-
∗ Corresponding author.
ever, the changes of catalyst surface structures in the process of
E-mail addresses: zhangqh@dhu.edu.cn (Q. Zhang), zhezhao@sit.edu.cn
(Z. Zhao). hybrids synthesis were neglected, which play a crucial role in

http://dx.doi.org/10.1016/j.mcat.2017.04.014
2468-8231/© 2017 Elsevier B.V. All rights reserved.
20 Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28

determining their photocatalytic activity. The CO2 adsorption and temperature, the white precipitate was collected, washed with dis-
photocatalysis reaction were carried out on the surface. As a con- tilled water and absolute ethanol for several times before dried at
sequence, surface structure is closely linked with the activity sites 80 ◦ C for 12 h.
and CO2 adsorption. If no abundant activity sites and adsorbed The Ce-promoted ZnO with different CeO2 content were typ-
CO2 molecules exist on the surface of semiconductor, it is invalid ically prepared by the impregnation method. The 200 mg of ZnO
that the photoexcited electrons transfer to the semiconductor sur- was mixed in a 30 mL Ce(NO3 )3 solution with different Ce3+ con-
face. Graciani [21] reported that the copper-ceria interface owned centration for 5 h. The slurry was then centrifuged, dried and then
adsorption/reaction sites (truly bifunctional sites) that affords calcined in air gas at 673 K for 4 h. The Ce/Zn molar ratios were
special reaction pathways for the CO2 → CH3 OH conversion and 0.08, 0.10, 0.14 and 0.17, which were measured by ICP-OES. The
showed highly efficient for the synthesis of methanol. Hoffmann Ce modified ZnO samples were denoted as xC/Z (x = 0.08, 0.10, 0.14
[22] found that the coordination of CO2 to surface exchange sites and 0.17). The La and Ce co-modified ZnO catalyst was prepared
on the ternary catalyst led to the formation of surface-bound CO2 using the same procedure method by immersing powdery ZnO into
and related carbonate species, which reduced the energy barrier a mix aqueous solution of constant Ce3+ and La3+ concentration. The
for conduction band electron transfer to CO2 . In the Sastre’s report Ce/Zn molar ratios measured were 0.14 and La/Ce molar ratio 0.09.
[23], the basic solids can increase CO2 conversion up to 2.21% with The La and Ce doubly modified ZnO samples were marked as L-C/Z.
total selectivity toward CH4 from 0.67%, due to their basicity solids
as good CO2 adsorbents. Similarly, a several of papers reported that
the MgO-modified TiO2 catalyst exhibited a higher CH4 selectivity
and yield in CO2 reduction due to the enhanced CO2 adsorption by 2.2. Characterization
MgO [24–26]. Undoubtedly, the effect of surface adsorption proper-
ties on the photocatalytic reduction of CO2 would be an interesting The chemical composition of the sample was determined by
and significant research in the field. The importance of the CO2 inductively coupled plasma-optical emission spectroscopy (ICP-
adsorption on the surface has been evidenced in previous litera- OES; Perkin Elmer, Optima 7000 DV). X-ray diffraction (XRD)
tures [21–26], but few researches discussed the effect of surface patterns were recorded by a PANalytical X’pert PRO diffractome-
structures in molecular level on the CO2 adsorption and reduction ter with a scanning rate of 4◦ min−1 , using Cu K␣ radiation at 40 KV
pathways. Consequently, a large uncertainty in the effect of surface and 40 mA. The average crystalline sizes of catalysts (dCeO2 ) were
structures on CO2 adsorption and intermediate variety and prod- calculated by the Scherrer equation based on the diffraction peak
ucts selectivity (CO, CH4 , CH3 OH and C2 -C3 hydrocarbons) needs to broadening. N2 adsorption-desorption was measured at 77 K using
clarify. an ASAP2020 analyzer after 200 ◦ C for 5 h in vacuum. The specific
ZnO, as a multifunctional nanomaterial because of its unique surface area of samples was determined using Brunauer-Emmett-
physical and chemical properties, has been investigated in pho- Teller (BET) method. The pore size and volume was calculated
tocatalytic applications specially [27,28], such as photocatalytic using the adsorption branch of the isotherm by means of Barrett-
degradation of dye [29–32] and water splitting [33–35]. Conse- Joyner-Halenda (BJH) method. UV–vis diffuse reflectance spectra
quently, ZnO was chosen as the model catalyst in this paper. We were obtained on Shimadzu UV-3600 spectrophotometer. For this,
firstly report the Ce modified ZnO and Ce and La co-modification of sample powders were mixed with BaSO4 spectroscopically, and
ZnO using a simple impregnation method and examine their cat- their absorption spectra were measured with respect to BaSO4 .
alytic properties for CO2 photoreduction. The effect of Ce content The nanostructures samples were observed using a Hitachi S-4800
and La species on the distribution of CH4 and CH3 OH of CO2 pho- field emission-scanning electron microscope (FE-SEM). Transmis-
toreduction is investigated. Meanwhile, we pay attention to the sion electron microscopy (TEM) and high-resolution TEM (HRTEM)
effect of Ce content and La species on adsorption capacity of CO2 images were recorded with a Jeol JEM-2100. X-ray photoelectron
molecular and the change of intermediate variety on the surface spectroscopy (XPS) experiments were carried out with a Thermo
of modified ZnO catalyst. Based on the results of characteriza- Fisher ESCALAB 250Xi system equipped with a hemispherical elec-
tion and reaction performance, the relationships between surface tron energy analyzer. The Al K␣ anode is operated at 15 kV and
properties, intermediate varieties (carbonate, ‘bicarbonate and car- 20 mA and all binding energies were calibrated by using the car-
boxylate) and products distribution were discussed systematically. bonaceous C 1 s line at 284.6 eV as reference. The PL spectra were
Lastly, a credible path way of CO2 → intermediate → CH3 OH and obtained on an Edinburgh Analytical Instrument FLS 920 spec-
CH4 in the photoreduction of CO2 with H2 O was rationally pro- trophotometer with an excitation wavelength of 325 nm. All in situ
posed. DRIFTS FTIR spectra of CO2 adsorption were recorded on Nicolet
6700 spectrometer equipped with a liquid nitrogen cooled MCT
detector, were displayed in absorbance unites, and acquired with
2. Experimental section a resolution of 4 cm−1 , using 32 scans. Prior to the adsorption, the
surface of sample was pretreated with high purity He for 60 min at
2.1. Catalyst preparation 673 K. The sample was cooled to 50 ◦ C, and purity CO2 gas continu-
ously adsorbed for 60 min at 30 ml/min. Subsequently, the physical
Potassium hydroxide, Zinc Nitrate Hexahydrate, Cerium(III) adsorption of CO2 was cleaned by purity He for 10 min at a certain
Nitrate Hexahydrate and Lanthanum(III) Nitrate Hexahydrate were temperature, and the IR spectra of CO2 adsorption with samples
purchased from Tansoole (China). All chemicals were used without were measured and recorded. The formation of possible radicals at
further purification. Water was obtained from a WP-UP-1810 water initial stages (∼3 min) was monitored by detecting its adduct with
purification system (Water Purifier, China). 5,5-dimethyl-1-pyrroline N-oxide (DMPO) via an electron param-
ZnO were prepared by hydrothermal method. In a typical pro- agnetic resonance (EPR) spectrometer (Bruker E500 spectrometer).
cedure, 1.25 mmol Zn(NO3 )2 ·6H2 O was dissolved in 25 ml water Before irradiation, CO2 was purged for 10 min into the reaction
under mild magnetic stirring at room temperature. Then a second slurry containing 50 mM DMPO and 50 mg of catalysts (0.14C/Z). No
solution (25 mL) containing KOH (1.0 g) was added dropwise into signals were observed for DMPO only. The instrument conditions
the above solution with vigorous stirring. At last, the resulting sus- used during the EPR measurement were as follow: modulation
pension was transfer into 100 mL Teflon autoclave. The autoclave amplitude, 2.00 G; microwave power = 10.03 mW; and light source,
was heated in an oven at 150 ◦ C for 20 h. After being cooled to room 200 W xenon.
Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28 21

Fig. 1. Schematic of experiment setup for photo catalytic CO2 reaction with H2 O to hydrocarbons.

2.3. Photocatalytic activity All the Ce modified ZnO samples exhibit identical hexagonal ZnO
crystal phase and the diffraction peaks are neither broaden nor nar-
As shown in Fig. 1, the photochemical experiments of CO2 pho- row, indicating that the addition of CeO2 does not have influence
toredution were performed in a self-designed quartz photoreactor on the crystalline structure and size of ZnO. No diffraction peaks of
(volume 120 ml). The light source was 450 W Xe lamp (Beijing CeO2 are detected for 0.08C/Z, probably due to that CeO2 species
Aulight Co., Ltd.) at UV–vis (␭ = 320–780 nm). The light intensities are highly dispersed on the surface of ZnO or the small amount of
measured by an optical power meter (Beijing Aulight Co., Ltd., CEL- CeO2 is beyond the detection limit of XRD. With the increase of
NP2000-10) were 580 mW cm−2 in the ranges of 320–780, and the Ce content, a new peak corresponding to the (111) plane of crys-
fraction of the UV light in the spectrum was approximately 10%. For talline CeO2 is observed at 28.6◦ and becomes stronger. Besides,
each test, 50 mg sample were dispersed on the bottom of a quartz the ceria modified ZnO were examined using Raman spectroscopy.
glass reactor. The reaction setup was vacuum-treated several times. The results are shown in Fig. S1. The spectra of Ce modified ZnO
Following that, 99.99% CO2 continuously passed through a water samples show a new Raman peak at 461 cm−1 , which attribute
bubbler to bring a mixed gas of CO2 and H2 O into the photoreactor to an F2g vibration peak reported in the literature for pure Ce02
at a flow rate of 30 mL/min. After 60 min, the inlet and outlet valves [35].The result of XRD and Raman indicates that the CeO2 nanopar-
of the reactor were closed. The temperature of the reactor was kept ticles were successful loaded in the surface of ZnO. Fig. 3 presents
at 473 K. The photocatalytic reaction was typically performed for the XRD patterns of La and Ce co-modified ZnO samples, and no
6 h. The amounts of formed CH3 OH and CH4 were analyzed by gas changes of the hexagonal ZnO peak are found, which implies that
chromatography (GC) quipped with flame ionization detector (FID). the crystal structures of the obtained ZnO are not affected by La
species. Moreover, the peak position of CeO2 for L-C/Z slightly shifts
3. Results to lower angle, which indicates an increase in the cell parameter
(the detail changes of CeO2 diffraction peak are described in Fig.
3.1. Characterization S2). This increase can be interpreted as the lattice substitution of
Ce4+ (0.97 Å) by La3+ (1.10 Å) accompanied with the formation of
The XRD patterns of the synthesized Ce modified ZnO, as well as ceria lanthanum solid solutions [36,37]. As shown in Table 1, the
those of pure ZnO and CeO2 , are shown in Fig. 2. It can be seen that average CeO2 crystallite size of the 0.14C/Z calculated by Scherrer
the peaks at 31.7, 34.4, 36.2, 56.5, 62.8 and 67.9◦ are ascribed to the equation is 4.8 nm, while Ce and La co-modified ZnO samples have
(100), (002), (101), (110), (103) and (112) planes of hexagonal ZnO larger crystallite sizes of CeO2 (9.4 nm) [37]. In brief, the incorpo-
(JCPDS file 89-1397) and the peaks appeared at 28.6, 33.1, 47.6, 56.4, ration of La3+ cations into the Ce4+ cations to forms a La-CeO2 solid
695 and 76.8◦ are assigned to the (111), (200), (220), (311), (400) solution and enhances crystal growth at the same time.
and (331) planes of cubic CeO2 (JCPDS file 34-0394), respectively. The nitrogen adsorption-desorption isotherms of three typical
samples (ZnO, 0.14C/Z and L-C/Z) at 77 K were shown in Fig. 4. The
0.14C/Z and L-C/Z samples presents the typical IV isotherms with
ZnO
a type H3 hysteresis loops (P/P0 > 0.7), indicating the presence of a
CeO2 porous structure. It is ascribed to the stacking of rare earth oxides
nanoparticles on the surface of pure ZnO. The surface area and
0.17C/Z
pore volume of all samples are summarized in Table 1. The sur-
face area and pore volume of ZnO are as low as 3.52 m2 g−1 and
Intensity (a.u.)

0.14C/Z 0.012 cm3 g−1 , respectively. It can be seen that the surface area and
pore volume gradually increase with the increasing CeO2 content
from 0.08 to 0.14. In addition, the surface area and pore volume
0.10C/Z
of L-C/Z markedly increase compared with 0.14C/Z, which are 3.7
and 4.4 times as high as those of pure ZnO, respectively. (The detail
0.08C/Z results are summarized in Table S1).
Fig. 5 shows the SEM images of ZnO, 0.14C/Z and L-C/Z. As shown
in Fig. 5a, the smooth rod-like products being ∼4 ␮m in length and
ZnO
CeO2 ∼1 ␮m in width are observed in the as-papered ZnO. With regard
to 0.14C/Z (Fig. 5b), it is noted that large CeO2 clusters are formed
20 30 40 50 60 70 80 on the surface of ZnO. Fig. 5c shows a significantly rough surface
2 Theta (degree)
with a large number of particles accumulating on the surface of
Fig. 2. The XRD patterns for ZnO, CeO2 , 0.08C/Z, 0.10C/Z, 0.14C/Z and 0.17C/Z.
ZnO, which explains the increased surface and volume showed in
22 Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28

ZnO
CeO2
Intensity (a.u.)

L-C/Z
L-C/Z

0.14C/Z

0.14C/Z

20 30 40 50 60 70 80 27 28 29 30 31
2 Theta (degree) 2 Theta (degree)

Fig. 3. The XRD patterns for 0.14C/Z and L-C/Z.

ticles surrounding 0.14C/Z are loosely floccule agglomerates, while


40
ZnO the nanoparticles on the surface of L-C/Z clearly display small grains
35 0.14C/Z and disperse homogeneously. The difference in surface morphology
L-C/Z is likely due to the increased nanocrystal after La species introduc-
Volume of adsorption (cm g )
3 -1

30
tion. The HRTEM images of CeO2 nanocrystal on the surface of ZnO
25 are shown in Fig. 6C and F. Although of the loosely floccule state
demonstrated in Fig. 6B, an obvious lattice fringe of nanocrystal
20 can be found in the HRTEM image of 0.14C/Z, which confirms that
the CeO2 nanoparticles are cubic (222) phase with the lattice spac-
15
ing of 0.156 nm. The mean size of CeO2 particles evaluated is about
10 4 nm for 0.14C/Z. This observation agrees with XRD data of 0.14C/Z.
The HRTEM measurements of L-C/Z demonstrate clearly that the
5 mean size of nanoparticles increases obviously to about 9 nm from
0 4 nm of 0.14C/Z. The change in particles size is due to the load-
ing of La as co-catalyst, which can promote the crystallization of
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 CeO2 nanoparticles, being consistent with the variety of calculated
P/P0
crystal size in Table 1. Another very important finding in HRTEM
Fig. 4. N2 adsorption-desorption isotherms of ZnO, 0.14C/Z and L-C/Z. figure is the slightly increased lattice spacing of L-C/Z compared to
0.14C/Z. For 0.14C/Z, the lattice spacing of (222) phase is 0.156 nm
and then increase slightly to 0.157 nm after addition of La. Hence,
Table 1. The transformation of morphology observed from the SEM
La as co-catalyst plays a significant role in the process of crystal for-
images agrees with the results of XRD.
mation, which conforms to the shift of (111) plane to lower angle
The particle size, morphology, and crystal structure of 0.14C/Z
observed in XRD patterns.
and L-C/Z are further evaluated by TEM and HRTEM, as shown
The UV–vis diffuses reflectance spectroscopy (DRS) curves of
in Fig. 6. The 0.14C/Z and L-C/Z exhibit relatively uniform rod-
pure ZnO, CeO2 , rare earth modified ZnO are performed in Fig. 7.
liked morphology. Moreover, rod-liked ZnO fringe is fuzzy, which
The ZnO displays a strong peak of light absorption (␭ < 400 nm), and
demonstrates the overlapping CeO2 nanoparticles located at the
CeO2 weak absorbs light with wavelength shorter than 600 nm. The
edge of ZnO. After magnification (Fig. 6B and E), the CeO2 nanopar-

Table 1
Textural properties, CeO2 crystal size and the energy of band gap (Eg ) of the catalysts.

Samples Surface area (m2 g−1 ) Pore volume (cm3 g−1 ) Crystal size (nm)a d/DXRD (nm) Band gap energy (ev)b

ZnO 3.52 0.012 ∼ ∼ 3.21


CeO2 63.0 0.165 29 ∼ 2.53
0.08C/Z 5.92 0.017 ∼ ∼ 3.18
0.10C/Z 8.93 0.026 4.1 ∼ 3.17
0.14C/Z 10.8 0.040 4.8 0.312 3.15
0.17C/Z 8.4 0.031 5.5 ∼ 3.13
L-C/Z 13.14 0.053 9.4 0.314 3.16
a
Crystal size and interplanar spacing of CeO2 are calculated by the (111) peak of XRD.
b
Bang gap energy calculated by using (h␭)n = k(h␭-Eg), n = 2.
Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28 23

Fig. 5. SEM images of the (a) ZnO, (b) 0.140C/Z and (c) L-C/Z.

Fig. 6. TEM micrographs of (A, B) 0.14C/Z and (D, E) L-C/Z samples.HR-TEM micrographs of (C) 0.14C/Z and (F) L-C/Z samples.

Ce modified ZnO show spectral response in a broad visible region deep level emission in the visible range is related to the defect
(␭ = 400–600 nm) owing to the photosensitizing effect of CeO2 and states, such as oxygen vacancies and zinc interstitials, which would
exhibit a distinctly decreased intensity of absorption in untraviolet- act as luminescence centers [44,45]. From the PL results, it can be
light regions than that of pure ZnO. The band energy gap of above observed that the xC/Z samples exhibit stronger peak intensity of
samples could be calculated by using (h␭)2 = k(h␭–Eg ). Plotting PL spectra compared with pure ZnO, which suggests the presence of
(h␭)2 versus h␭ based on the spectral response give the extrapo- higher singly ionized oxygen vacancies and zinc interstitials in xC/Z
lated intercept corresponding to the Eg value. As shown in Table 1, [46]. The results indicate the Ce addition consequently affect the
the optical band gap energies of the 0.08C/Z, 0.10C/Z, 0.14C/Z and defect states [47,48]. Moreover, the red-shift of emission peak from
0.17C/Z are slightly lower than that of pure ZnO (3.21 eV), which 560 nm to 595 nm was observed for all Ce modified ZnO, which may
indicates that the enhanced ability to absorb visible-light of Ce be due to the smaller conduction band offset between CeO2 and
modified ZnO samples endow the composites to be a promising ZnO. The phenomenon is similar to the result researched by Sunita
photocatalyst for solar-driven application. On the other hand, the [49]. As shown in Fig. S2, the emission peak intensity of La and Ce
light absorption of La and Ce co-modified ZnO decrease, indicating co-modified ZnO slightly decreases compared to Ce modified ZnO,
that La as co-catalyst slightly weakened the light absorption of the which can be attributed to the fact that La can act as an “electron
entire UV–vis range. reservoir” to trap the photogenerated electrons from semiconduc-
Fig. 8 depicts the photoluminescence (PL) spectra of pure ZnO tors and retard the recombination of electron-hole pairs. Thus,
and modified ZnO catalysts in the wavelength range of 400–700 nm. it may contribute to the photoactivity enhancement of the L-C/Z
As displayed in Fig. 8, all samples show a broad peak of PL inten- [50–52].
sity emission at about 500–650 nm, which is commonly referred Fig. 9 shows in situ DRIFT spectra obtained after CO2 adsorption
to a deep level of trap-state emission [43]. It is known that the on rare earth modified catalysts at 325 K. A number of IR bands
24 Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28

bidentate-carbonate(m-CO3 2− and b-CO3 2− ). According to litera-


ZnO tures [53–56], a shoulder peak at 1690 cm−1 is ambiguous, due
0.14C/Z
to the difficulty of distinguishing the contribution of bicarbon-

Absorbance(a.u.)
ate or carboxylates. The bands located at 1608 and 1305 cm−1
L-C/Z
are attributed to the symmetrical stretching vibration and asym-
Absorbance(a.u.)

metric stretching vibration of surface bidentate carbonate species


0.08C/Z (b-CO3 2− ), which is produced by CO2 coordination with an unsatu-
0.10C/Z rated O2− . Meanwhile, the band located at 1415 cm−1 is indicative
of CO2 adsorbed as monodentate species. The 1495 cm−1 band is
200 300 400
Wavelenth(nm)
500 600 700 800
assigned to the bicarbonate species (HCO3 − ), which is produced by
0.17C/Z the reaction of CO2 with surface OH groups [57,58]. Furthermore,
the band at 1250 cm−1 is ascribed to carboxylate species (HCOO− ).
0.14C/Z As showed in Fig. 9A, it can be observed distinctly that the feature
CeO2
of b-CO3 2− species at 1608 cm−1 is a shoulder peak and the peak
200 300 400 500 600 700 intensity is lower than that of HCO3 − species for pure ZnO. How-
Wavelength(nm) ever, for 0.14C/Z, the intensity of HCO3 − much higher than that of
b-CO3 2− , which is difficult to be observed. After rigorous calcula-
Fig. 7. UV–vis diffuse reflectance spectra of the as-prepared pure ZnO, xC/Z and tion, the peak intensity of b-CO3 2− increases distinctly, while the
L-C/Z.
growth rate of HCO3− is slowly, and the ratio of b-CO3 2− /HCO3 −
reached a plateau for after the content of Ce exceed 0.14. This
result demonstrated that the surface of pure ZnO was primarily
(a) ZnO (d) dominated by b-CO3 2− accompanied with minor HCO3 − , while the
(c)
(b) 0.08C/Z surface of Ce modified ZnO was covered with more HCO3− and less
Realative intensity (a.u.)

(c) 0.14C/Z (e) CO3 2− . As shown in Fig. 9B, the similar spectra with various types
of carbonate-like species are obtained on the L-C/Z samples, which
(d) 0.17C/Z (b) suggested that b-CO3 2− , HCO3 − , and HCOO− are the major interme-
(e) L-C/Z
diates on the surface of La-Ce co-modified ZnO sample. La and Ce
co-modified led to enhanced peak intensity of b-CO3 2− and HCO3 − ,
indicating appropriate La species are beneficial for the formation of
b-CO3 2− and HCO3 − intermediate. (Fig. S4 shows the IR spectra of
the CO2 molecules adsorbed on La and Ce co-modified ZnO).
(a)

3.2. Photoreduction of CO2 with H2 O

400 450 500 550 600 650 700 To demonstrate the photocatalytic performance of as-obtained
Wavelength(nm) ZnO and rare earth modified ZnO samples, their photoreduction
experiments of CO2 with H2 O under simulated solar irradiation
Fig. 8. Room temperature photoluminescence spectra of ZnO, Ce modifeidZ and
L-C/Z.
were carried out, with the results as shown in Fig. 10. From Fig. 10a,
it can be seen that the yield of CH4 continuously increases with
the irradiation time for ZnO and Ce modified ZnO. Compared with
in the range of 1800–1000 cm−1 are observed, and these bands ZnO, CeO2 prepared by hydrothermal method owned a remarkably
are mainly due to various types of carbonate-like species includ- higher CH4 yield in 6 h. Thus, the deposition of CeO2 nanoparticles
ing carboxylate species (HCOO− ), bicarbonate (HCO3 − ), mono- and results in a quantitatively improved CH4 formation and the CH4
b-carbonate(CO3 )
2-

(A)
m-carbonate(CO3 )

(B)
2-
carboxylate(HCO2 )

bicarbonate(HCO3 )
-

b-carbonate(CO3 )
2-

carboxylate( HCO2 )
-
Absorbance(a.u.)

Absorbance(a.u.)

L-C/Z
(d)
(c)
(b) 0.14C/Z
(a)

2000 1800 1600 1400 1200 2000 1800 1600 1400 1200
-1 -1
Wavenumbers(cm ) Wavenumbers(cm )

Fig. 9. In situ DRIFTS spectra of the CO2 molecules adsorbed on (a)ZnO, (b) 0.08C/Z, (c) 0.14C/Z, (d) 0.17C/Z and co-modified ZnO at 325 K.
Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28 25

30 45
ZnO (a) 40
ZnO
(b)
25 0.08C/Z 0.08C/Z
0.10C/Z 35 0.10C/Z
0.14C/Z 0.14C/Z

Yield of CH3OH(umolg )
-1
Yield of CH4 (umolg )

20 30 0.17C/Z
-1

0.17C/Z
CeO2 CeO2
25
15 L-C/Z L-C/Z
20

10 15

10
5
5
0
0

0 1 2 3 4 5 6 0 1 2 3 4 5 6
ReactionTime (h) ReactionTime (h)

Fig. 10. Yields of CH4 (a) and CH3 OH (b) as functions of irradiation time over ZnO, CeO2 , C/Z and L-C/Z.

yields rising continuously with the increasing of CeO2 content. The after Ce introduction, the ratio value of CH4 to CH3 OH is obvi-
0.14C/Z sample displays the best photoactivity (25.1 ␮mol g−1 ) for ously continuous rising with the increase of Ce. Despite of the
CO2 and water to CH4 , which is about 4.2 times higher than that of lower photoactivity, the 0.17C/Z catalyst exhibits the highest ration
ZnO. From Fig. 10b, it is a remarkable fact that, the variation ten- value (2.08) of CH4 to CH3 OH among the as-prepared samples. The
dency of CH3 OH yield is strikingly different with that of CH4 for 0.14C/Z exhibits the maximal CH4 and CH3 OH yield, and the ratio
pure CeO2 and Ce modified ZnO. The CH3 OH yield of CeO2 has a of CH4 to CH3 OH remains relatively high level (1.87). Therefore,
maximum value at about 1.2 h but subsequently decreases along the appropriate CeO2 addition greatly improves the CH4 selectiv-
with irradiation time. Therefore, a certain amount of CeO2 effec- ity, slightly increases CH3 OH selectivity, and markedly enhancing
tively reduced the CH3 OH yield. The CH3 OH yield of 0.08C/Z and CH4 yield finally. It is concluded that CeO2 species contribute to
0.17C/Z is 3.7 and 3.4 ␮mol g−1 , which is 0.28 and 0.26 times than selective formation of CH4 from the mixture of the CO2 and H2 O,
that of pure ZnO, respectively. Only 0.14C/Z catalyst shows com- which was verified by the fact that the CH3 OH disappears after
parative CH3 OH yield for CO2 and water to CH3 OH. In addition, 6 h irradiation time (showed in Fig. 10b). In the report of Wang
Fig. 9a also clearly shows the yield of CH4 from CO2 reduction as [6], ordered mesoporous CeO2 -TiO2 composite exhibited excellent
a function of irradiation time for the L-C/Z catalysts under visible photocatalytic activity in the reduction of CO2 with H2 O and the
light. The CH4 yield of L-C/Z was 26.8 ␮mol g−1 , which is slightly CH4 yield increased rate exceed that of CO, which was similar to our
higher than the 0.14C/Z (24.8 ␮mol g−1 ). When it comes to the vari- experimental phenomena. Besides, the CH3 OH yield of co-modified
ation of CH3 OH yield, the L-C/Z shows observably higher yield of L-C/Z rise to 40.7 ␮mol g−1 from 13.4 ␮mol g−1 (0.14C/Z), while the
CH3 OH (40.7 ␮mol g−1 CH3 OH yield), which was 3.1 times than that CH4 yield increase to 26.8 ␮mol g−1 from 24.8 ␮molg−1 (0.14C/Z).
of 0.14C/Z sample. (Blank tests are showed in Fig. S5 and the effect Obviously, CH3 OH formation is more favorable than CH4 after La
of La content on the react activity is summarized in Fig. S6). introduction. In conclusion, La species contribute to the selective
In order to clearly present the influence of rare earth modi- formation of CH3 OH from the mixture of the CO2 and H2 O, while Ce
fication to the efficiency of CO2 photoreduction, the cumulative species effectively promoted selective formation of CH4 . It indicates
products yields of ZnO, CeO2 and modified ZnO were compared that the La and Ce species play different roles on the photocatalysis
in Fig. 11. For ZnO, the yields of CH4 and CH3 OH are 5.7 and of CO2 and H2 O to hydrocarbons. (The changes of products yields
12.8 ␮mol g−1 , and the ratio of CH4 to CH3 OH is 0.44. Notably, for 0.14C/Z and yL-C/Z are described in Fig. S7).

4. Discussion
CH4
40 2.0 According to the published literature, CO2 photochemical reduc-
CH3OH tion follows different reaction pathways to yield the end different
35
products through different intermediates [1,59,60]. Various surface
Ratio of CH4 and CH3OH

30
1.5 reactions and the processes of multiple electron/proton transfers
Yield (umolg )

associated with a variety of final products will take place, which


-1

25 are listed below [60,61].


1.0
20 H2 O + 2 h+ → 1/2O2 + 2H+ ; (1)
15
0.5
2H+ + 2e− → H2 ; (2)

CO2 + 2H+ + 2e− → CO + H2 O;


undetect

10
(3)
5 + −
0.0 CO2 + 2H + 2e → HCO2 H; (4)
0
ZnO 0.08C/Z 0.10C/Z 0.14C/Z 0.17C/Z L-C/Z CeO2 CO2 + 4H+ + 4e− → HCHO + H2 O; (5)
+ −
Fig. 11. The cumulative CH4 and CH3 OH yield and their ratio value for ZnO and Ce CO2 + 6H + 6e → CH3 OH + H2 O; (6)
modified ZnO after visible light irradiation of 6 h. Error bars were calculated from
three independent measurements. CO2 + 8H+ + 8e− → CH4 + 2H2 O; (7)
26 Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28

u''' v'''
u v
u'' u'
v'' v'

0.14C/Z

Intensity (a.u.)
17.57%

0.10C/Z
15.66%
Fig. 12. Proposed reaction pathways for CO2 photoreduction to hydrocarbons.
0.08C/Z
11.79%

920 910 900 890 880 870


The gas products may include CO, CH4 , and H2 (via water split- Binding Energy (ev)
ting), and the liquid products may have HCOOH, HCHO, and CH3 OH.
The above photo catalytic reactions depend on three factors: (1) Fig. 13. The Ce 3d XPS spectra of the 0.08C/Z, 0.10C/Z and 0.14C/Z.
light absorption of photocatalysts, (2) transfer of charge carriers,
and (3) use of the charge carriers in the processes (showed in ratios continuously increase with increasing Ce amount, indicat-
Fig. 12). In our experiment, the adsorption of light is decreased ing the progressive enhancement of the surface −Ce3+ · · ·O species.
(Fig. 7) and the PL intensity (Fig. 8) is increased after Ce modified, Besides, the O 1 s spectra of Ce co-modified catalysts are presented
which indicates the Ce modified catalyst has decreased capacity of in Fig. 14. The peak with low binding energy is attributed to the
light adsorption and shorter lifetime of charger carriers. Changes lattice oxygen ions of the crystalline network [41]. Two peaks at
of two general aspects are unfavorable for photocatalytic conver- higher binding energy correspond to oxygen species and hydroxyl
sion of CO2 to hydrocarbons, which might lead to the deceased groups located on the surface, respectively. It can be seen that oxy-
activity of 0.08C/Z and 0.10C/Z compared to ZnO (Fig. 11). But the gen species on the surface increase with the increase of Ce, which
performance of 0.14C/Z was obviously increased compared with further indicates the amount of surface the −Ce3+ · · ·O species grad-
ZnO. Considering the results of in situ CO2 -FTIR, it is found that the ually increase. The results of Ce and O XPS forcefully evidence the
capacity of CO2 adsorption is enhanced notably. We conclude that forementioned inference. Similarly, Liu [25] found that the pho-
the photocatalytic reaction rate for CO2 reduction on the pure ZnO toactivity of defective TiO2 was remarkably higher than that of
catalysts is slow due to the limited CO2 adsorption and the forma- defect-free TiO2 . This enhancement was primarily attributed to the
tion of intermediates on the catalyst surface. In other words, if there creation of oxygen vacancies and Ti3+ on the surface, which could
is no enough CO2 adsorption and formation of intermediates on the promote CO2 activation by generating CO2 -intermediates. Besides,
surface of catalysts, the following process of C O bond breaking Park [64] confirmed Ceria nanoparticles in contact with titania have
and C H bond formation is impossible to complete. It is reasonable Ce3+ and Ce4+ states of almost equal stability. In virtue of the previ-
that the CO2 adsorption and formation of intermediates on surface ous reports, it is that infferred that defect emerged on the interface
of Ce modified ZnO is rate-determining step in the process of CO2 of the Ce modified ZnO mixed-metal oxide promotes intermedi-
to hydrocarbons. Moreover, it is worthy serious consideration that ates formation and notably increases the CH4 yields. Scheme 1
the CH4 yield of Ce0.5/Z-r catalyst is 4.2 times higher than ZnO, and schematically shows that the formation and reaction procedure of
CH3 OH yield slightly increases. It indicates that appropriate CeO2 CO2 adsorption intermediate over (Ce, La)/ZnO materials.
is helpful to the photocatalytic conversion of CO2 to CH4 . As shown In addition, in situ DRIFTS results in Fig. 9(b) demonstrate the
in Fig. 8, it is found that the b-CO3 2− intensity of 0.14C/Z is much more favorable formation of different intermediates (include b-
higher than that of pure ZnO, and the peak intensity of HCO3 − has CO3 2− , HCO3 − , and HCOOH) after La and Ce co-modified. As shown
little change. The results of in situ CO2 -FTIR are consistent with the in Fig. 11, interestingly, the methanol yield of L-C/Z is 40.7 ␮mol g−1
ratio of CH4 and CH3 OH for photoreduciton of CO2 and H2 O. Based after 6 h illumination time, which is 3.1 times than that of 0.14C/Z,
on these phenomena observed, we reasonably infer that b-CO3 2−
is a possible intermediate for the production of CH4 , while HCO3 −
is possibly related to the formation of CH3 OH. According to the (2)
(3)
reports in the literature [53,62,63], the b-CO3 2− groups is formed
by the interaction of CO2 with an acid metal ion and its neighboring 43.58%
46.52% (1)
basic oxygen (Me4+ -O2− ) centers or vacancy oxygen on the surface 9.90%
Intensity (a.u.)

of semiconducts, and the HCO3 − species is produced by the reac- 0.14C/Z


tion of CO2 with surface OH groups. Consequently, we propose that
metal ion and its neighboring oxygen or vacancy oxygen are the 45.24%
33.03% 21.72%
catalysis active sites of CH4 formation, and surface OH groups are
closely linked with CH3 OH synthesis from CO2 photoreduction. 0.08C/Z
In order to verify above-mentioned conjecture, the surface
structures of Ce modified ZnO were character by the X-ray pho- 38.43% 41.32%
toelectron spectroscopy. Fig. 13 shows the XPS spectra of Ce 3d for 20.25%
Ce modified ZnO. The labels v and u indicate the spin-orbit cou-
pling 3d5/2 and 3d3/2 , respectively. XPS peaks denoted as u, u , u ZnO
and v, v , v are attributed to Ce4+ while u’ and v’ are assigned to
537 536 535 534 533 532 531 530 529 528 527 526 525
Ce3+ , following the convention established by Burroughs et al. [38]. Binding Energy (ev)
Based on the deconvoluted peak area, the Ce3+ /Ce4+ ratios are listed
in Fig. 13. It can be noticed that the intensity ratio of Ce3+ /Ce4+ Fig. 14. The O 1 s XPS spectra of the ZnO, 0.08C/Z and 0.14C/Z.
Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28 27

Scheme 1. The surface reaction mechanism of CO2 photoreduction to CH4 and CH3 OH.

while the CH4 yield increases slightly. But the intermediates HCO3 − C H bond formation (as showed in Scheme 1), respectively. (2)
feature of in situ DRIFTS CO2 -FTIR has not great increase. In order The Me4+ -O2− centers or oxygen vacancies on the surface of semi-
to rationally understand the experiment result, the transfer rate of conducts demonstrate faster react rate with separated electrons
charge carrier influence on the surface reaction mechanism of CO2 than surface −OH groups. The excess separated electrons will react
photoreduction is scientifically evaluated. The photoluminescence with the surface OH groups due to longer life time of electrons. On
spectra presented in Fig. 8 shows that the peaks at 595 nm decrease the basis of above, the performance result of L-C/Z catalysts were
in intensity. These decreases in intensity imply that La species reasonably and scientifically explicated. For the Ce modified ZnO
retard the recombination of electron-hole pairs and improve the samples, the process (1) presented in Scheme 1 is the major reac-
transfer of charge carriers to the surface sites. Combine to the fact tion path. Despite low transfer capacity of induced electrons, the
of remarkable increased CH3 OH yield; we can infer that the induced CO2 adsorption was rate-determining process because the surface
electron after La introduction transfers to the activity sites of CO2 Ce4+ -O2− centers or oxygen vacancies of Ce modified ZnO catalysts
convert into CH3 OH, and enhanced formation of CH3 OH. Despite rapidly react with a small amount of separated electrons. At last, the
enhanced capacity of electron transfer, the CH4 yield of L-C/Z had CH4 yield of 0.14C/Z is 4.2 times higher than ZnO after appropriate
not significant increase. It is attributed to the electron separated Ce modified due to the observably increased b-CO3 2− intermedi-
after La and Ce co-modified is reacted with surface OH groups rather ates, which are possible intermediates for the production of CH4 .
than Me4+ -O2− centers or oxygen vacancies. In consideration of the When comes to La and Ce co-modified ZnO catalysts, electron sep-
increased photoluminescence (PL) spectra for Ce modified ZnO, we arated after La co-modified was reacted with surface OH groups for
conclude that the Me4+ -O2− centers or oxygen vacancies have pri- the longer lifetime of induce electrons, and the amount of HCO3 −
ority to react with induced electrons when semiconducts own the intermediates (directly related to the product of CH3 OH) increases.
weak ability of electron-hole separation. That is to say, the reaction Both of two factors lead to the increased methanol yield of L-C/Z,
rate of Me4+ -O2− centers or oxygen vacancies with induced elec- which is 3.1 times than that of 0.14C/Z sample after 6 h illumination
trons is much faster than surface OH groups, and the great amount time. Namely, the process (2) begins to dominate in the process of
of electrons react with rapidly with Me4+ -O2− centers or vacancy CO2 to hydrocarbons. The yield of CH4 slight increased because no
oxygen in spite of low ability of electron separation (proved by the more electrons react with the increased b-CO3 2− intermediates.
increased photoluminescence spectra for Ce modified ZnO).
It is demonstrated above that active species (b-CO3 2− , HCO3−
intermediates) play important roles for the mechanism. ESR spin- 5. Conclusions
trapping technique was conducted to confirm the probability of
formation OH, CO2 , H and CH3 radicals of them or their deriva- In summary, we have demonstrated that Ce modified and La-Ce
tive radicals. The EPR characterizations probed using DMPO were co-modified ZnO composites with rod structure and varied compo-
showed in Fig. S8. Clear signals of DMPO-CH3 and DMPO-OH sitions, which were synthesized via a simple impregnation route.
adducts were observed, and DMPO-H or DMPO-CO2 adducts were The photocatalytic performance of as obtained samples indicates
not found. The EPR signals were similar to reported studies [22,65]. that introduction of CeO2 species to ZnO can significantly promote
The absence of DMPO-H and DMPO-CO2− adducts is attribute to the the selectivity and yield of CH4 , while the La and Ce co-modified
fast rate of hydrocarbon formation on the surface of 0.14C/Z at room ZnO effectively increases the catalytic process of CO2 to CH3 OH.
temperature [65]. The presence of OH and CH3 radicals strongly Combined the characterization results, the CO2 adsorption over Ce
supported the surface reaction mechanism of CO2 photoreduction modified ZnO catalysts is the rate-determining process, and Ce can
showed in Scheme 1. greatly improve the capacity of CO2 adsorption and b-CO3 2− inter-
Based on the above discussion, it is rationally inferred that (1) mediates formation on the Ce4+ -O2− centers or oxygen vacancies
Me4+ -O2− centers or oxygen vacancies are the active sites of CH4 which are the activity sites of CH4 formation. Another important
formation, and surface OH groups are closely linked with CH3 OH finding is that La as co-catalyst helps with the separation of pho-
synthesis from CO2 photoreduction. CO2 molecular react with two togenerated electron-hole pairs. The electrons own long lifetime
active sites, forming b-CO3 2− , HCO3 − intermediates and produce and can react with surface OH groups to forme HCO3 − species,
to CH4 and CH3 OH after the process of C O bond breaking and which may be a more active intermediate for CH3 OH production.
This study provides a new information to correlate the surface
28 Q. Guo et al. / Molecular Catalysis 436 (2017) 19–28

chemistry (surface sites/groups, reaction intermediates) with CO2 [25] L.J. Liu, C.Y. Zhao, D. Pitts, H.L. Zhao, Y. Li, Catal. Sci. Technol. 4 (2014)
photoreduction activity. It is practical and helpful to enhance the 1539–1546.
[26] L.J. Liu, C.Y. Zhao, H.L. Zhao, D. Pitts, Y. Li, Chem. Commun. 49 (2013)
transformation efficiency of CO2 phororeduction and selectively 3664–3666.
synthesize deficient industrial chemicals. [27] C.G. Silvaa, M.J. Sampaio, S.A.C. Carabineiro, J.W.L. Oliveira, D.L. Baptista, R.
Bacsa, B.F. Machado, P. Serp, J.L. Figueiredo, A.M.T. Silvaa, J.L. Faria, J. Catal.
316 (2014) 182–190.
Acknowledgements [28] M.J. Sampaio, J.W.L. Oliveira, C.I.L. Sombrio, D.L. Baptista, S.R. Teixeira, S.A.C.
Carabineiro, C.G. Silva, J.L. Faria, Appl. Catal. A 518 (2016) 198–205.
This research is supported by the Youth Science and Technology [29] C. Lu, Y. Wu, F. Mai, W. Chung, C. Wu, W. Lin, C. Chen, J. Mol. Catal. 310 (2009)
159–165.
Excellence Sail Plan of Shanghai (14YF1410800) and the National [30] C.C. Chen, H.J. Fan, J.L. Jan, J. Phys. Chem. C 112 (2008) 1197–11962.
Natural Science Foundation of China (21302127 and 21502116). [31] C.C. Chen, J. Mol. Catal. A 264 (2007) 82–92.
[32] F.D. Mai, C.C. Chen, J.L. Chen, S.C. Liu, J. Chromatogr. A 1189 (2008) 355–365.
[33] R. Wang, X. Xu, Y. Zhang, Z. Chang, Z. Sun, W.F. Dong, Nanoscale 7 (2015)
Appendix A. Supplementary data 11082–11092.
[34] J. Han, Z. Liu, K. Guo, X. Zhang, T. Hong, B. Wang, Appl. Catal. B 179 (2015)
Supplementary data associated with this article can be found, 61–68.
[35] W.Y. Hernandez, M.A. Centeno, F. Romero-Sarria, J.A. Odriozola, J. Phys. Chem.
in the online version, at http://dx.doi.org/10.1016/j.mcat.2017.04.
C 113 (2009) 5629–5635.
014. [36] Y.B. Han, H. Yu, W.P. He, Int. Shan, J. Hydrogen Energy 38 (2013)
10293–10304.
[37] D.R. Abd El-Hafiz, M.A. Ebiad, R.A. Elsalamony, L.S. Mohamed, RSC Adv. 5
References
(2015) 4292–4303.
[38] P. Burroughs, A. Hamnett, A.F. Orchard, G. Thornton, J. Chem. Soc. Dalton
[1] T. Inoue, A. Fujishima, S. Konishi, K. Honda, Nature 277 (1979) 637–638. Trans. 17 (1976) 1686–1698.
[2] G. Centi, S. Perathoner, ChemSusChem 3 (2010) 195–208. [41] B.M. Reddy, L. Katta, G. Thrimurthulu, Chem. Mater. 22 (2010) 467–475.
[3] L. Hammarstrom, S. Hammes-Schiffer, Acc. Chem. Res. 42 (2009) 1859–1860. [43] S. Sarkar, A. Makhal, T. Bora, S. Baruah, J. Dutta, S.K. Pal, Phys. Chem. Chem.
[4] S.C. Roy, O.K. Varghese, M. Paulose, C.A. Grimes, ACS Nano 4 (2010) Phys. 13 (2011) 12488–12496.
1259–1278. [44] P. Jiang, J.J. Zhou, H.F. Fang, C.Y. Wang, Z.L. Wang, S.S. Xie, Adv. Funct. Mater.
[5] N. Zhang, S.X. Ouyang, P. Li, Y.J. Zhang, G.C. Xi, T. Kako, J.H. Ye, Chem. 17 (2007) 1303–1310.
Commun. 47 (2011) 2041–2043. [45] K. Vanheusden, C.H. Seager, W.L. Warren, D.R. Tallant, J.A. Voigt, Appl. Phys.
[6] Y.G. Wang, B. Li, C.L. Zhang, L.F. Cui, S.F. Kang, X. Li, L.H. Zhou, Appl. Catal. B Lett. 68 (1996) 403–405.
130 (-131) (2013) 277–284. [46] J. Lee, H.S. Shim, M. Lee, J.K. Song, D. Lee, J. Phys. Chem. Lett. 2 (2011)
[7] W.Y. Lin, H. Frei, J. Am. Chem. Soc. 127 (2005) 1610–1611. 2840–2845.
[8] W.B. Hou, W.H. Hung, P. Pavaskar, A. Goeppert, M. Aykoland, S.B. Cronin, ACS [47] L.Y. Chen, Y.T. Yin, RSC Adv. 3 (2013) 8480–8488.
Catal. 1 (2011) 929–936. [48] Y.Z. Chen, D.Q. Zeng, K. Zhang, A. Lu, L. Wang, D.L. Peng, Nanoscale 6 (2014)
[9] L. Collado, A. Reynal, J.M. Coronado, D.P. Serrano, J.R. Durrant, V.A. de la Pena 874–881.
O’Shea, Appl. Catal. B 178 (2015) 177–185. [49] S. Khanchandani, S. Kundu, A. Patra, A.K. Ganguli, J. Phys. Chem. C 116 (2012)
[10] W.N. Wang, W.J. An, B. Ramalingam, S. Mukherjee, D.M. Niedzwiedzki, S. 23653–23662.
Gangopadhyay, P. Biswas, J. Am. Chem. Soc. 134 (2012) 11276–11281. [50] S. Martha, K.H. Reddy, K.M. Parida, J. Mater. Chem. A 2 (2014) 3621–3631.
[11] S.J. Xie, Y. Wang, Q.H. Zhang, W.P. Deng, Y. Wang, ACS Catal. 4 (2014) [51] Q. Li, B.D. Guo, J.G. Yu, J.R. Ran, B.H. Zhang, H.J. Gong, Yan, J. Run, J. Am. Chem.
3644–3653. Soc. 133 (2011) 10878–10884.
[12] M.M. Gui, S.P. Chai, B.Q. Xu, A.R. Mohamed, RSC Adv. 4 (2014) 24007–24013. [52] N. Zhang, Y. Zhang, Y.J. Xu, Nanoscale 4 (2012) 5792–5813.
[13] C.J. Wang, R.L. Thompson, J. Baltrus, C. Matranga, J. Phys. Chem. Lett. 1 (2010) [53] A.M. Turek, I.E. Wachs, J. Phys. Chem. 96 (1992) 5000–5007.
48–53. [54] J. Saussey, J.C. Lavalley, C. Bovet, J. Chem. Soc. Faraday Trans. 78 (1982)
[14] M.L. Li, L.X. Zhang, X.Q. Fan, Y.J. Zhou, M.Y. Wu, J.L. Shi, J. Mater. Chem. A 3 1457–1463.
(2015) 5189–5196. [55] H. Noei, C. Woll, M. Muhler, Y.M. Wang, J. Phys. Chem. 115 (2011) 908–914.
[15] W.G. Tu, Y. Zhou, Q. Liu, S.C. Yan, S.S. Bao, X.Y. Wang, M. Xiao, Z.G. Zou, Adv. [56] T. Hikov, A. Rittermeier, M.B. Luedemann, C. Herrmann, M. Muhlerb, R.A.
Funct. Mater. 23 (2013) 1743–1749. Fischer, J. Mater. Chem. 18 (2008) 3325–3331.
[16] P.Q. Wang, Y. Bai, P.Y. Luo, J.Y. Liu, Catal. Commun. 38 (2013) 82–85. [57] S. Fujita, M. Usui, H. Ito, N. Takezawa, J. Catal. 157 (1995) 403–413.
[17] W.G. Tu, Y. Zhou, Q. Liu, Z.P. Tian, J. Gao, X.Y. Chen, H.T. Zhang, J.G. Liu, Z.G. [58] Y. Zhang, R.Q. Yang, N. Tsubaki, Catal. Today 132 (2008) 93–100.
Zou, Adv. Funct. Mater. 22 (2012) 1215–1221. [59] G.R. Dey, J. Nat. Gas. Chem. 16 (2007) 217–226.
[18] O.K. Varghese, M. Paulose, T.J. LaTempa, C.A. Grimes, Nano Lett. 9 (2009) [60] G.H. Liu, N. Hoivik, K.Y. Wang, H. Jakobsen, Sol. Energy Mater. Sol. Cells 105
731–737. (2012) 53–68.
[19] J. Lv, W.F. Fu, C.Y. Hu, Y. Chen, W.B. Zhou, RSC Adv. 3 (2013) 1753–1757. [61] H.Q. Sun, S.B. Wang, Energy Fuels 28 (2014) 22–36.
[20] H.C. Shown, Y.C. Hsu Chang, C.H. Lin, P.K. Roy, A. Ganguly, C.H. Wang, J.K. [62] A. Fisher, A.T. Bell, J. Catal. 178 (1998) 153–173.
Chang, C.I. Wu, L.C. Chen, K.H. Chen, Nano Lett. 14 (2014) 6097–6103. [63] L.F. Liao, C.F. Lien, D.L. Shieh, M.T. Chen, J.L. Lin, J. Phys. Chem. B 106 (2002)
[21] J. Graciani, K. Mudiyanselage, F. Xu, A.E. Baber, J. Evans, S.D. Senanayake, D.J. 11240–11245.
Stacchiola, P. Liu, J. Hrbek, J.F. Sanz, J.A. Rodriguez, Science 345 (2014) [64] J. Park, J. Graciani, E. Jaime, S. Dario, S.D. Senanayake, L. Barrio, P. Liu, J.F. Sanz,
546–550. J. Hrbek, J.A. Rodriguez, J. Am. Chem. Soc. 132 (2010) 356–363.
[22] H. Park, H.H. Ou, A.J. Colussi, M.R. Hoffmann, J. Phys. Chem. A 119 (2015) [65] N.M. Dimitrijevic, B.K. Vijayan, O.G. Poluektov, T. Rajh, K.A. Gray, H.Y. He, P.
4658–4666. Zapol, J. Am. Chem. Soc. 133 (2011) 3964–3971.
[23] F. Sastre, A. Corma, H. García, J. Am. Chem. Soc. 134 (2012) 14137–14141.
[24] S.J. Xie, Y. Wang, Q.H. Zhang, W.P. Deng, Y. Wang, ACS Catal. 4 (2014)
3644–3653.

Vous aimerez peut-être aussi