A facile one step route to synthesize WO3 nanoplatelets for CO oxidation and photodegradation of RhB: Microstructural, optical and electrical studies

Article abstract of DOI:10.1039/c6ra13500e

Two-dimensional nanoplatelets of tungsten trioxide (NP-WO₃) were synthesized at room temperature, using a wet chemical method, without any surfactants or templates; aqueous mineralization was obtained by simply adjusting the pH. The resulting nanostructures were characterized using X-ray diffraction combined with Rietveld refinements, Raman and UV-Vis spectroscopies. Their morphologies and sizes were analyzed by scanning and electron microscopies. The electrical, optical, catalytic and photocatalytic properties of the NP-WO₃ nanoplatelets were then determined and compared to the ones of pseudospherical (PS-WO₃) standard samples. Nanoplatelets as well as pseudospherical particles crystallized in the single orthorhombic WO₃ phase. The Rietveld refinement calculations and HRTEM analyses revealed a strong distortion of the WO₆ octahedra, according to the W-O splitting. The electrical conductivity of WO₃ compact pellets showed that both samples were semi-conducting with a weak difference in activation energies. Using a homemade photocatalytic device, the NP-WO₃ particles used as photocatalyst in an aqueous medium, exhibited a significant efficiency to decompose rhodamine B over their large exposed surface (010), compared to PS-WO₃ particles. These NP-WO₃ particles were also used as the catalytic material for oxidation in an air-CO gas flow. They exhibited catalytic activity higher than the one in the PS form.

Full text of DOI:10.1039/c6ra13500e

View Article Online
View Journal
RSC Advances
This article can be cited before page numbers have been issued, to do this please use: N. Dirany, M.
Arab, V. Madigou, C. Leroux and J. R. GAVARRI, RSC Adv., 2016, DOI: 10.1039/C6RA13500E.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. This Accepted Manuscript will be replaced by the edited,
formatted and paginated article as soon as this is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/advances

Page 1 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
A facile one step route to synthesize WO3 Nanoplatelets for CO oxidation and
photodegradation of RhB: microstructural, optical and electrical studies
N. Dirany, M. Arab*, V. Madigou, Ch. Leroux and J. R. Gavarri
Université de Toulon, IM2NP, UMR CNRS 7334, BP 20132, 83957, La Garde, France
*Corresponding author: madjid.arab@univ-tln.fr
Abstract
Twoꢀdimensional nanoplatelets of tungsten trioxide (NPꢀWO3) were synthesized at room
temperature, using a wet chemical method, without any surfactants or templates: aqueous
mineralization was obtained by simply adjusting the pH. The resulting nanostructures were
characterized using Xꢀray diffraction combined to Rietveld refinements, Raman and UV–Vis
spectroscopies. Their morphologies and sizes were analyzed by scanning and electron
microscopies. The electrical, optical, catalytic and photocatalytic properties of NPꢀWO3
nanoplatelets were then determined and compared to the ones of pseudospherical (PSꢀWO3)
standard samples. Nanoplatelets as well as pseudo spherical particles crystallized in the single
orthorhombic WO3 phase. The Rietveld refinement calculations and HRTEM analyses
revealed a strong distortion of WO6 octahedra, according to the WꢀO splitting. The electrical
conductivity of WO3 compacted pellets showed that both samples were semiꢀconducting with
a weak difference in activation energies. Using a homemade photocatalytic device, the NPꢀ
WO3 particles used as photocatalyst in aqueous medium, exhibited a significant efficiency to
decompose Rhodamine B over their large exposed surface (010), compared to PSꢀWO3
particles. These NPꢀWO3 particles were also used as catalytic material for the oxidation in an
airꢀCO gas flow. They exhibited catalytic activity higher than the one of PS form.
Keywords: tungsten trioxide, nanoplatelets, octahedra distortion, band gap, CO oxidation,
photocatalysis.
I.
Introduction
In recent years, various studies have clearly demonstrated the relevance of certain
chemical syntheses to control the sizes and morphologies of nanomaterials, both in the basic
sciences as in industrial applications. The nanoparticles having low dimension morphologies
1D and 2D can present physical and chemical properties strongly differing from the ones of
bulk materials. They can be utilized as high effective catalysts, and applied in nanodevices for
electronic, magnetic, and optical applications.1ꢀ7
1

RSC Advances
Page 2 of 30
View Article Online
DOI: 10.1039/C6RA13500E
Tungsten trioxide WO3 is an important nꢀtype semiconductor with high stability and wide
tunable band gap, around 2.6 eV at room temperature.5,8,9 The gap of the structure is
depending upon bonding–antibonding W–O interaction, which correlating to the overlap in
each direction. As a semiconductor material, WO3 has a broad range of potential applications
in various fields such as photocatalysis, heterogeneous catalyst,10,11 gas sensor,12
13
14
electrochromic devices, water splitting to produce H2 and O2 and solar energy devices15.
The properties of WO3 are critically dependent on structure, size and morphology.16
Therefore, a great effort has been devoted to synthesize tungsten oxide with well controlled
shapes and defined sizes with powerful performances. WO3 has many polymorphs depending
on the synthesis method and the temperature range. The idealized structure of WO3 is cubic,
with a three dimensional network of corner sharing octahedra. The polymorphs correspond to
more or less distortions of this network, tilting of the octahedra or cation displacements,
which lead then to different crystallographic structures of WO3.17 At temperature ranges
between ambient and 500 °C, the monoclinic and orthorhombic phases are predominant.
In order to control the morphology and size of tungsten oxides, many efforts have been
devoted to the syntheses route. Several methods were developed to obtain different
morphologies of WO3 such as hydrothermal reaction,18 template assisted methods,19 chemical
vapor deposition20 and electroꢀdeposition.21 Tungsten oxide with particular grain shapes has
been obtained, including oneꢀdimensional nanoscale building blocks22 and other advanced
shapes such as flowerꢀlike,23 urchinꢀlike24 and hierarchical hollow architectures.25 In this
case, GimenoꢀFabra et al. obtained tungsten oxide nanoplates using nearꢀ and superꢀcritical
water in a continuous flow system with tungsten (VI) ethoxide as precursor and calcination
temperatures above 350°C.26 Gehong Zhang et al. fabricated WO3 nanosheets of ~200ꢀ500 nm
via a fast and organic additivesꢀfree hydrothermal method.27 These methods require, however,
complicated operating procedures and/or the addition of various surfactants, for a better
control of the morphology. However, it is still a significant scientific challenge to develop
simple and reliable synthetic methods with controllable morphologies and uniform sizes
through mild and low cost chemical routes.
Among various nanostructures, twoꢀdimensional inorganic nanomaterials have attracted
considerable attention because they could offer new opportunities in applied surface
materials. The 2D sheet structure generally exhibits high surface energies, provide active sites
for catalysts and sensors and can be used as building blocks to construct complex
2

Page 3 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
nanostructures. 28 In addition, due to their huge light exposed surface area, they show supreme
advantages for photocatalysis compared to the 1D and 3D counterparts.27
In this work, we present for the first time a wellꢀcontrolled synthesis of orthorhombic
quadrangular nanoplates of WO3 (called NPꢀWO3), using a mineralization process free from
any oneꢀstep template, without addition of surfactant. We compare the properties of these
nanoplatelets to the ones of WO3 powders having pseudospherical form (called PSꢀWO3) and
considered as standard samples. The samples were characterized by Xꢀray diffraction (XRD)
and Raman spectroscopy. The morphological and local observations were obtained from
scanning electron microscopy (SEM), transmission electron microscopy (TEM), high
resolution transmission electron microscopy (HRTEM) coupled to the energy dispersive Xꢀ
ray spectroscopy (EDS). Finally, a systematic comparison between the properties of the NP
and PS WO3 samples was undertaken, with the studies of the electrical conductivity, the
catalytic activity in presence of CO gas, the energy gap determination and finally the
photocatalytic activity in presence of rhodamine B, in aqueous solution.
II.
Experimental setup
1. Synthesis and structural characterization
The NPꢀWO3 particles were synthesized using sodium tungsten Na2WO4.2H2O as precursor
by a simple free template precipitation method. In a typical synthesis, a given amount of
precursor was dissolved in a 20 mL of distilled water under continuous stirring, and HCl (2M)
diluted solution was added until the formation of pale yellow precipitates. The solid was
transferred into 60 mL of diluted HCl solution (2M). After continuous stirring for 24 h, the
mixed solution was aged at 80°C for 96 h. The yellow precipitate was filtered, washed with
distilled water several times and finally calcined at 500 °C for 2h in air, with a heating rate of
5 °C minꢀ1.
The PSꢀWO3 powders were synthesized by dissolving 6g of ammonium tungstate hydrate
(NH4)10(W12O41).6H2O in 30 mL of absolute ethanol solution. 3 mL of HNO3 was added to
29
the solution and then stirred by a magnetic stirrer at 70 °C. The solution was heated until
the solvent was evaporated. The yellow precipitate was filtered, washed with distilled water,
dried at 80 °C. The obtained sample was annealed at 500 °C for 2h in air with a heating rate
of 5°C minꢀ1.
3

RSC Advances
Page 4 of 30
View Article Online
DOI: 10.1039/C6RA13500E
The morphological characterizations of the powder were conducted by Scanning Electron
Microscope (SUPRA 40Vp Colonne Gemini Zeiss SEM operated at 10Kv) and Transmission
Electron Microscope (Tecnai G2 operating at 200 kV, with a point to point resolution of 0.25
nm). The crystal structure of WO3 was determined via Xꢀray diffraction XRD (EMPYREAN
Panalytical diffractometer) with a Cu.Kα source. Xꢀray patterns were compared with those of
the International Centre for Diffraction Data (ICDD). The DBWS 2.16 software was used to
refine the structural parameters of the different products.30,31 The crystal cell parameters of the
as synthetized materials were calculated from the Rietveld method. The program uses the
NewtonꢀRaphson algorithm to refine and quantify the intensities and profile of the peaks.
Diffraction peak profile adjustment is performed using the modified PseudoꢀVoigt function
containing the mixtures of Lorentzian and Gaussian contributions. The minimization was
carried out using the reliability index parameters such as the Bragg factor (RBragg), comparing
to the calculated and observed intensities (Ical and Iobs), the expected factor (Rexp), the
weighted profile factor (Rwp) and the goodness of fit χ. All these parameters were used as
numerical criteria of the quality of refinement calculations. From the optimized profile, we
used the DebyeꢀScherrer (DS) approach to calculate the particles size using the following
relationship: 32ꢀ34
ꢀ λ
D = ꢁ ꢂꢃꢄꢅ
(1)
where D is the crystallite size, K is the form factor (K = 0.9 in the case Gaussian profiles), λ is
ꢆ
ꢇ
the XꢀRay radiation wavelength (λ= 0.15406 nm), ∆ 2θ is the angular broadening of Bragg
peaks resulting from the size effect and θ is the Bragg angles in radians. The integral breadth
β value was calculated using a gaussian approximation of peak profiles such as βꢈ =
ꢆ
ꢇꢈ
∆ 2θ ꢉꢊꢋꢌ − ∆ 2θ ꢉꢍꢊꢎꢏ. ∆ 2θ
ꢆ
ꢇꢈ
ꢆ
ꢇ
ꢆ
ꢇ
and ∆ 2θ
ꢉꢍꢊꢎꢏ
are respectively the full width at half
ꢉꢊꢋꢌ
maximum (FWHM) of Bragg peak of the synthesized samples and the standard WO3 sample
ꢆ
ꢇ
obtained after a thermal treatment of 24 h at 600°C. The difference ∆ 2θ is the broadening
resulting from size effect. A theoretical specific surface area SA was calculated using the
approximation of particles having the form of square sheets, with an average thickness e, an
average linear side size l, and having a density ρ ( 7.28 g/cm3) :
1
2
2
ꢒ
Sꢐ = ꢑ
+ ꢒ ꢓ ꢒ
(2)
e
ρ
l
To highlight the efficiency of the morphology, we compare the studied properties of the
nanoplatelets to the pseudospherical nanoparticles. In the case of spherical particles, we use
4

Page 5 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
35
⁄
the classical formula of the specific surface area: Sꢐ = 6 ρd where d is the mean particle
diameter (or coherence length). The apparent size (e, l and d) were determined from SEM and
TEM images. Considering the NP geometry with defined dimensions (l and e), the percentage
⁄∑
ꢔ of each facet can be calculated as following: τꢕꢖ ꢗꢘꢘ , ꢘꢗꢘ ,ꢆꢘꢘꢗꢇ = Sꢕ ꢕꢖ ꢗꢘꢘ , ꢘꢗꢘ ,ꢆꢘꢘꢗꢇ Sꢕ,
ꢆ
ꢇ ꢆ
ꢇ
ꢆ
ꢇ ꢆ
ꢇ
where Sꢕ is the surface area of the i facet.
Raman spectroscopy was used to analyze the characteristic vibration modes associated to the
bonds present in the obtained material according to the constituent’s element. The equipment
used to record the various vibration spectra was a spectrometer Horiba JobinꢀYvon HR800
LabRam, spatially resolved to 0.5 microns, by means of an optical microscope with a 100X
objective. The 514.5 nm line of an Arꢀion laser was used as the excitation source; the
photonic power applied to the samples was limited to 5 ꢁW with an acquisition time of 30
seconds.
2. Properties
Optical analysis: The analysis of UVꢀVisible diffuse reflectance spectra (DRS) of the
different samples has been done using a LAMBDA 190s UV/Vis/NIR spectrometer along
with150ꢀmm integrating sphere. A small volume powder sample was manually pressed into a
3.2 cm powder cup which was clamped (attached) on the external port of the integrating
sphere. The measures were performed in the range of 250ꢀ800nm at room temperature. The
baseline was carried out using a calibrated Spectral on reflectance standard.
Raman Spectra were obtained using a spectrometer Horiba JobinꢀYvon HR800 LabRam,
spatially resolved to 0.5ꢁm, by means of an optical microscope with a 100 objective.
Electrical analysis: The electrical measurements were performed using electrical impedance
spectroscopy (EIS) under air at atmospheric pressure, in the same temperature range explored
in catalytic experiments. The SOLATRON (SI 1260 AC) impedance analyzer was used in the
frequency range of 1–106 Hz. For a given temperature, each measurement was recorded after
stabilization of the heating process. The samples were compacted in the form of pellets at a
pressure of 4 kbar, and then placed between two cylindrical platinum electrodes.
Catalytic properties: The catalytic studies were performed using a homemade device
allowing the reaction of a powder catalyst with a gas flow containing a gas to be oxidized, at a
fixed temperature. The product resulting from the gasꢀsolid interaction was analyzed by
Fourier Transform InfraRed spectroscopy (FTIR, Bruker–Tensor 27), coupled to a quadrupole
5

RSC Advances
Page 6 of 30
View Article Online
DOI: 10.1039/C6RA13500E
mass spectrometer (MS, MKS – CIRRUS). This study was conducted with 1250 ppm of CO
diluted in air and FTIR spectra were acquired every 30s. For further information, the details of
the experiment are reported in a previous work.36
Photocatalytic activity. The photocatalytic activity of samples was characterized from the
UV irradiation of aqueous solutions in which Rhodamine B (RhB) and NP or PS active
particles of WO3 were present at fixed concentrations and at a stabilized temperature. The
degradation of RhB was observed under UV light using an 80 W mercury lamp, operated at
254 nm (or 4.88 eV). For this purpose, 250 mg of WO3 powders was dissolved in 250 mL of
RhB (10 mg/L) and then exposed to ultrasound for 5 min to eliminate aggregates. Prior to the
irradiation, the suspension was stirred for 60 min in darkness to reach an adsorptionꢀ
desorption equilibrium. After this time, the light source was turned on. During the reaction,
samples were taken from the reactor at different time intervals. Given their low values, the
concentrations of RhB in these irradiated solutions can be considered as proportional to the
observed intensity of the RhB absorption peaks at 554 nm. Thus, the degradation rate of RhB
was analyzed using a UVꢀvisible spectrophotometer (Shimadzu UVꢀ24646) and expressed as
Cn/C0, where Cn is the concentration of RhB for each irradiated time, and C0 is the initial
concentration.
III.
Results and discussion
1. Structural and morphological characterization
The synthesis method leads to well crystallized grains for both samples, though with different
morphologies. The SEM images show that the powders consist either of large quantity of
platelets (Fig. 1.a) and particles without shape that can be considered as pseudoꢀspherical
grains (Fig. 1.b). The images show also that the grains with different morphologies tend to
form a dense aggregate, particularly for PS particles. The insert TEM images show a
quadrangular NPꢀWO3 with a mean lateral size of 175 nm and an average thickness of 40 nm.
The PSꢀWO3 grains are smaller than the nanoplatelets, with a size distribution centered on 60
nm. The platelet sample presents a relative uniform size distribution compared to the
pseudospherical powder.
Fig. 1: SEM images of the WO3 powders: a) NPꢀWO3 and b) PSꢀWO3. The TEM pictures in
insert show the size dispersion of the grains
6

Page 7 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
Fig. 2 shows the XRD patterns of the samples. Both patterns correspond to well crystallized
tungsten oxide phase. The observed Bragg peaks of the diffractograms are characteristic of a
single phase WO3. Moreover, no characteristic residual peaks from other crystalline
impurities were detected. At the synthesis temperature used, WO3 can be either monoclinic or
orthorhombic,37 with very closed lattice parameters. Thus, Rietveld refinements were done
with the γꢀWO3 monoclinic structure P121/C (N°14)38 as well as the βꢀWO3 orthorhombic
phases Pnma (n°62)37 and Pmnb (n°60).39
The refinement is significantly better with the orthorhombic structure (reliability factors Ri <
10%) compared to the monoclinic phase. The results of the Rietveld refinements are
summarized in Table 1, where the refined lattice parameters are reported. The initial atom
coordinates have been exported from CIF datasheet of WO3 (836ꢀICSD database) structure
which has a low shift of Bragg peaks compared to our experimental diffraction patterns.
According to the DebyeꢀScherrer calculation, we obtained a mean size of 64.3 (± 1.0) nm and
82.9 (± 1.0) nm for NPꢀWO3 and PSꢀWO3 respectively.
Fig. 2: Rietveld analyses (space group Pnma n°62): observed and calculated XRD profiles of
the structures: NPꢀWO3 and PSꢀWO3.
Due to the specific Wyckoff positions, the coordinate’s atoms were fixed to start the
refinement procedure. We first refined the W coordinates then those oxygen atoms of the
WO3 structures (due to their low scattering factors). The refinements were carried out with
orientation factors, anisotropic and isotropic respectively for platelets and pseudospheres.
Moreover, the difference between XRD pattern profiles associated with each morphology can
be related to the degree of crystallization and to the local order of atoms. The refinement
position gave set structural parameters with low deviation, but sufficient to distinguish the
two structures. According to the of XRD pattern profile, the samples show different textures,
which characterized by different intensities of the main peaks at 23.10°, 23.59° and 24.32°
corresponding to (002), (200) and (020) plans. For the isotropic pseudospheres the intensities
are similar, but for nanoplatelets, we obtained a higher intensity for (020) orientation than
(002) and (200), confirming the anisotropic form.
7

RSC Advances
Page 8 of 30
View Article Online
DOI: 10.1039/C6RA13500E
Table 1: Refinement structural parameters and Coordinates atoms of WO3 phase based on
two octahedra structure (Space group of Pnma). Refinement calculation based on CIF
datasheet of ICSD (Collection code 836).
The obtained refinement parameters were exploited by Carine software for the cell
representation. The crystal structure of orthorhombic WO3 is shown in fig. S2 (supporting
information). The Schema presents a multilayer configuration of alternating octahedral groups
WO6 connected in three space directions, which the atoms of W are located in the center of
the octahedra and oxygens are at the vertices, thus each oxygen is forming a connection WꢀOꢀ
W.
Using Carine software and our refinement results, we extracted the main bond lengths and
bond angles, reported in Table 2. It should be noted that the refinement was occurred with a
multiplicity of order two (two octahedra Oct1 and Oct2), according to the agreement phase
datasheet.
In this case, we obtained a layered structure based on the alternating and distorted octahedra
WO6. In WO6 regular octahedrons, two chemical bonds are perpendicular to a plane in which
the four other chemical bonds are located. In the present structures, the two shorter bonds vary
but are closed to 1.7 Å for Oct1 and Oct2. The four longer bonds strongly vary in the range
1.99 to 2.23 Å, for Oct1 and Oct2. The OꢀO distances to the vertices between nearest
neighbors, vary slowly from 2.49 to 2.78 Å. The OꢀO corner distances of the octahedra vary
from 3.6 to 3.84 Å while the WꢀW distance reaches 3.95 Å: all these values characterize flat
irregular octahedra. These distortions were also observed with the variation of OꢀWꢀO angles,
as summarized in Table 2.
Generally, WO3 exhibits a perovskite structure even in the orthorhombic phase at higher
temperature. So, the irregular bond distances and angles of polyhedron are at the origin of the
deformation and distortion. For both samples, the W positions shift from the midpoints of the
octahedral to form zigzag chains in the a and b directions. In this case, the orthorhombic
system shows alternation of long and short W–O bonds in both the a and c directions, but not
in the b direction. However, the splitting in c direction is larger for platelets than spheres
octahedral.
Table 2: Selected bond distances and angles in different nanostructure of WO3, with two
optimized WO6 octahedral of the unit cell.
8

Page 9 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
The transmission electron microscopy images and associated electron diffraction patterns
showed that each nanoplatelet is a single crystal. Most encounted platelets appear as square,
although rectangular ones can also be found (Fig. 3b). When the nanoplatelets lay flat, the
corresponding zone axes show a quadratic symmetry for the strong spots, remaining of the
ideal cubic structure of WO3. Fig. 3a and 3c can be distinguished by their superstructure spots
(with regards to the cubic structure). Fig. 3a shows weak superstructure spots along two
perpendicular directions as Fig. 3c has strong superstructures spots along one direction. These
supplementary spots could only be indexed in the orthorhombic β−WO3 phase with space
group 62 (Pnma), with two perpendicular zone axes [010] and [001] respectively.
Nanoplatelets laying edge on brought information along the growth direction of the platelets.
Again, the strong spots of the diffraction patterns (Fig.
4a and 4b) exhibit a quadratic symmetry, showing that the zone axis is perpendicular to those
of fig. 3a and 3c. The weak superstructure spots could be indexed in the β−WO3 structure and
Fig. 4a and 4b correspond to [100] zone axis. Depending on the platelets, more (Fig. 4b) or
less streaks (Fig. 4a) along b* can be observed in the diffraction patterns, indicating defects in
the piling up of (0k0) planes, thus along the main growth direction of the nanoplatelets.
HRTEM images taken along the same zone axis show the irregular piling up of (0k0) planes
(Fig. 4c). Variations of more than 10 % of the interreticular distances were measured.
Dislocations are revealed in the Inverse Fast Fourier Transform (Fig. 4d) (HRTEM images
and diffraction patterns are in supplementary information).
Fig. 3: TEM image and electronic diffraction patterns of flat laying WO3 platelets: a) and c)
correspond respectively to perpendicular [010] (a) and [001] (c) zone axes of the
orthorhombic βꢀWO3 structure.
According to the obtained results (table 2), the interreticular distance can be identified as the
heavy atoms distances WꢀW and they vary in three directions. So, the measured distance from
the HRTEM image (fig. 4c) in the c direction is in agreement with that deduced from
refinement calculations: 3.94 Å. Based on the above TEM observation and structural analysis,
we conclude that the top and bottom facets are (010) and lateral facets are (100) and (001).
From fig. S2b (supporting information), we can see that (010) facets of the orthorhombic
9

RSC Advances
Page 10 of 30
View Article Online
DOI: 10.1039/C6RA13500E
structure are O terminated planes. Compared with the (010) facets, there are more W atoms
exposed on the surfaces of (100) and (001) planes (fig. S2a and S2c). It’s indicated that the
(010) plan have more active oxygen sites. Therefore, the anisotropic property of the
nanoplatelets will exhibit different physical/chemical properties than isotropic PSꢀWO3.
Fig. 4: HRTEM image and electron diffraction pattern of WO3 platelets laying edge on a) and
c) correspond to the [100] zone axis of the orthorhombic structure, c) is an HRTEM image
showing the piling up of (020) planes. The Inverse Fast Fourier Transform d) emphasizes the
defects in the piling up of the plans, pointed by arrows.
Raman analysis : According to previous research,40ꢀ42 the Raman activity of phonons is
caused by the successive deformation of the ideal cubic lattice. Most peaks below 200 cmꢀ1
are usually attributed to lattice modes, whereas the mid and high frequency regions
correspond to deformation and stretching modes, respectively. Since inorganic compounds
have vibration bands mainly below 1200 cmꢀ1, an investigation of Raman spectroscopy of
WO3 show mainly seven vibration bands in range of 50ꢀ1000 cm−1, as reported in fig. 5. In the
range of 50 to 150 cm−1, we observe a typical characteristic peaks of the orthorhombic
43ꢀ45
crystalline phase at low frequencies, which are associated to lattice vibrational modes.
The peaks at 273 and 330 cm−1 are assigned to the bending vibration OꢀWꢀO. 44,46
Fig. 5: Raman spectra of WO3 structures
The bands in the range 700ꢀ850 cmꢀ1 are attributed to either the antisymmetric stretch of Wꢀ
OꢀW bonds or the symmetric stretch of OꢀWꢀO bonds.47,48 Thus, the intense peaks centered at
808 and 718 cm−1 are typical Raman peaks of crystalline phase of WO3, which correspond to
the stretching vibrations of the bridging oxygen43,49 and these are assigned to WꢀO stretching
(ν), WꢀO bending(δ) and OꢀWꢀO deformation (γ) modes, respectively.46,50 Close to the high
frequency Raman bands of PS and NP samples, we also observe large undulations with two
different aspects in PS and NP samples (maximum close to 640 nm): this could be ascribed to
specific defects in the lattice, probably linked to a nonꢀstoichiometry in oxygen associated
with local distortions. In addition, the two highꢀfrequency Raman bands of NPꢀWO3 particles
are characterized by thinner profiles and higher maximum compared to the ones of PSꢀWO3
10

Page 11 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
particles. These observations are correlated with the different profiles of the low frequency
bands (lattice modes). In this case, the absorption band at 135 nm is higher for PS particles.
As reported in the table 2, the WꢀO bond is smaller for PS particles (1.68Ǻ) than NP, which
the shorter bond rather promotes more overlap orbitals.
Generally, these differences observed for PS and NP samples could be due to a better
crystallization level coupled to the anisotropy of the platelets for the NP samples. Taking into
account to the WO6 distortions and the presence of the linear defects in the plane of the plates,
nanoplates present deformation and stretching modes more intense than pseudospherical, in
agreement to the bonds and angles changes deduced from XRDꢀrefinement and observed
defects from HRTEM images.
2. UV–Vis Diffuse Reflectance
As the values of optical band gap energies generally condition the photocatalytic activity of
materials, it was necessary to characterize the influence of morphology on optical properties
of the as prepared NP and PS WO3. Let us recall that the current values of band gaps of WO3
materials are found to be close to 2.6 eV. The UVꢀvisible diffuse reflectance spectroscopy
(DRS) at room temperature was used to determine the band gap energies, noted Eg, of Np and
PS WO3.
The intrinsic absorption edge of the powders can be evaluated and discussed in terms of the
indirect interband transition. The optical energy band gap is obtained by using the Kubelkaꢀ
Munk method.51 This method consists in plotting F(R∞) versus photon energy and extracting
Eg by a linear fit (see Fig 6). R∞ designates the absolute reflectance of the sample at infinite
thickness and F(R∞) is the conversion of the diffuse reflectance measurements R∞ to F(R∞)
according to the following equation:
ꢝ = ꢀ(3)
ꢆ
ꢇ
ꢗꢚꢛ
ꢜ
ꢆ
F Rꢙ
ꢇ
=
ꢈꢛ
ꢄ
ꢜ
where R∞ designates the absolute reflectance of the sample at infinite thickness. K represents
the molar absorption coefficient and S is the scattering coefficient. F(R∞) of the different
samples treated at 500C, are plotted in fig. 6. It is expected to show a linear behavior in the
higher energy region, which should correspond to a strong absorption near the absorption
edge. Using linear fitting method, extrapolating the linear portion of this straight line to zero
absorption edge gives the optical energy band gap of the powders.
11

RSC Advances
Page 12 of 30
View Article Online
DOI: 10.1039/C6RA13500E
Fig. 6: Variation of F(R∞) as a function of irradiation energy, for the PS and NP samples, R∞
being determined from UVꢀVis absorbance spectra.
The band gap absorption of the WO3 with two morphologies was estimated to be 2.65 and
2.82 eV for pseudospheres and nanoplatelets, respectively. The difference between the gap
can be attributed to different factors such shape, crystallite size,52 lattice parameters,53 degree
of structural orderꢀdisorder in the lattice and the MossꢀBurstein effect.54ꢀ56 The change of
these factors depends on the distortion polyhedra as mentioned in the table 2, which the bond
length and angles vary from one morphology to another. In our case, it should be remarked
that the band gap of NP particles is greater than the one of the PS particles: this could be
ascribed to the difference in morphology and sizes of particles, keeping in mind that strong
anisotropy and smaller sizes are observed for the NP particles. It is well known that the band
gaps Eg generally increase as crystallite sizes decrease, in the case of semiconducting
particles.52
Moreover, the electronic structures of pure WO3 phases may be interpreted in terms of W–O
bond splitting, i.e. different bond length in the same direction. As reported by Chatten et al.,8
the presence of W–O bond splitting induces an increase in the band gap though the magnitude
of the splitting in the three directions. In the case of the orthorhombic phase, the structure
show fairly strong splitting in two directions. The magnitude of the band gap in our materials
is dominated by the splitting in the c direction (1.907 to 2.244 Å) and also splitting in a
direction (1.692 to 1.864 Å), in the same order to those obtained by DFT calculation.
According to the data of the table 2, the splitting is higher for NP than PS oxides, therefore a
large gap for NP 2.82 eV. This behavior resulted in shorter W–O bond, increased overlap of
O2p and W5d and increased bondingꢀantibonding interaction.57
3. Electrical Impedance Spectroscopy
Electrical measurements were obtained with a series of Nyquist representations for the
optimized WO3 sample as a function of temperature. To interpret the semicircles of fig. 7a, a
specific electrical equivalent circuit was used to fit the experimental data. This involved a
resistance R in parallel with a constant phase element ZCPE corresponding to the form Zꢂꢞꢟ
ꢇꢎ
=
ꢆ
JAω ; where A has the dimension of a capacity (Farad) only for n=1:
12

Page 13 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
ꢇꢎ
1
1
ꢆ
ꢒ = ꢒ + JAω
(4)
Z
R
Generally, the exponent n characterizes grain heterogeneity and sample porosity. In the case
of heterogeneous samples having a large distribution of grain sizes, the n values are lower
than 1. However in the case of ionic diffusion and reaction, this CPE model could be used to
represent the electrical signals in competition with the classical Warburg model mainly valid
to represent diffusion at the electrodes.58 The obtained results show a unique main semicircle
for both samples corresponding to grain core electrical behavior, as reported in the fig. S5
(supporting information). In the case of the nanoplatelets (Fig. S5a and S5d), a linear segment
appears at a lower frequency, suggesting a typical Warburgꢀlike behavior corresponding to
electrodes signals. It should be recalled that, as the electrodes are metallic (Pt), they allow
electron conduction, and behave as blocking electrodes for ionic conduction. In fact, the
observation of such a signal is generally attributed to gas interaction with electron exchange at
the electrodes. This is probably linked to the common reactions: O + e− ↔ O− and/or ½O2 +
−
2e− ↔ O2 . We also note that the diameter of the semicircles decreases according to the
temperature. The linear segment disappears between 150 and 350 °C. At low temperature, the
Warburg behavior is due to the adsorption and desorption of gases (oxygen and water
molecules). However, at high temperature, the Warburg behavior is mainly caused by the
conduction and the accumulation of electrons on the electrode surface.
Fig. 7: a) An example of Nyquist representation of the electrical impedance obtained for NPꢀ
WO3. b) Conductivity σ of the samples as a function of the reciprocal temperature.
The intersection of the semicircles and the horizontal axis Z’ corresponds to the R value. The
temperature dependent activation energies have been calculated from the Arrhenius plot of
data using the following equation:
ꢟ
σ = K exp ꢠ− ꢢ ꢡꢤꢥ
(5)
ꢣ
where Ea is activation energy for conduction, K a preꢀexponential factor and T the sample
temperature in Kelvin. In this expression, the conductivity σ is obtained from the classical
l
relation: σ = ꢒ , where l and S are respectively the thickness and surface of cylindrical
RS
13

RSC Advances
Page 14 of 30
View Article Online
DOI: 10.1039/C6RA13500E
pellet in contact with electrodes. This conductivity depends on effective porosity and
microstructure of the compacted pellet. It should be noted that this expression is only valid for
pure electron conduction, and is generally used as a good approximation in the case of mix
conduction. In fact, it was necessary to determine these activation energies in various
temperature ranges. These variations of activation energies could be associated with electrical
evolutions involved by lattice expansion or charges carriers modifications in solids. The
activation energies have been calculated taking into account the total resistance R.
Fig. 7b shows the evolution of the logarithm of conductivity as a function of the reciprocal
temperature. The conductivity curves can be divided into two parts depending on the
temperature. At low temperature, we observe a first increase of log(σ) as T increases up to
75°C: this can be associated with extrinsic transport of impurities including adsorbed gases.
Then we observe a strong decrease of conductivity due to gas desorption and water molecules
elimination up to 125°C for PS sample and up to 200°C for NP sample. The desorption
process is faster for an isotropic system PS as compared to anisotropic nanoplatelets. Finally,
the conductivity increases again, following a classical semiconductor behavior. Close to 300
°C, it appears a small change in slope, giving rise to a continuous electrical modification. This
particular trend can be due to a modification in charge carriers (the intrinsic carriers playing a
prominent role at high temperature) or to a structural transformation. A similar phenomenon
has been reported by other group for dc resistance measurement of pure WO3.59,60 They have
explained this change by a monoclinic → orthorhombic WO3 phase transition.61,62
The electrical transition could be associated with lattice expansion or charges carriers
modifications in solids. The activation energies have been calculated taking into account the
total resistance R. In Table S1 we have reported the activation energies at each step. At high
temperature (T> 500 K), the activation energies can be connected with the formation of
intrinsic defects (Schottky defects, oxygen vacancies) in the bulk. At low temperature T<400
K, extrinsic defects resulting from adsorbed molecules or impurities from initial constituents,
are dominant in transport properties. If the material is subjected to air atmosphere, the
exchange with oxygen plays the major role in the composition of the solid at high temperature
and its nonꢀstoichiometry.
4. Catalytic applications
4.1.CO oxidation
14

Page 15 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
Catalytic analyses were carried out by FTIR spectra acquisitions, every 30s. In presence of
active samples, the spectra are characterized by the appearance of a new absorption band
corresponding to the emission of CO2 resulting from CO conversion, following the wellꢀ
known conversion reaction CO + ½ O2→ CO2.
Fig. 8a and 8b represent the time dependent evolutions of the CO2 FTIR band intensity at
various temperatures, for NPꢀWO3 sample (data of PSꢀWO3 are in supplementary
information). The decrease in CO band intensities is clearly correlated with the increase in
CO2 band intensities. At 350°C, the conversion reaches a maximum level (the CO band
increases and the CO2 band reaches a maximum value) for NP. The solid–gas interactions are
based on three main physicochemical steps: gas adsorption process, surface reactions and gas
desorption process. This process was relative to reactions between adsorbed CO gases and
adsorbed oxygen (O2) from air, an interaction described by the Langmuir–Hinshelwood
mechanism.63
Fig. 8: Catalytic activity of WO3 nanoplatelets versus time. Evolution of the intensities of
FTIR absorption bands: a) CO oxidation into b) CO2.
The calculation of the surface area of the two NP and PS samples based on the XRD size
effects (D values) delivered the values of 9 m2/g for NP sample and 14 m2/g for PS sample.
According to these values, the (010) facets contribute about 70% of the surface area with
respect to other facets. From these data, the catalytic efficiencies can be expressed per surface
unit. As temperature increased, CO conversion occurred up to a maximum of oxidation
activity. Fig. 9 reports the catalytic efficiency, normalized to the specific surface of each
sample. These data clearly show that the conversion efficiency strongly increases for
nanoplates than pseudospheres and remain stable until 500°C. The catalytic performance of
the platelets is at least two times higher than the spheres particles.
Generally, the catalytic activity can be attributed to any increase in specific surface of
samples, which is not the case for our samples. The catalytic reactivity normalized to the
surface unit may be linked both to the morphology and the nature of exposed surfaces. As
reported by ab initio calculation, the plan platelets (020) exhibit more active oxygen sites than
other plan orientations, as shown by fig. S2b.64 However, orientation (200) and (002) present
15

RSC Advances
Page 16 of 30
View Article Online
DOI: 10.1039/C6RA13500E
a surface state conditioned by a mixture of sites of W and O atoms, thus less reactive.
Therefore, the pseudospherical particles are more isotropic with equal distribution of active
sites. As a consequence, they show less oxygen sites than nanoplates.
Fig. 9: Normalized catalytic efficiency to the specific surface area
So regarding the electrical behavior, the conductivity of the fabricated WO3 samples increased
with temperature, mainly due to high mobile electrons. For the same temperature range, the
catalytic conversion of CO into CO2 was improved. The conduction is due to the charge
carrier displacement in the oxide grains and occurs through the barriers at the generated
interfaces. The increased capacity to exchange electrons with external molecules (CO and O2)
of the anisotropic microstructure of thin platelets is at the origin of this improved conversion
of CO into CO2.
Analyzes of the used samples after catalytic applications, show no structural changes. In
addition to the general appearance of the powder, unchanged color, the samples were checked
by XRD. A slight decrease of the peak intensities was observed, however the Bragg position
remain identical (Fig. S1 of supporting information), thus demonstrating the stability of the
oxides. This stability is enhanced by the presence of air as an oxygen source for the
regeneration of the active sites.
4.2. Photodegradation of RhB
The photocatalytic activities of nanoplates and pseudospherical oxides were evaluated from
the degradation reaction of RhB in aqueous solution, under UV irradiation. Irradiation in the
absence of WO3 samples shows that the degradation of RhB is extremely lower, lower than
5% during the experiment time. Fig. 10 shows the absorption spectra of the RhB solution for
different times during UV lights irradiation in the presence of the catalyst. The characteristic
peak intensity of RhB at 554 nm decreases progressively with irradiation times and the
absorption peak almost disappears after 350 min irradiation for WO3 nanoplatelets, but still
persists for the WO3 pseudospheres at the same irradiation time. The shift of the main
absorbance band from 554 nm to 520 nm with irradiation time indicates the removal of ethyl
16

Page 17 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
groups one by one.65 Photocatalysis of RhB occurred under UV irradiation until complete
degradation without altering the morphologies and powders remain yellow.
Fig. 10: a) Absorption spectra of RhB solution with photocatalyst NP and PS of WO3 under
UV irradiation
As observed in fig. 11, the photocatalytic efficiency of NPꢀWO3 and PSꢀWO3 reaches 93 %
and 87 % respectively after irradiation for 350 min. The photodegradation rate constants k for
the two catalysts were estimated from firstꢀorder reaction kinetics by using the classical
relationship Cn/C0 = exp(ꢀkt). The reaction rate constant are k=0.0072 minꢀ1 for NPꢀWO3
particles and k = 0.0050 minꢀ1 for PSꢀWO3 particles (see insert Fig. 11). The higher
photocatalytic activity observed for NPꢀWO3 particles can be ascribed to the large active
surface areas (70% of Sa), with large quadrangular faces (010) absorbing more UV light,
small thicknesses allowing easy migration of electronꢀhole pairs to the surface, and more
active adsorption and photocatalytic reaction sites.66
Fig. 11: Normalized Photocatalytic activity (Cn/C0)*.The insert corresponds to the Y =ꢀ
Ln(Cn/C0) function of time t delivering the constant k : k=0.0072 min ꢀ1 for NPꢀWO3 ; k =
0.0050 min ꢀ1 for PSꢀWO3, correlation coefficient R=0.976
The absence of new absorption band during the reduction of RhB in the UVꢀVis spectra (fig
10) indicates that RhB is decomposed thoroughly into CO2 and H2O according to the
following mechanism.67 Under irradiation, photogenerated electrons would transfer from the
valence band to the conduction band of WO3, allowing the formation of highly active
electron–hole pairs according to the reaction WOꢦ + hν → hꢧ + eꢚ. The efficiency of this
step depends particularly on the structural and electronic properties of the semiconductor.
Furthermore, the high platelet gap (2.82 eV) allows a longer residence time of the charge
carriers in the conduction band, compared to spheres (2.65 eV). As a consequence, the
photogenerated h+ and eꢀ then contribute to the oxidation and reduction reaction. The hole is
quickly converted to the hydroxyl radical (OH*) upon oxidation of the adsorbed water. So,
electrons most often react with dissolved O2 with formation of superoxide radical. They can
17

RSC Advances
Page 18 of 30
View Article Online
DOI: 10.1039/C6RA13500E
further react and produce HO* radicals and thus contribute to oxidation of organic materials
giving rise to water and CO2 gas. The formed hydroxyl radical is the major reactant, which is
responsible for oxidation of RhB over WO3.68 The highest photodegradation of the
nanoplatelets is due to their exposed crystal surfaces. Han et al.64 demonstrated that shape
controlled WO3 nanocrystals with highly exposed (020) facets exhibit higher sensitivity
compared to nanoparticles with other prominent facets or anisotropic particles. Similar trend
was found in the case of TiO2 nanosheets and trapezohedron shaped In2O3 nanoparticles
which exhibit high index facets.69,70
The differences in the photodegradation capability were attributed to the presence of
unsaturated cations on the (020) surfaces, i.e. five and fourꢀfold coordinated W ions, which
could effectively chemisorb both oxygen species (e.g. Oꢚꢈ , Oꢈꢚ,Oꢚ) and the target molecules.
Liu et al.71 indicated that the surface energy of crystal faces in shape controlled WO3
nanocrystals is in the following order of 1.56 Jmꢀ2 for (002) > 1.54 Jmꢀ2 for (020) > 1.43 Jmꢀ2
for (200) and suggested that (002) and (020) are the most active in surface mediated reactions.
A semiconductor with different crystal structures not only has different band gap but also has
defects (as revealed for nanoplatelets) that can be affect the electron–hole recombination rate.
Then, the reactivity of the photocatalyst is affected by both the morphology and the crystal
structures.72 In addition to the state surface, the nanostructures with corners and edges as traps
(like nanoplatelets) promote more adsorption than nanostructures without these traps (like
nanospheres). Due to their low coordination numbers, corners and edges in the case of
platelets or cubic shapes, have more catalytic activity in comparison with other flat surfaces,
as reported by M. Farhadian et al.73 The NPꢀWO3 (with their corners and edges) in
comparison to the PSꢀWO3 (with high surface area), have the most photocatalytic activity.
The photodegradation process is favored by the large surface area of the platelets, particularly
the (010) facets (70% of the surface area), showing a high amount of catalytic reaction sites
for the adsorption of reactant molecules and to more surface OHꢀ, therefore allows a high
degradation rate of organic pollutant.
As reported in this last part of the CO oxidation, the oxides remain stable without color and
structure alteration after rhodamine photodegradation. The XRD patterns occurs in a similar
manner before applications.
4. Conclusions
18

Page 19 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
We have successfully synthesized well crystallized single phase WO3 platelets morphology,
based on free template aqueous mineralization processes. The oxides were identified in the
orthorhombic structure, close to monoclinic phase, with a large exposed (020) facet than (002)
and (200). The microstructural analysis revealed the main orientation along the [020] plan
with the presence of the linear defect parallel to (020). In addition, the distortion and
deformation of the octahedra are critically dependent upon the W–O and O–O bonds. The
splitting bonds are at the origin of the studied properties changes.
The photocatalytic performances of the asꢀprepared WO3 are investigated for the RhB
degradation, which the photodegradation activity of nanoplatelets was significantly higher
than WO3 with a spherical morphology. However, the CO transformation over tungsten
material remains incomplete. According to the semiconductor behavior of the oxides, the CO
oxidation based on the oxydoꢀreduction mechanism shows a favorable charge transfer for CO
catalytic detection. The nanoplatelets morphology, with corners and edges, have the best
catalytic activity in comparison with nanospheres. The large exposed (020) facets of the
nanoplatelets show a high amount of catalytic reaction sites, for the conversion of reactant
molecules and for photodegradation of organic pollutant with more surface OHꢀ.
Acknowledgements
The authors thank Pr. J.Ch. Valmalette for his assistance with the Raman measurements. We
gratefully acknowledge the Regional Council of ProvenceꢀAlpesꢀCôte d’Azur, General
Council of Var, and the agglomeration community of Toulon Provence Mediterranean for
their financial supports in the framework of "M2D2" project.
References:
1.
2.
D. Chen, L. Gao, A. Yasumori, K. Kuroda and Y. Sugahara, Small, 2008, 4, 1813.
D. Chen, H. Wang, R. Zhang, L. Gao, Y. Sugahara and A. Yasumori, J. Ceram. Process. Res,
2008, 9, 596.
3.
B. I. Kharisov, Recent Pat Nanotechnol, 2008, 2, 190.
4.
5.
6.
7.
X. Shen, G. Wang and D. Wexler, Sens. Actuator B-Chem., 2009, 143, 325.
X. Li, G. Zhang, F. Cheng, B. Guo and J. Chen, J. Electrochem. Soc., 2006, 153, H133.
J. Su, X. Feng, J. D. Sloppy, L. Guo and C. A. Grimes, Nano Lett., 2010, 11, 203.
D. Chen and J. Ye, Adv. Funct. Mater., 2008, 18, 1922.
8.
9.
10.
11.
12.
R. Chatten, A. V. Chadwick, A. Rougier and P. J. Lindan, J. Phys. Chem. B, 2005, 109, 3146.
S. R. Bathe and P. Patil, Smart Mater. Struct., 2008, 18, 025004.
M. Gondal, A. Hameed and Z. Yamani, J. Mol. Catal. A: Chem., 2004, 222, 259.
H. Liu, S. Huang, L. Zhang, S. Liu, W. Xin and L. Xu, Catal. Commun., 2009, 10, 544.
J. Ou, M. Yaacob, J. Campbell, M. Breedon, K. Kalantar-Zadeh and W. Wlodarski, Sens.
Actuator B-Chem., 2012, 166, 1.
13.
J.d. R. M. Neto, R.M. Torresi and S. I. C. de Torresi, J. Electroanal. Chem., 2016, 765, 11.
19

RSC Advances
Page 20 of 30
View Article Online
DOI: 10.1039/C6RA13500E
14.
15.
16.
W. Zhao, Z. Wang, X. Shen, J. Li, C. Xu and Z. Gan, Int. J. Hydrogen Energy, 2012, 37, 908.
H. Zheng, Y. Tachibana and K. Kalantar-zadeh, Langmuir, 2010, 26, 19148.
M. Ahmadi, S. Sahoo, R. Younesi, A. P. Gaur, R. S. Katiyar and M. J. Guinel, J. Mater. Sci.,
2014, 49, 5899.
17.
18.
19.
C. Chacón, M. Rodríguez-Pérez, G. Oskam and G. Rodríguez-Gattorno, J. Mater. Sci. Mater.
Electron., 2015, 26, 5526.
M. Szabó, P. Pusztai, A.-R. Leino, K. Kordás, Z. Kónya and Á. Kukovecz, J. Mol. Struct., 2013,
1044, 99.
K. P. Musselman, G. J. Mulholland, A. P. Robinson, L. Schmidt-Mende and J. L. MacManus-
Driscoll, Adv. Mater., 2008, 20, 4470.
20.
21.
S. Gubbala, J. Thangala and M. Sunkara, Sol. Energy Mater. Sol. Cells, 2007, 91, 813.
M. Deepa, A. Srivastava, S. Lauterbach, S. Shivaprasad and K. Sood, Acta Mater., 2007, 55,
6095.
22.
23.
24.
F. Zheng, M. Zhang and M. Guo, Thin Solid Films, 2013, 534, 45.
Q.-H. Li, L.-M. Wang, D.-Q. Chu, X.-Z. Yang and Z.-Y. Zhang, Ceram. Int., 2014, 40, 4969.
Z. Gu, T. Zhai, B. Gao, X. Sheng, Y. Wang, H. Fu, Y. Ma and J. Yao, J. Phys. Chem. B, 2006, 110,
23829.
25.
S. Q Guo, M. M Zhen, M. Q. Sun, X. Zhang, Y. P. Zhao, L. Liua, RSC. Adv. , 2015, 5, 16376.
26.
27.
M. Gimeno-Fabra, P. Dunne, D. Grant, P. Gooden and E. Lester, Chem. Eng. J., 2013, 226, 22.
G. Zhang, W. Guan, H. Shen, X. Zhang, W. Fan, C. Lu, H. Bai, L. Xiao, W. Gu and W. Shi,
Eng. Chem. Res., 2014, 53, 5443.
Ind.
28.
29.
A.Gurio, Nanoscale, 2011, 3, 154.
C. Gómez-Solís, D. Sánchez-Martínez, I. Juárez-Ramírez, A. Martínez-de la Cruz and L. M.
Torres-Martínez, J. Photochem. Photobiol, 2013, 262, 28.
30.
31.
R. A. Young, The Rietveld Method. I. U. Cr, Oxford University Press Inc New York.
F. W. B. Lopes, M. Arab, H. P. Macedo, C. P. de Souza, J. F. de Souza and J. R. Gavarri, Powder
technol, 2012, 219, 186.
32.
L. V. Azároff, Elements of X-Ray Crystallography; McGraw-Hill Book Company: New York, NY,
USA.
33.
34.
35.
R. Jenkins, R. Snyder, Introduction to X-Ray Powder Diffractometry, New York, NY, USA, 1996.
R.C. Pullar, M.D. Taylor, A. K. Bhattacharya, J. Eur. Ceram. Soc., 1998, 18, 1759.
A. K. Tripathi, M. K. Singh, M. C. Mathpal, S. K. Mishra and A. Agarwal, J. Alloys Compd., 2013,
549, 114.
36.
M. David, M. Arab, F. Guinneton, E. Flahaut, B. Bakiz, P. Nowakowski and J. R. Gavarri, Chem.
Eng. J., 2011, 171, 272.
37.
38.
39.
40.
E. Salje, Acta Cryst. B, 1977, 33, 574.
B. Loopstra and H. Rietveld, Acta Cryst. B, 1969, 25, 1420.
T. Vogt, P. M. Woodward and B. A. Hunter, J. Solid State Chem., 1999, 144, 209.
J. Pfeifer, C. Guifang, P. Tekula-Buxbaum, B. Kiss, M. Farkas-Jahnke and K. Vadasdi, J. Solid
State Chem., 1995, 119, 90.
41.
42.
43.
44.
45.
I. Hargittai, M. Hargittai, V. Spiridonov and E. Erokhin, J. Mol. Struct., 1971, 8, 31.
U. O. Krašovec, A. Š. Vuk and B. Orel, Electrochim. Acta, 2001, 46, 1921.
G. De Wijs and R. De Groot, Electrochim. Acta, 2001, 46, 1989.
A. Rougier, F. Portemer, A. Quede and M. El Marssi, Appl. Surf. Sci., 1999, 153, 1.
M. Regragui, M. Addou, A. Outzourhit, J. Bernede, E. El Idrissi, E. Benseddik and A.
Kachouane, Thin Solid Films, 2000, 358, 40.
46.
M. Daniel, B. Desbat, J. Lassegues, B. Gerand and M. Figlarz, J. Solid State Chem., 1987, 67,
235.
47.
48.
49.
B. M. Weckhuysen, J.-M. Jehng and I. E. Wachs, J. Phys. Chem. B, 2000, 104, 7382.
T. Kubo and Y. Nishikitani, J. Electrochem. Soc., 1998, 145, 1729.
P. Tägtström and U. Jansson, Thin Solid Films, 1999, 352, 107.
20

Page 21 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
50.
51.
E. Salje, Acta Cryst. A, 1975, 31, 360.
A. Durairajan, D. Balaji, K. K. Rasu, S. M. Babu, Y. Hayakawa and M. Valente, J. Lumin., 2015,
157, 357.
52.
53.
54.
A. P. Alivisatos, 1996, 271,933.
X. Yang, Z. Fu, G. Liu, C. Zhang, Y. Wei, Z. Wu and T. Sheng, RSC Adv., 2015, 5, 70220-70228.
I. L. Rosa, A. P. A. Marques, M. T. Tanaka, D. M. Melo, E. R. Leite, E. Longo and J. A. Varela, J
fluoresc, 2008, 18, 239.
55.
56.
57.
R. Lacomba-Perales, J. Ruiz-Fuertes, D. Errandonea, D. Martínez-García and A. Segura,
Europhys. Lett., 2008, 83, 37002.
V. Longo, E. Orhan, L. Cavalcante, S. Porto, J. Espinosa, J. A. Varela and E. Longo, Chem. Phys.,
2007, 334, 180.
M. Boulova, N. Rosman, P. Bouvier and G. Lucazeau, J. Phys.: Condens. Matter, 2002, 14,
5849.
58.
59.
60.
Z. Li, R. Liu, J. Wang, Z. Xu, Y. Xie and B. Wang, Sci. Tech. Adv. Mater., 2007, 8, 566.
G. Kiss, Z. Pintér, I. Perczel, Z. Sassi and F. Réti, Thin Solid Films, 2001, 391, 216.
A. Labidi, C. Jacolin, M. Bendahan, A. Abdelghani, J. Guerin, K. Aguir and M. Maaref, Sens.
Actuator B-Chem., 2005, 106, 713.
61.
L. Depero, S. Groppelli, I. Natali-Sora, L. Sangaletti, G. Sberveglieri and E. Tondello, J. Solid
State Chem., 1996, 121, 379.
62.
63.
64.
65.
66.
67.
68.
P. Woodward, A. Sleight and T. Vogt, J. Phys. Chem. Solids, 1995, 56, 1305.
H. Over and M. Muhler, Prog. Surf. Sci., 2003, 72, 3.
X. Han, X. Han, L. Li and C. Wang, New J. Chem, 2012, 36, 2205.
J. Huang, L. Xiao and X. Yang, Mater. Res. Bull., 2013, 48, 2782.
J. Yu and L. Qi, J. Hazard. Mater., 2009, 169, 221.
H. Zhang, J. Yang, D. Li, W. Guo, Q. Qin, L. Zhu and W. Zheng, Appl. Surf. Sci., 2014, 305, 274.
R. Gao, J. Stark, D. W. Bahnemann and J. Rabani, J Photochem Photobiol A Chem, 2002, 148,
387.
69.
70.
X. Han, Q. Kuang, M. Jin, Z. Xie and L. Zheng, J. Am. Chem. Soc., 2009, 131, 3152.
X. Han, X. Han, L. Sun, S. Gao, L. Li, Q. Kuang, Z. Xie and C. Wang, Chem.Commun., 2015, 51,
9612.
71.
72.
Y. P. Xie, G. Liu, L. Yin and H.-M. Cheng, J. Mater. Chem., 2012, 22, 6746.
Y. Zheng, C. Chen, Y. Zhan, X. Lin, Q. Zheng, K. Wei and J. Zhu, J. Phys. Chem C, 2008, 112,
10773-10777.
73.
M. Farhadian, P. Sangpout and G. Hosseinzadeh, Journal of Energy Chemistry, 2015, 24,
171-177.
21

RSC Advances
Page 22 of 30
View Article Online
DOI: 10.1039/C6RA13500E
Fig. 1: SEM images of the agglomerate WO3 powders: a) NPꢀWO3 and b) PSꢀWO3. The TEM
pictures in insert show the dispersed grains
I
I
obs
cal
I
(cal-obs)
Nanoplates
Bragg position
)
s
t
i
n
u
.
b
r
a
(
y
t
i
s
n
e
t
Pseudosphers
n
I
20
25
30
35
40
45
50
55
60
65
70
2
θ (degrees)
Fig. 2: Rietveld analyses (space group Pnma n°62): observed and calculated XRD profiles of
the structures: NPꢀWO3 and PSꢀWO3.

Page 23 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
Fig. 3: TEM image and electronic diffraction patterns of flat laying WO3 platelets: a) and c)
correspond respectively to perpendicular [010] (a) and [001] (c) zone axes of the
orthorhombic βꢀWO3 structure.
Fig. 4: HRTEM image and electron diffraction pattern of WO3 platelets laying edge on a) and
c) correspond to the [100] zone axis of the orthorhombic structure, c) is an HRTEM image
showing the piling up of (020) planes. The Inverse Fast Fourier Transform d) emphasizes the
defects in the piling up of the plans, pointed by arrows

RSC Advances
Page 24 of 30
View Article Online
DOI: 10.1039/C6RA13500E
Stretching modes
Nanoplates
Pseudospheres
8
0
8
Lattice modes
2
8
1
7
Deformation modes
7
3
7
2
)
4
6
u
.
a
(
5
3
1
y
2
9
t
i
s
n
0
3
3
e
t
n
I
50
75
100
125
200
400
600
800
1000
Raman shift (cm-1)
Fig. 5: Raman spectra of WO3 structures
4
3
2
1
0
Nanospheres
Nanosheets
Linear fit sheets
Linear fit spheres
)
R
(
Eg = 2,82 eV
F
Eg = 2,65 eV
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Photo energy (eV)
Fig. 6: Variation of F(R∞) as a function of irradiation energy, for the PS and NP samples, R∞
being determined from UVꢀVis absorbance spectra.

Page 25 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
5M
4M
3M
2M
1M
0
b)
a)
-4
Nanoplates
a4
Pseudospheres
25°C
75°C
-5
-6
-7
-8
)
m
2
c
/
a1
S
(
a2
200°C
σ
σ
σ
σ
50°C
g
o
l
a3
75°C
2M
125°C
Surface conduction effect
3.0 3.5
0
4M
6M
8M
10M
1.5
2.0
2.5
103/T (K)
Real part impedance Z' (Ohm)
Fig. 7: a) An example of Nyquist representation of the electrical impedance obtained for NPꢀ
WO3. b) Conductivity σ of the samples as a function of the reciprocal temperature.
0,7
CO
b)
200 °C
250 °C
300 °C
350 °C
400 °C
450 °C
350°C
450°C
200 °C
250 °C
300 °C
350 °C
400 °C
450 °C
0,25
0,20
0,15
0,10
0,05
0,00
0,6
0,5
0,4
0,3
0,2
0,1
0,0
200°C
0
20
40
60
80
100
0
20
40
60
80
100
Time (min)
Time (min)
Fig. 8: Catalytic activity of WO3 nanoplatelets versus time. Evolution of the intensities of
FTIR absorption bands: a) CO oxidation into b) CO2.

RSC Advances
Page 26 of 30
View Article Online
DOI: 10.1039/C6RA13500E
0,12
0,10
0,08
0,06
0,04
0,02
0,00
1,2
1,0
0,8
0,6
0,4
0,2
0,0
CO _ Nanoplates
CO _ pseudospheres
CO2 _Nanoplates
CO2 _pseudospheres
0
100
200
300
400
500
Temperature (°C)
Fig. 9: Normalized catalytic efficiency to the specific surface area
1,2
1,2
1,0
0,8
0,6
0,4
0,2
0,0
0 min
0 min
1,0
10 min
40 min
10 min
40 min
0,8
0,6
0,4
0,2
0,0
100 min
160 min
100 min
160 min
230 min
230 min
330 min
350 min
330 min
350 min
400
450
500
550
600
650
700
400
450
500
550
Wavelength (nm)
600
650
700
Wavelength (nm)

Page 27 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
Fig. 10: a) Absorption spectra of RhB solution with photocatalyst NP and PS of WO3 under
UV irradiation
3,0
0,14
WO3 pseudospheres
2,5
0,12
-1
k
= 0,00724 min
nanoplates
2,0
1,5
1,0
0,5
0,0
WO3 nanoplates
0,10
0,08
0,06
0,04
0,02
0,00
-1
= 0,005 min
k
pseudospheres
0
50
100
150
200
250
300
350
400
Times (min)
0
50
100
150
200
250
300
350
400
Times (min)
Fig. 11: Normalized Photocatalytic activity (Cn/C0)*. The insert corresponds to the Y =ꢀ
Ln(Cn/C0) function of time t delivering the constant k: k=0.0072 min ꢀ1 for NPꢀWO3 ; k =
0.0050 min ꢀ1 for PSꢀWO3.

RSC Advances
Page 28 of 30
View Article Online
DOI: 10.1039/C6RA13500E
Table 1: Refinement structural parameters and Coordinates atoms of WO3 phase based on
two octahedra structure (Space group of Pnma). Refinement calculation based on CIF
datasheet of ICSD (Collection code 836).
x
y
z
x’
y’
z’
Coordinates
atoms
Nanoplatelets
Pseudospheres
0,0276(1) 0,2500(0) 0,0260(5)
0,0289(0) 0,2500(0) 0,5375(0)
W1
W2
O1
O2
O3
O4
O5
O6
O7
O8
0,0276(0) 0,2500(0) 0,0285(4)
0,0259(5) 0,2500(0) 0,5355(6)
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,50000
0,50000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,00000
0,50000
0,50000
0,2573(3) 0.2500(0) 0,0633(1) 0,2598(3) 0.2500(0) 0,0565(1)
0,2517(3) 0.2500(0) 0,5025(9) 0,2646(4) 0.2500(0) 0,5055(5)
0,0324(8) 0.2500(0) 0,2458(7) 0,0522(5) 0.2500(0) 0,2460(8)
0,004759 0.2500(0) 0,7792(1) 0,0003(0) 0.2500(0) 0,7719(7)
a = 7,5347(5) A ; b = 7,3131(3) A ;
c = 7,6941(9) A
a = 7,5239(0) A ; b = 7,3222(7)A ;
c = 7,6892(2) A
Structural
parameters
α = γ = β = 90.0°
α = γ = β = 90.0°
Rp= 9.589 %; Rwp= 12.668 %;
Rp= 7.962 %; Rwp= 10.297 %;
Reliability
factors
R
exp= 6.122%.
R
exp= 6.728 %
Goodness of fit χ2 = 4.280
Goodness of fit χ2 = 2.169
.

Page 29 of 30
RSC Advances
View Article Online
DOI: 10.1039/C6RA13500E
Table 2: Selected bond distances and angles in different nanostructure of WO3, with two
optimized WO6 octahedral of the unit cell.
WO6 octahedra
Nanoplates
Spheres
groups
Oct1
Oct2
Oct1
Oct2
Axes
1,692
1,754
1,851
1,851
1,907
2,090
78.8
1,700
1,864
1,864
1,869
2,188
2,244
70.0
1,683
1,760
1,855
1,855
1,983
1,996
76.5(2)
1,811
1,828
1,861
1,861
2,124
2,235
76.0
WꢀO
a
b
c
Bond (Å)
Error: 0.072Å
79.4
80.2
82.6(2)
77.6
82.9(2) 80.6(2) 82.7(2) 81.5(2)
83.2(2) 81.2(2) 97.2(2) 82.1(2)
96.4(2) 95.2(2) 97.5(2) 94.9(2)
97.3(3) 98.0(2) 102.9 97.8(2)
104.6 104.9(2) 103.9 103.4(2)
OꢀWꢀO
Angle (º)
162.
150.1
157.6
175.1
2,631
2,631
2,634
2,634
2,688
2,688
2,821
2,821
161.3
179.6
179.5
2,535
2,535
2,547
2,547
2,662
2,662
2,713
2,713
153.2
159.2
179.0
2,609
2,609
2,705
2,705
2,705
2,705
2,779
2,779
176.1
176.6
2,496
2,496
2,616
2,616
2,642
2,642
2,709
2,709
OꢀO Bond (Å)
Error = 0.055 Å
Axes
OꢀO corner
distance (Å)
3,840
3,660
3,600
a
b
c
3.760
3.660
3.670
3.81
3.71
3.79
3.79
3.70
3.94
a
b
c
WꢀW
distance (Å)
(*): multiplicity; a, b and c indicate the three direction.