Effect of Temperature and pH on Phase Transformations in Citric Acid Mediated Hydrothermal Growth of Tungsten Oxide

Article abstract of DOI:10.1002/ejic.201701156

The temperature-dependent composition of suspension during citric acid mediated crystallization of tungsten trioxide (WO₃) from sodium tungstate was studied by in situ Raman spectroscopy. Additionally, microwave-assisted hydrothermal synthesis experiments combined with ex situ analysis by X-ray diffraction and SEM were performed to analyze the effect of pH on the eventually, isothermally, obtained crystal phase and morphology. The Raman results suggest that WO₃·2H₂O precipitates from the tungstate solution upon acidification to pH 0.5 at room temperature. This is first transformed to WO₃·H₂O initiating at T = 70 °C. At temperatures above 170 °C, the crystallization of phase-pure monoclinic WO₃ with well-defined plate-like morphology was observed at pH 0.5. Using the microwave-assisted hydrothermal synthesis procedure shows that increasing the pH to values of 1.5 or 2 results in significant or dominant formation of hexagonal WO₃, respectively. Comparing the activity of selected samples in photocatalytic oxidation of propane using visible light, demonstrates the presence of hydrate phases or hexagonal WO₃ is detrimental to performance.

Full text of DOI:10.1002/ejic.201701156

A Journal of
Accepted Article
Title: Effect of temperature and pH on phase transformations in citric
acid-mediated hydrothermal growth of tungsten oxide
Authors: Kasper Wenderich, Johannes Noack, Anne Kärgel, Annette
Trunschke, and Guido Mul
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Inorg. Chem. 10.1002/ejic.201701156
Link to VoR: http://dx.doi.org/10.1002/ejic.201701156

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
Effect of temperature and pH on phase transformations in
citric acid-mediated hydrothermal growth of tungsten oxide
Kasper Wenderich,*[a, b] Johannes Noack,[c, d] Anne Kärgel, Annette Trunschke and Guido Mul
[c]
[c]
[a]
2]
Abstract: The temperature dependent composition of suspension
during citric acid-mediated crystallization of tungsten trioxide (WO
decomposition when part of a Z-scheme configuration.[ Methods
include sol-gel methods,[ solvothermal
3]
3
)
to synthesize WO
3
[5]
from sodium tungstate was studied by in situ Raman spectroscopy.
Additionally, microwave-assisted hydrothermal synthesis experiments
combined with ex situ analysis by X-Ray diffraction and Scanning
Electron Microscopy were performed to analyze the effect of pH on
the eventually, isothermally, obtained crystal phase and morphology.
synthesis,[4] electrochemical etching, spray pyrolysis,[6] chemical
vapor deposition,[ and template directed synthesis. The
7]
[8]
hydrothermal synthesis route gives access to WO
3
9]
particles with
[
[10]
a well-defined morphology, such as nanorods, nanowires,
[11]
nanocubes,[11c] octahedra,[12] nano-
nanoplates/nanosheets,
10b,
13]
and flower-like morphologies.[14] Often capping
The Raman results suggest that WO
tungstate solution upon acidification to pH 0.5 at room temperature.
This is first transformed to WO · H O initiating at T = 70 °C. At
temperatures above 170 °C, the crystallization of phase-pure
monoclinic WO with well-defined plate-like morphology was
3
2
· 2H O precipitates from the
urchins[
agents are used to control the morphology and crystal phase of
the samples. These capping agents adsorb on specific surface
facets of a crystal during hydrothermal growth, and thus influence
the final morphology.[ Although citric acid-mediated growth of
3
2
15]
3
observed at pH 0.5. Using the microwave-assisted hydrothermal
synthesis procedure shows that increasing the pH to values of 1.5 or
plate-like WO
3
in microwave-assisted hydrothermal synthesis has
been reported previously by Sungpanich et al.,[ a detailed
analysis of the effect of process conditions during synthesis (pH
and temperature) on the growth mechanism and obtained crystal
phase and morphology, has not yet been reported.
16]
2
3
results in significant or dominant formation of hexagonal WO ,
respectively. Comparing the activity of selected samples in
photocatalytic oxidation of propane using visible light, demonstrates
the presence of hydrate phases or hexagonal WO
performance.
3
is detrimental to
In this work, the phase transitions of tungstates initiated by
acidification and increasing temperature up to 170 °C, were
investigated by in situ Raman spectroscopy.[
17]
Additionally,
microwave-assisted hydrothermal synthesis experiments
combined with ex situ analysis by X-Ray diffraction and Scanning
Electron Microscopy were performed to in particular analyze the
effect of pH on morphology and phase composition of the final
product after extended isothermal crystallization time, reduced
from 17 hours in the Raman-autoclave to 2 hours in the
Introduction
One of the most studied semiconductor materials in various
applications (optics, sensors, catalysts), is tungsten trioxide
(
WO
small bandgap (ranging from 2.4 to 2.8 eV). Due to the latter
property, WO is able to absorb visible light in the blue range of
3 3
). WO is chemically stable, non-toxic and has a relatively
microwave methodology. Finally, a selected number of WO
3
[1]
samples obtained by microwave-assisted hydrothermal synthesis
were tested for their photocatalytic activity through the oxidation
3
the visible spectrum, and able to induce photocatalytic
transformations, such as in waste water treatment, and in water
of propane,[18] to determine the WO
performance.
3
phase providing optimized
[
a]
b]
Dr. K. Wenderich, Prof. Dr. G. Mul
Results and Discussion
Photocatalytic Synthesis Group, Faculty of Science and Technology
MESA+ Institute for Nanotechnology, University of Twente
PO Box 217, 7500 AE Enschede, The Netherlands
Dr. K. Wenderich
In situ Raman spectroscopy of crystal phases during
hydrothermal synthesis of WO
[
3
Centre for Innovation Competence SiLi-nano®
Martin-Luther-University Halle-Wittenberg
Karl-Freiherr-von-Fritsch-Str. 3, 06120 Halle (Saale), Germany
E-mail: kasper.wenderich@chemie.uni-halle.de
URL: http://www.sili-nano.de/de/detail/kasper-wenderich.html
Dr. J. Noack, A. Kärgel, Dr. A. Trunschke
Department of Inorganic Chemistry
Fritz-Haber-Institute der Max-Planck-Gesellschaft e.V.
Faradayweg 4-6, Berlin, Germany
Dr. J. Noack
Figure 1 shows the Raman spectra in the typical range of ν(W=O),
ν(O-W-O) and δ(O-W-O) vibrations measured during the heating
stage of the hydrothermal reaction of sodium tungstate at a pH of
[
c]
d]
0
1
.5 in the presence of citric acid, and up to a temperature of
70 °C. At room temperature, the acidification of the tungstate
solution leads to the precipitation of tungstic acid. This formation
can be observed by the Raman bands in the spectra.[19] In its
[
BasCat – UniCat BASF JointLab
Technische Universität Berlin
Sekr. EW K-01 Hardenbergstraße 36, 10623 Berlin, Germany
monohydrate form (WO
layers of octahedrally coordinated WO
vertices in the plane. The dihydrate form (WO
structurally quite similar with additional water molecules
3
· H
2
O), tungstic acid is composed of
(H O) units connected by
· 2H O) is
5
2
4
3
2
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
intercalated between the layers. Both phases can be identified in
the Raman spectra by their characteristic ν(W=O) bands at 945
3 2
Figure S3 reveals that in Figure 1 no sign of WO · H O is
observed at pH = 0.5 at low temperatures, while its existence is
observed in Figure S3 in addition to the dihydrate, by the band at
-1
and 958 cm , respectively. Additionally, bands from 632 to 650
cm-1 and at 884 cm-1 are found, which correspond to ν(O-W-O)
vibrations. While at low temperatures up to approximately 70 °C
the dihydrate is the major phase, during heating of the solution,
its bands gradually decrease in intensity for the benefit of the
-1
945 cm . This is likely due to the higher pH employed during the
experiment shown in Figure S3. The SEM images of the two as-
discussed experiments with 170 and 150 °C reaction
temperatures (Figures S2 and S4) show plate-like particles
exhibiting sharp edges, reflecting the monoclinic crystal structure
monohydrate. Starting from 150 °C, WO
3
1
· H
2
O is consumed and
bands at 806, 714, 323 and 269 cm- appear, indicating the
formation of monoclinic WO From those measurements,
WO · 0.33H O phase (which would exhibit intense bands at 945,
05, 680 cm ) as a possible intermediate in the transformation to
monoclinic WO
is not clearly identified.[19c] Also, Raman bands
3
of WO . This overall anisotropy in the particle dimensions is also
3
.
observed in the XRD patterns by an under-estimation of the
reflections involving the crystallographic a-axis and has thus been
considered in the Rietveld fit-model.
3
2
-1
8
3
from the citric acid acting as structure directing agent added in the
synthesis are not observed (Figure S7). The XRD pattern of the
precipitate after washing and drying (Figure S2) exclusively
Besides analysis of the final products, the transients observed in
the in situ Raman experiments were complemented by XRD and
SEM investigations of samples drawn from the hydrothermal
vessel at different points during the hydrothermal reaction at
3
shows peaks corresponding to monoclinic WO .
1
50 °C. The XRD patterns in Figure 2a show the formation of the
monohydrate phase precipitating from the acidified solution of
Na WO at temperatures above 100 °C and its decomposition to
monoclinic WO starting at 150 °C. This phase transformation is
best seen by decreasing intensity of the reflections at 2θ = 16.5,
2
4
3
25.6 and 52.7° and appearance of a trifold of reflections at 2θ =
2
3.1, 23.6 and 24.4°. As seen e.g. by the reflections at 2θ = 14.1°
°
and 18.0 appearing after 1 hour of hydrothermal synthesis, little
3 2
amounts of WO · 0.33H O are formed in addition, which slowly
decompose during the reaction at 150 °C. In the Raman spectra,
these traces are not observed due to a strong overlap of the
various bands. The SEM images (Figure 2b) of the samples
drawn during heating at 130 °C and after 1, 7 and 17 h of
isothermal reaction at 150 °C, reflect the transformation of the
irregularly shaped hydrate phase into the nicely defined
3
rectangular shape of the final monoclinic WO . Defects in the
crystal morphology are reduced and the mean particle size gets
more uniform as a result of Ostwald ripening with increasing
reaction time at 150 °C.
Figure 1. Raman spectra measured in situ during hydrothermal synthesis during
the heating of an aqueous sodium tungstate solution (0.05 M) at pH = 0.5 in the
presence of citric acid. The band at 746 cm-1 marked with * is assigned to the
sapphire window of the Raman probe. The reaction temperature profile is given
in Figure S1.
3
To further investigate the formation of monoclinic WO ,
experiments were conducted at 150 °C, which is the initial
transformation temperature from the monohydrate to monoclinic
WO . Also the pH was increased to a value of 1. A similar behavior
3
is observed in this experiment compared to the previous one as
seen from the in situ Raman spectra (Figure S3). Reaching
1
50 °C, the band at 806 cm-1 is increasing in intensity as a
function of time, while the characteristic bands of the
-1
monohydrate phase at 945 and 639 cm vanish. From 1 to 15
hours of reaction time, the absolute and relative intensities of the
bands barely change, indicating complete phase transformation
already occurred one hour after the isothermal point was reached.
Again, the XRD pattern of the precipitate after washing and drying
shows exclusively peaks corresponding to monoclinic WO
Figure S4). Comparison of the Raman spectra in Figure 1, and
3
(
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
observed in small quantities at temperatures of 230 °C and above.
At a pH of 2, a significant amount of precipitate is only formed at
T = 150 °C and T = 170 °C, which contains the hexagonal crystal
phase. Above these temperatures, insufficient precipitate is
formed to analyze. XRD patterns of samples synthesized at a pH
o
of 1.5 and 2.0 at 170 C are depicted in Figure S5. At a pH of 1,
pure monoclinic WO
a significant amount of WO
amount of WO · 0.33H O. At T = 170 °C, WO
but traces of WO · 0.33H O are still left. At a pH of 0.5 formation
of WO · 0.33H O is not observed anymore, but when the
temperature is too low (T = 150 °C), WO · H O is still observed.
3
is formed at T ≥ 200 °C. When T = 150 °C,
· H O is present, as well as a small
· H O is absent,
3
2
3
2
3
2
3
2
3
2
3
2
Figure 2a demonstrates that sufficient time at the isothermal stage
o
at 150 C will result in full transition of WO
3
· H
2
O to monoclinic
3
WO .
Table 1. Formed crystal phases during microwave-assisted hydrothermal
synthesis at different pH and different T, as measured through X-ray
diffraction. m-WO
hexagonal WO . N.A. means not available, meaning the amount of formed
precipitate was so low that this could not be analyzed.
3 3 3
is an abbreviation for monoclinic WO , h-WO for
3
T
150 o
C
170 o
C
200 °C
230 °C
250 °C
pH
0.5
60%
m-WO
3
m-WO
3
m-WO
3
m-WO
3
m-WO
0%
WO
3
4
3
· H
2
O
O
1
62%
94%
m-WO
6%
m-WO
3
m-WO
3
m-WO
3
m-WO
3
· H
·
3
3
WO
2%
3
2
3
WO ·
6%
0.33H
2
O
WO
3
0.33H
2
O
1.5
h-WO
WO
.33H
3
h-WO
WO
0.33H
3
h-WO
WO
0.33H
3
h-WO
WO
0.33H
3
h-WO
WO
0.33H
3
Figure 2. (a) XRD patterns of samples drawn from the reactor at different times
during the synthesis of WO at 150 °C and a pH of 1.0. Samples were drawn
3
3
·
3
·
3
·
3
·
3
·
2
O
0
2
O
2
O
2
O
2
O
during the heating process and after reaching the desired synthesis temperature
o
m-WO
3
m-WO
3
of 150 C (defined as t = 0 h). (b) SEM images of samples (after withdrawal)
after reaching 130 °C and hydrothermal reaction at 150 °C for 1, 7 and 17 h.
2
h-WO
3
h-WO
3
N.A.
N.A.
N.A.
Microwave-assisted hydrothermal synthesis experiments.
The structure of the individual phases is also reflected in the
overall particle morphology (see Figure S6). In order to compare
The Raman spectroscopy measurements provided in situ
information on the effect of temperature on crystal phase
transitions during hydrothermal synthesis, but required significant
analysis time. To reduce the amount of required solution and to
enhance the isothermicity of the reactor, reactions were
performed in a microwave-assisted hydrothermal reactor, to study
the effect of pH at different temperatures on crystal morphology.
As seen from Table 1, the pH value as well as the reaction
temperature play a crucial role in determining the final crystal
a set of pure WO
for the reaction temperature, the samples synthesized at a pH of
.5 are further compared. Table 2 shows the BET surface areas
3
samples with similar reaction conditions except
0
2
-1
in m g as a function of synthesis temperature. An increase in
reaction temperature results in a slight decrease in BET surface
3
area of the monoclinic WO samples, which is possibly the result
of the nanoplates becoming smoother and larger in size. The
sample prepared at 150 °C consists of a mixture with hydrate
phase, which causes a slightly higher surface area.
phase of WO
is formed at high temperatures and low pH values during the
synthesis. At a pH of 1.5 hexagonal WO and WO · 0.33H O are
predominantly formed. At this pH, monoclinic WO is only
3 3
that is formed. Generally, a monoclinic WO phase
3
3
2
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
Table 2. BET surface area as a function of temperature of samples
synthesized through microwave-assisted hydrothermal synthesis at
pH = 0.5.
T (°C)
BET (m2 g-1)
Bandgap (eV)
150
170
200
230
250
17.61
14.76
13.84
13.44
13.41
2.55
2.69
2.69
2.69
2.68
Kubelka-Munk and Tauc plots of the samples obtained by
microwave-assisted hydrothermal synthesis at a pH of 0.5 are
depicted in Figure 3. The (indirect) bandgaps deducted from the
Tauc plots are depicted in Table 2. For each m-WO sample, an
3
identical bandgap of 2.7 eV is calculated, in consistency with
literature.[ For the WO
1]
sample synthesized at a pH of 0.5 and
3
o
T = 150 C, determining the bandgap is more complicated, as
there are two crystal phases present. From the as-obtained Tauc
plot, we estimate the bandgap to be at 2.55 eV. The slightly lower
2
bandgap might be the result of H O groups in the crystal lattice,
inducing defects.
3
Figure 3. (a) Kubelka-Munk and (b) Tauc plots of WO obtained through
microwave-assisted hydrothermal synthesis at pH = 0.5 at variable temperature.
Discussion on crystal phase transformations in (microwave-
assisted) hydrothermal synthesis.
As observed in the previous sections, the pH and the synthesis
temperature play a crucial role in determining the obtained crystal
phase of the WO
in more detail. At low pH (≤ 1) and temperatures of 170 ºC or
higher, the final product is phase-pure monoclinic WO . This trend
3
. Here, we discuss the role of these parameters
3
is both observed for the ‘conventional’ hydrothermal synthesis
experiments (Figures 1 and S3) and the microwave-assisted
hydrothermal synthesis experiments (Table 1). We speculate that
the basis for obtaining this phase-pure monoclinic WO
formation of tungstic acid seeds (tungsten oxide hydrates in the
form of WO · 2H O or WO · H O) at these low pH values at room
3
lies in the
3
2
3
2
temperature, which also involves coordination of citric acid
molecules. This is evidenced by the visible formation of a yellow
precipitate during the preparation of the solution used in the
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
hydrothermal synthesis procedure. The formation of such a yellow
precipitate is not observed at a pH of 1.5 and 2, nor in the absence
of citric acid at any pH. At these conditions, tungsten will be
present in the form of a dissolved tungstate rather than a tungsten
oxide hydrate. Considering the Pourbaix diagram of tungsten,
explained by poisoning of the catalyst surface by adsorption of
oxidation products. A repetition of this particular experiment (runs
13 to 18 in Figure 4a), after the coating was used to construct the
curve of Figure 4b, shows stabilization in the CO formation rate
2
takes place to approximately the same value as in runs 3 to 8.
Figure 4b shows that initially the relationship between the amount
of CO formed and the reaction time is linear. When the same
2
coating was exposed to the same set of experiments (cycle 2 in
Figure 4), comparable behavior was observed.
2 4
WO
occurs at a pH higher than 2.[20] We
deprotonation of H
speculate that similar effects occur in our experiments at a pH of
.5 and 2. Tungsten will be present in an anionic form, preventing
1
citric acid to interact properly with these tungsten seeds. This
prevents formation of a tungsten precipitate. Unfortunately we
were not able to resolve the chemical nature of the interaction of
citric acid with the tungstic acid seeds experimentally. Although
we can detect citric acid in aqueous solution using Raman
spectroscopy (Figure S7), citric acid could unfortunately not be
detected in solution during the in situ hydrothermal treatment
2 3
Figure 5 depicts the average rate of CO formation over WO
photocatalyst synthesized at variable conditions, corrected for
their BET surface area. When the synthesis temperature was 170,
o
200, 230 or 250 C, the rate is approximately similar (around 6.3
to 7 µmol CO
2
m-2 h-1). The sample synthesized at 150 oC is
-
2
considerably less active than the others (around 2.9 µmol CO m
2
-1
(
Figures
1
and S3). Likely, the intensity of the peaks
h ). Obviously, the main difference between this sample and the
others is the presence of both monoclinic WO and WO · H O,
whereas in the other samples only monoclinic WO is present.
This underlines that the presence of the WO · H O seems
detrimental for the photocatalytic activity. A likely explanation for
this behavior would be that the hydrate in WO · H O acts as a
corresponding to tungsten species is much stronger than the
intensity of the peaks corresponding with the citric acid in solution.
During the hydrothermal synthesis at elevated temperatures,
3
3
2
3
3
2
crystal phase transitions WO
3
· 2H
2
O  WO
3
· H
2
O  m-WO
3
take place due to decomposition of the hydrate phases.
A
3
2
possible explanation for the role of the citric acid in the
charge carrier recombination trap site, i.e. the lifetime of the
has been provided by Biswas et al.[
21]
excited electrons and holes is dramatically reduced.
crystallization of m-WO
3
o
Prior to hydrothermal treatment, acidification could result in
polycondensation of WO
2-. However, this can be prevented by
the introduction of citric acid. This allows the formation of seeds
of WO hydrates. When these are formed, citric acid will adsorb
The hexagonal WO
3
sample synthesized at 170 C and a pH of 2
4
has been tested on its photocatalytic activity as well. However,
the activity under visible light is inferior to its monoclinic
counterpart (Figure S8). In previous studies, Nagy et al. and
3
on specific facets of the seeds, changing the order of free
energies of the facets. In such a way, the growth rates of the
different facets are affected, ultimately resulting in the formation
Adhikari et al. demonstrate that the bandgap of hexagonal WO
is larger than the bandgap of monoclinic WO
.[23] Here, it is very
likely that the hexagonal WO also has a larger bandgap than the
monoclinic WO . Illumination took place at 420 nm, which
corresponds to a photon energy of 2.95 eV. It is plausible that the
bandgap of the hexagonal WO exceeds 2.95 eV, hence
3
3
3
of WO
3
nanoplates.
3
It is reasonable to assume that the citric acid is decomposed
during the hydrothermal synthesis process. This is amongst
others evidenced by a slight increase in pressure (Figure S1),
3
explaining the negligible activity in photocatalytic propane
oxidation.
2
which is likely the result of CO formation due to citric acid
decomposition. Furthermore, we observe in Figure S3 after 10
hours in the hydrothermal synthesis procedure the formation of
-1
broad peaks at 939 and 969 cm . Formation of such peaks could
be the result of citric acid decomposition in other species, such as
acetonedicarboxylic acid and acetoacetic acid.[22]
Photocatalytic activity measurements.
To demonstrate the relevance of this work for photocatalysis, the
samples synthesized by the microwave hydrothermal method at
a pH of 0.5 were analyzed for photocatalytic activity in the
oxidation of propane. As CO
production of CO was taken as a basis for photocatalytic activity.
A typical measurement is included in Figure 4. Figure 4a
demonstrates the rate of CO formation based on analysis of CO
quantity obtained after 10 minutes of illumination, whereas Figure
b demonstrates the yield of CO as a function of increasing
reaction time. In the dark-measurements hardly any CO is
formed, whereas considerably large amounts of CO are formed
upon illumination at 420 nm. Figure 4a demonstrates that the CO
2
was very dominantly formed, the
2
2
2
4
2
2
2
2
formation rate slightly decreases as a function of increasing light
exposure time (runs 3 to 8). The decrease in rate might be
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
Figure 5. Average CO
2
production rates during photocatalytic oxidation of
3
propane using coatings of WO nanoplates obtained through microwave-
assisted hydrothermal synthesis at a pH of 0.5, corrected for BET surface area.
Conclusions
In this manuscript, insight is provided in the phase transformations
occurring upon hydrothermal synthesis of WO
tungstate solution containing citric acid as
temperature and pH. If the pH is 1 or lower, mainly tungsten oxide
dihydrate (WO · 2H O) is formed. During heating, starting from
0 °C, the tungsten oxide dihydrate is first transformed to tungsten
oxide monohydrate (WO · H O). At temperatures above 150 °C,
WO · H O is transformed mainly into monoclinic WO . To a minor
extent, WO · 0.33H O is also formed, but this crystal phase is
slowly decomposed into monoclinic WO in the isothermal stage.
3
aqueous sodium
a
function of
3
2
7
3
2
3
2
3
3
2
3
Figure 4. CO
propane in visible light (420 nm) over a coating of WO
at pH = 0.5 and T = 200 °C in a microwave. The same set of experiments was
repeated (cycles 1 & 2). The dashed line in (a) represents the moment that the
measurements of (b) took place.
2
formation rates (a) and yield (b) in photocatalytic oxidation of
nanoplates, synthesized
Microwave-assisted hydrothermal synthesis experiments indicate
that there is a tipping point in pH in the synthesis: underneath this
3
3
pH, the dominant crystal phase is monoclinic WO , provided the
temperature is high enough to prevent formation of hydrated WO
3
.
Furthermore, BET measurements have demonstrated that an
increase in reaction temperature results in more uniform samples,
while Kubelka-Munk and Tauc plots indicate that the bandgap
3
remains constant for phase-pure monoclinic WO .
The photocatalytic activity of the samples synthesized through
microwave hydrothermal synthesis at a pH of 0.5 has been tested
in the photo-oxidation of propane under visible light (420 nm). We
observe that the synthesis temperature affects the amounts of
CO
photocatalytically inactive hydrates (WO
resulting in a decrease of photocatalytic activity. When the
temperature is sufficiently high, only monoclinic WO is formed
and the photocatalytic activity is considerably higher. The
monoclinic WO samples did not show a significant difference in
2
formed: when the synthesis temperature is too low,
3
· H O) will be present,
2
3
3
photocatalytic activity.
The studies as performed in this paper can contribute greatly to
understanding crystal growth of WO
this in optimizing crystal facet engineering of WO
3
using citric acid and to use
. For example,
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
we have recently published studies on structure-directed
employing diffuse reflectance spectroscopy. To this end, a DRS-cell
(Harrick, Praying Mentis) and a UV-Vis spectrophotometer (Thermo
3
deposition of Pt on WO particles synthesized in a Teflon-lined
748 acid digestion vessel.[24]
Scientific, Evolution 600) were combined. Barium sulfate (BaSO ) was
4
4
used as a reference. From as-measured diffuse reflectance spectra
Kubelka-Munk and Tauc plots were constructed. Using the latter, the
indirect bandgap was determined by deriving the intersection of the slope
and the baseline, as described by Montini et al.[25]
Experimental Section
Photocatalytic activity measurements: Photocatalytic activity
In situ Raman spectroscopy of crystal phases during hydrothermal
measurements were performed with plate-like WO
using microwave-assisted hydrothermal synthesis. Photocatalytic
oxidation of propane (C ), of which the products were analyzed by gas
chromatography, was used as prove reaction. To this end a 2 mL batch
reactor in combination with an Agilent 7820 GC system (containing a
Varian CP7584 column and a Methanizer-FID combination) as described
by Fraters et al. was used.[18a] Before the samples were loaded into the
reaction cell, coatings of these samples on glass substrates (25.3 mm x
25.8 mm) were made through dropcasting, using a stock slurry solution of
3
samples synthesized
synthesis of WO
from Sigma-Aldrich, Merck and Alfa Aesar. In a typical hydrothermal
experiment, 4.33g of sodium tungstate dihydrate (Na WO · 2H O) and
.60g of citric acid (CA) were dissolved, each in 30 g of Millipore water.
3
: All chemicals of analytical grades were purchased
3 8
H
2
4
2
2
The solutions were further diluted in an additional 200 mL of Millipore water
and filled into an MED1075 autoclave giving a total W-concentration of
0.05 M and a molar ratio of approximately W/CA = 1:1. The pH was
adjusted to 0.5 or 1.0 using concentrated HCl (32%). The solution was
-
heated to the reaction temperature of 150-170 °C with a rate of 1.5 K min
3 mL Millipore H
measured using a HI 1332 probe in combination with a pH 209 pH meter
from Hanna instruments) with 150 mg of as-synthesized WO . The glass
substrates were cleaned through sonication in acetone first, then in ethanol,
followed by treatment for 30 minutes in a mixture of H O, H (≥35%) and
NH OH (28-30% NH content) with a ratio of 5:1:1. After treatment, the
2
O (pH = 2, pH adjusted with concentrated HCl and
1
and kept isothermal at this reaction temperature for 17 hours. During the
entire experiment, the solution was stirred at 300 rpm. In situ Raman
3
measurements were performed using
Spectrometer RXN1 equipped with
a
Kaiser Optics Raman
a
fiber-optic probe-head (laser
2
2 2
O
wavelength at 785 nm, 125 mW). Samples were withdrawn from the vessel
at different temperatures and reaction times for X-ray diffraction (XRD) and
scanning electron microscopy (SEM) investigations of the precipitates,
after washing with water and drying. Temperature/pressure profiles of the
experiments can be found in the supplementary information (Figure S1).
To monitor the pH, an InLab Routine Pro model from Mettler Toledo was
4
3
substrates were put on a heating plate preheated at 100 °C. 750 µL of the
slurry solution was dropcasted on each glass plate. After evaporation of
the water a uniform coating was obtained. To ensure tightness of the
reactor cell, coating was removed (by scraping with a spatula) so that an
inner circle with a diameter of 1.25 cm remained (corresponding with a
o
2
used for temperatures up to 80-90 C. During the hydrothermal synthesis,
sample area of 0.61 cm ). The mass of the sample present on the coating
for temperatures above 120 oC, the pH could also be measured using a
was calculated to be approximately 7.05 mg.
2
ZrO electrode together with a Ag/AgCl reference electrode (both from
Corr instruments). The as-prepared products were washed with water,
then either with ethanol or a 1:2 water/ethanol mixture and were dried
overnight at a maximum temperature of 100 °C.
As-made coatings on glass substrates were loaded into the reaction cell
for photocatalytic testing. Prior to the measurements, the batch reactor
2 2
was purged with a gas mixture containing 80 vol% N , 19.5 vol% O and
5
000 vol ppm propane (C ) for a minimum of 21 minutes. Afterwards,
3 8
H
Microwave-assisted
Microwave-assisted hydrothermal synthesis studies have been performed
with a Multiwave PRO from Anton Paar using 80 ml quartz vessels. In a
typical synthesis, 2.64 g of Na
each dissolved in 80 mL H
together and the pH was adjusted to values in a range of 0.5 to 2 using
diluted HCl (the pH was determined using an InLab Routine Pro Model
hydrothermal
synthesis
experiments:
the valves of the batch reactor were closed, followed by illumination for 10
minutes with a 420 nm LED (intensity 6.2 mW/cm at the coating surface).
2
Afterwards, the reactor was purged with He and the gas mixture was
analysed by gas chromatography. For each coating, the following
sequence of runs was performed: i) two runs in the dark, followed by six
runs under illumination, ii) a sequence of runs under different illumination
times, iii) a sequence of runs under different reaction times in the dark and
iv) again, two runs in the dark, followed by six runs under illumination. An
2
WO
4 2
· 2H O and 1.68 g of citric acid were
2
O. Afterwards, the solutions were mixed
2
from Mettler Toledo). After dilution with H O to a total volume of 160 mL,
2
the solution was equally distributed over four cuvettes, which were placed
in the microwave hydrothermal synthesis oven. The reaction was
conducted at temperatures in the range of 150 °C to 250 °C for 2 hours
using a heating rate of 10 K min . It should be noted that the heating rate
is different compared to the experiments with the MED1075 autoclave. The
phase composition of the products was studied by using powder X-ray
diffraction (XRD), using a STOE STADI P transmission diffractometer
equipped with a primary focusing germanium monochromator (CuKα1
radiation) and a linear position sensitive detector. Diffraction data were
analyzed by whole pattern Rietveld fitting using the program TOPAS
area of 0.61 cm was illuminated in the experiments.
2
We will focus on the formation of CO , as selectivity towards this
-1
compound was by far the highest. The reaction rate r of CO
(
2
formation
-
1
-
1
[
1
8
a
]
m
o
l
g
h
)
w
a
s
c
a
l
c
u
l
a
t
e
d
a
s
f
o
l
l
o
w
s
,
i
n
s
p
i
r
e
d
b
y
F
r
a
t
e
r
s
e
t
a
l
.
:
ꢀ
∙ꢁꢂꢃꢄ∙ꢅ
ꢆ∙ꢇ
ꢈ
ꢎꢏꢐꢐ
ꢑ
푟 =
∙
∙
(1)
ꢉ∙ꢊꢋꢌꢍ
(Bruker AXS). With this software, not only can we carefully monitor which
crystal phases are formed, but also in which ratio they are present. The
morphology was studied by scanning electron microscopy (SEM) using a
Hitachi S-4800 electron microscope operating at 1.5 kV in secondary
electron (SE) mode. Nitrogen physisorption measurements were
performed at liquid N temperature on a Quantachrome Autosorb-6B
2
analyzer. Prior to measurement, the samples were outgassed in vacuum
at 150 °C for 3 h. The specific surface area was calculated according to
In equation 1, P is the pressure inside the batch reactor (Pa), V the gas
3
3
-1 -1
volume inside the reactor (m ), R the gas constant (m Pa mol K ), T the
temperature inside the reactor (K), m the mass of the catalyst (g), 푆 the
ꢒꢓꢇ
2
-1
BET surface area of the sample (m g ), t the illumination time (s) and 푋
ꢔꢕꢄ
the gas fraction of CO
the WO samples need to stabilize over multiple runs to obtain a constant
propane oxidation rate. Therefore, we will focus in this manuscript mainly
2
present inside the reactor. Our experience is that
the multipoint Brunauer–Emmett–Teller method (BET). For the WO
3
3
samples synthesized at a pH of 0.5, the bandgap was determined by
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
on the last three runs of the 4th sequence. From these runs we calculated
[9]
a) J. Zhu, S. Wang, S. Xie, H. Li, Chemical Communications 2011, 47,
4403-4405; b) J. Wang, E. Khoo, P. S. Lee, J. Ma, Journal of Physical
Chemistry C 2008, 112, 14306-14312; c) R. F. Mo, G. Q. Jin, X. Y. Guo,
Materials Letters 2007, 61, 3787-3790.
the average reaction rate of CO
2
formation. By repeating this
measurement using either additional measurements on the same coating
and/or other coatings, a final average reaction rate including error bars
was calculated.
[10] a) Z. Gu, H. Li, T. Zhai, W. Yang, Y. Xia, Y. Ma, J. Yao, Journal of Solid
State Chemistry 2007, 180, 98-105; b) Z. Gu, T. Zhai, B. Gao, X. Sheng,
Y. Wang, H. Fu, Y. Ma, J. Yao, Journal of Physical Chemistry B 2006,
1
10, 23829-23836.
11] a) X. Su, F. Xiao, Y. Li, J. Jian, Q. Sun, J. Wang, Materials Letters 2010,
4, 1232-1234; b) J. Wang, P. S. Lee, J. Ma, Journal of Crystal Growth
Acknowledgements
[
6
This project was funded by the Dutch National Research School
Combination Catalysis Controlled by Chemical Design (NRSC-
Catalysis). From the University of Twente, we would like to thank
Vera Smulders for her help in performing Raman measurements
concerning citric acid in aqueous solution. Furthermore, from the
Fritz Haber Institute, we would like to acknowledge Jasmin Allan
for performing XRD measurements and we would like to thank
Frank Girgsdies for his help in analyzing the XRD patterns. Also
we want to thank Gisela Weinberg for making HR-SEM images
and Maike Hashagen for performing BET surface area
measurements.
2009, 311, 316-319; c) Y. P. Xie, G. Liu, L. Yin, H. M. Cheng, Journal of
Materials Chemistry 2012, 22, 6746-6751.
[
12] Z. G. Zhao, Z. F. Liu, M. Miyauchi, Chemical Communications 2010, 46,
3321-3323.
[
13] S. Jeon, K. Yong, Journal of Materials Chemistry 2010, 20, 10146-10151.
14] a) Y. Zhao, H. Chen, X. Wang, J. He, Y. Yu, H. He, Analytica Chimica
Acta 2010, 675, 36-41; b) J. Yu, L. Qi, Journal of Hazardous Materials
2009, 169, 221-227; c) S. K. Biswas, J. O. Baeg, International Journal of
Hydrogen Energy 2013, 38, 3177-3188.
[
[
[
[
15] G. Liu, J. C. Yu, G. Q. Lu, H. M. Cheng, Chemical Communications 2011,
4
7, 6763-6783.
16] J. Sungpanich, T. Thongtem, S. Thongtem, Ceramics International 2012,
8, 1051-1055.
3
17] J. Noack, F. Rosowski, R. Schlögl, A. Trunschke, Zeitschrift fur
Anorganische und Allgemeine Chemie 2014, 640, 2730-2736.
Keywords: Crystal engineering • Tungsten oxide • Hydrothermal
synthesis • Raman spectroscopy • Photocatalysis
[18] a) B. D. Fraters, R. Amrollahi, G. Mul, Journal of Catalysis 2015, 324,
19-126; b) M. Khajeh Aminian, M. Hakimi, Catalysis Science and
1
Technology 2014, 4, 657-664.
[
1]
a) H. Zheng, J. Z. Ou, M. S. Strano, R. B. Kaner, A. Mitchell, K. Kalantar-
Zadeh, Advanced Functional Materials 2011, 21, 2175-2196; b) J. Kim,
C. W. Lee, W. Choi, Environmental Science and Technology 2010, 44,
[
19] a) A. G. Souza-Filho, V. N. Freire, J. M. Sasaki, J. Mendes Filho, J. F.
Julião, U. U. Gomes, Journal of Raman Spectroscopy 2000, 31, 451-454;
b) A. Baserga, V. Russo, F. Di Fonzo, A. Bailini, D. Cattaneo, C. S. Casari,
A. Li Bassi, C. E. Bottani, Thin Solid Films 2007, 515, 6465-6469; c) M.
Gotić, M. Ivanda, S. Popović, S. Musić, Materials Science and
Engineering B: Solid-State Materials for Advanced Technology 2000, 77,
6849-6854; c) A. Kudo, Y. Miseki, Chemical Society Reviews 2009, 38,
253-278.
[2]
a) S. H. S. Chan, T. Yeong Wu, J. C. Juan, C. Y. Teh, Journal of
Chemical Technology and Biotechnology 2011, 86, 1130-1158; b) R. Abe,
M. Higashi, K. Domen, ChemSusChem 2011, 4, 228-237.
193-201; d) M. F. Daniel, B. Desbat, J. C. Lassegues, B. Gerand, M.
Figlarz, Journal of Solid State Chemistry 1987, 67, 235-247.
20] E. Patrick, M. E. Orazem, J. C. Sanchez, T. Nishida, Journal of
Neuroscience Methods 2011, 198, 158-171.
[
3]
4]
M. Deepa, A. G. Joshi, A. K. Srivastava, S. M. Shivaprasad, S. A.
Agnihotry, Journal of the Electrochemical Society 2006, 153, C365-C376.
a) H. G. Choi, Y. H. Jung, D. K. Kim, Journal of the American Ceramic
Society 2005, 88, 1684-1686; b) Z. G. Zhao, M. Miyauchi, Journal of
Physical Chemistry C 2009, 113, 6539-6546.
[
[
[
[
[
21] S. K. Biswas, J. O. Baeg, S. J. Moon, K. J. Kong, W. W. So, Journal of
Nanoparticle Research 2012, 14.
22] C. H. Munro, W. E. Smith, M. Garner, J. Clarkson, P. C. White, Langmuir
[5]
[6]
[7]
[8]
G. Gu, B. Zheng, W. Q. Han, S. Roth, J. Liu, Nano Letters 2002, 2, 849-
1995, 11, 3712-3720.
851.
23] a) D. Nagy, D. Nagy, I. M. Szilagyi, X. Fan, RSC Advances 2016, 6,
M. Regragui, M. Addou, A. Outzourhit, J. C. Bernéde, E. El Idrissi, E.
Benseddik, A. Kachouane, Thin Solid Films 2000, 358, 40-45.
33743-33754; b) S. Adhikari, D. Sarkar, RSC Advances 2014, 4, 20145-
20153.
N. Shankar, M. F. Yu, S. P. Vanka, N. G. Glumac, Materials Letters 2006,
[
24] K. Wenderich, A. Klaassen, I. Siretanu, F. Mugele, G. Mul, Angewandte
Chemie International Edition 2014, 53, 12476-12479.
60, 771-774.
B. C. Satishkumar, A. Govindaraj, M. Nath, C. N. R. Rao, Journal of
Materials Chemistry 2000, 10, 2115-2119.
[
25] T. Montini, V. Gombac, A. Hameed, L. Felisari, G. Adami, P. Fornasiero,
Chemical Physics Letters 2010, 498, 113-119.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201701156
FULL PAPER
Entry for the Table of Contents
FULL PAPER
In situ Raman spectroscopy and
microwave-assisted hydrothermal
synthesis experiments show a pH of
Crystal engineering
Kasper Wenderich*, Johannes Noack,
Anne Kärgel, Annette Trunschke, Guido
Mul
0
.5 and a temperature above 170 ºC
are required to obtain plate-like,
phase-pure monoclinic WO
3
Page No. – Page No.
nanoplates from a solution containing
sodium tungstate and citric acid.
These phase-pure samples show
better photocatalytic performance in
the oxidation of gaseous propane than
Effect of temperature and pH on
phase transformations in
citric acid-mediated hydrothermal
growth of tungsten oxide
samples containing hydrated WO
phases.
3
This article is protected by copyright. All rights reserved.