Fast response-recovery time toward acetone by a sensor prepared with Pd doped WO3 nanosheets

Article abstract of DOI:10.1039/c9ra04429a

Pd-WO₃ nanosheets were synthesized through a one-step hydrothermal method using Na₂PdCl₄ solution as the palladium source and sodium tungstate as the tungsten source, and were used to detect acetone. After being characterized by TEM, XRD, BET and XPS, it was found that Pd doped on the surface of WO₃ nanosheets was mainly present as metal palladium, and the specific surface area increased after doping. In addition, the effect of Pd doping on gas sensing properties was studied. When the Pd-doped amount was 2 at%, sensors fabricated with the composites had the best gas sensing performance. Under a 100 ppm acetone atmosphere, the response time was 1 s and the recovery time was 9 s. The detection limit for acetone was 50 ppb at the optimum working temperature of 300 °C, and the selectivity for acetone was excellent under 100 ppm atmosphere (S_acetone/S_ethanol = 5.06). The excellent gas sensing properties of this material are mainly attributed to the high catalytic activity and the catalytic spill-over effect of the Pd nanoparticles, which provided additional active sites for the sensitive materials.

Full text of DOI:10.1039/c9ra04429a

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Fast response–recovery time toward acetone by
a sensor prepared with Pd doped WO3 nanosheets†
Cite this: RSC Adv., 2019, 9, 28439
Ke He,‡a Zhao Jin,‡a Xiangfeng Chu,b Wenjie Bi,a Wu Wang,a Chao Wanga
a
*
and Shantang Liu
Pd-WO3 nanosheets were synthesized through a one-step hydrothermal method using Na2PdCl4 solution
as the palladium source and sodium tungstate as the tungsten source, and were used to detect acetone.
After being characterized by TEM, XRD, BET and XPS, it was found that Pd doped on the surface of WO3
nanosheets was mainly present as metal palladium, and the specific surface area increased after doping.
In addition, the effect of Pd doping on gas sensing properties was studied. When the Pd-doped amount
was 2 at%, sensors fabricated with the composites had the best gas sensing performance. Under
a 100 ppm acetone atmosphere, the response time was 1 s and the recovery time was 9 s. The detection
limit for acetone was 50 ppb at the optimum working temperature of 300 ꢀC, and the selectivity for
acetone was excellent under 100 ppm atmosphere (Sacetone/Sethanol ¼ 5.06). The excellent gas sensing
properties of this material are mainly attributed to the high catalytic activity and the catalytic spill-over
effect of the Pd nanoparticles, which provided additional active sites for the sensitive materials.
Received 12th June 2019
Accepted 25th August 2019
DOI: 10.1039/c9ra04429a
rsc.li/rsc-advances
a high selectivity. By increasing the specic surface area of the
sensing material, modifying its the surface and adjusting its
Introduction
defects, its gas sensing performance can be greatly
improved.12–14 Surface modication of materials is an effective
method to improve the gas sensing performance of gas sensors,
and the morphology also has some inuence. Compared with
other structures,15–17 nanosheets have many advantages, such as
large specic surface areas, abundant surface adsorption
centers and unique electronic transmission performance.18
While they have broad application prospects in gas sensors, Pd-
WO3 nanosheets composites with fast response–recovery time
have rarely been reported for gas sensing applications.
As society has developed, the emission of volatile organic
compounds (VOCs) such as acetone, formaldehyde and toluene
has increased, which seriously endangers human health.1,2 For
instance, it was found that some trace VOCs present in human
breath are indicators for different diseases. For example, the
marker gas for type-1 diabetes is acetone (>1.8 ppm),3 so there is
great practical signicance and value in developing gas sensors
to rapidly monitor acetone gas.4,5 Therefore, it is becoming
increasingly important to study various methods and technol-
ogies to rapidly detect and monitor VOCs.2,3,6,7
In this paper, Pd-WO3 nanosheets composites were prepared
by one-step hydrothermal method. It was found that the sensors
fabricated with the composites had a fast response, low detection
limit, high selectivity and high sensitivity. This is due to Pd
nanoparticles on the surface of the WO3 nanosheets, thus
providing more active sites for acetone reaction. The catalytic
spill-over effect of precious metal Pd reduced the activation
energy required for the reaction and the response–recovery time.
Currently, gas sensors are widely used as cost-effective
detection methods, and metal oxide semiconductor (MOS) gas
sensor are inexpensive, easily integrated and have long service
lives. These advantages arise from the change in conductivity
caused by the interaction between gas molecules and the
surface of metal oxide semiconductors (MOS).8 WO3 is a typical
n-type semiconductor with great potential for the detection of
harmful and toxic gases such as acetone, CO, NO2, H2S and
NH3.9–11 However, it is difficult for pure WO3 to simultaneously
have a fast acetone response–recovery, a low detection limit and
Experimental section
Materials
aKey Laboratory for Green Chemical Process of Ministry of Education, School of
Chemistry and Environmental Engineering, Wuhan Institute of Technology,
Xiongchu Avenue, Wuhan 430073, China. E-mail: stliu@wit.edu.cn
bSchool of Chemistry and Chemical Engineering, Anhui University of Technology,
Maanshan 243002, Anhui, P. R. China
Sodium tungstate dehydrate (Na2WO4$2H2O, 99.5%), sodium
tetrachloropalladate(II) (AR), citric acid (CA, >99.5%), D-
(+)-glucose (AR), hydrochloric acid (36–38%, AR) and liquid
chemicals (including methanal, methanol, ethanol, aqua
ammonia, acetone, toluene, and n-hexane) were purchased
from Sinopharm Chemical Reagent Co. Ltd, China. All reagents
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c9ra04429a
‡ Ke He and Zhao Jin contributed equally to this work.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 28439–28450 | 28439

View Article Online
RSC Advances
Paper
used in our experiments were of analytical grade and used process in a test chamber with an ambient relative humidity of
without further purication.
about 25% and room temperature near 25 ꢀC. Statistics were
performed on the resulting measurements. The sensor device
and the measurement system are reported by our previous
work.20,21 First, sensors were placed into the chamber and pre-
heated to their operational temperature, then the evaporator
was turned on for one minute, and a certain amount of
measured VOCs were injected into the test chamber. The upper
cover of the test bench was opened aer the response reaching
a constant value, then the sensor began to recover in air and the
sensor response returned to the initial value. The response
result of the sensor in this work was dened as S ¼ Ra/Rg (the
reducing gas) or Rg/Ra (the oxidizing gas), where Ra and Rg are
the resistances of the sensor in air and the target gas, respec-
tively. The response and recovery times were dened as the time
required to reach 90% of the total resistance change in
adsorption and desorption.22
Synthesis
WO3 and Pd-WO3 nanosheets were synthesized through
a hydrothermal method. In this work, a typical process was as
follow:19 2 mmol of Na2WO4$2H2O was dissolved in 60 mL of
deionized water to prepare a transparent solution, which was
dened as solution A. Aer vigorous stirring, 3 mmol of citric
acid and 10 mmol of glucose were added into solution A, which
was dened as solution B. Aer vigorous stirring for 10 min,
3 mL of HCl solution (12 M) was added into solution B. The
solution B was vigorous stirred for 30 min and transferred into
a 100 mL Teon-lined autoclave, which was then heated at
120 ꢀC for 24 h. Keep the autoclave at room temperature until it
had cooled, the resultant product was ltered, washed with
water and anhydrous ethanol for three times. Finally, the
product was dried at 60 ꢀC in a vacuum oven for 12 h and
ꢀ
annealing in a nitrogen atmosphere at 400 C for 2 h. Pd-WO3
Results and discussions
Structure and micro-morphology characterization
nanosheets were synthesized through a one-step hydrothermal
method. For doping, different molar ratios of Na2WO4$2H2O
and Na2PdCl4 (1 wt%) were dissolved in distilled water to get 1,
2, 3, 5 and 8 at% Pd-WO3 nanosheets. Other experimental steps
were the same as above.
Fig. 1 showed the XRD patterns of the hydrothermally processed
product when WO3 was doped with Pd (0, 1, 2, 3, 5 and 8 at%).
Strong and sharp diffraction peaks are observed in this gure,
which indicates that the samples are highly crystalline. The
XRD pattern in Fig. 1 shows that all diffraction peaks in the
spectrum of the as-prepared samples can be well indexed to
a monoclinic phase of WO3 (JCPDS no. 43-1035; a ¼ 0.7297 nm,
b ¼ 0.7539 nm, c ¼ 0.7688 nm, b ¼ 90.91ꢀ), corresponding to the
following planes: (002), (020), (200), (120), (112), (202), (114) and
(420), respectively.23 The clear reections were assigned to the
orthorhombic phase of WO3. Furthermore, peaks of the Pd
nanoparticles are also observed in the spectrum, indicating that
the Pd was loaded onto the surface of WO3. The XRD pattern of
Pd-WO3 nanoparticles conrms that Pd was presented in the
Characterization
X-ray powder diffraction (XRD) was performed using a Shi-
madzu XRD-6000 diffractometer with a CuKa1 irradiation over
a 2q range from 10 to 90ꢀ to analyze the crystal phases of the
samples. The morphologies of the as-obtained products were
characterized and analyzed by transmission electron micro-
scope (TEM). High-resolution TEM (HR-TEM) images were ob-
tained using a JEOL JEM-2100 microscopy, operated at an
acceleration voltage of 200 kV, to characterize the sizes and
crystal structures of WO3 and Pd-WO3 nanosheets. The surface
areas were measured on a surface area analyzer (Micromeritics
ASAP 2020 V3.00H) and calculated using the Brunauer–
Emmett–Teller (BET) equation. X-ray photoelectron spectros-
copy (XPS, K-alpha) was used to investigate electron congura-
tions of the surfaces of WO3 and Pd-WO3 nanosheets.
Fabrication and test of the sensor
The as-synthesized products were ground and mixed with H2O,
ethylene glycol, and glycerol (1 : 2 : 2 mass ratio) to form
a paste. Then, the paste was dropped and spot-coated onto an
Al2O3 ceramic plate substrate (area of 8 mm ꢁ 10 mm ꢁ 0.65
mm), onto which a pair of Au interdigitated electrodes (elec-
trode width of 0.2 mm and gap width of 0.2 mm) had been
previously printed. Aerwards, the resulting chip was dried at
120 ꢀC for 2 h, and then heated at 400 ꢀC for another 2 h to
anneal under nitrogen atmosphere to ensure all solvents were
evaporated. The nal sensor was obtained aer aging at 200 ꢀC
for an additional 24 h. Finally, the gas sensing properties of the
sensors were tested by a static measurement system produced
by CGS-1TP (Beijing Elite Tech. Co., Ltd., China), which test
Fig. 1 XRD patterns of bare WO3 samples with 0 at%, 1 at%, 2 at%, 3
chamber is 18 L. The measurement was processed by a static at%, 5 at% and 8 at% Pd doping.
28440 | RSC Adv., 2019, 9, 28439–28450
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
metallic (Pd0) state, which was indexed to the face centered resolution TEM image of the WO3 nanosheets is shown in
cubic structure of Pd (JCPDS no. 87-0641), corresponding to the Fig. 2c–f. The clear two-dimensional ordered lattice structure
following planes: (111), (200) and (220), respectively.24–26 The indicates that the acquired WO3 nanosheets are orthorhombic
appearance of the palladium diffraction peak indicates that the structure. It can be seen from low-magnication TEM images
experimentally doped palladium is supported on the surface of that the doping has no effect on the morphology of WO3.
tungsten trioxide.
Furthermore, the HR-TEM images shown in Fig. 2c for mono-
Then, the morphology and inner crystal structures of the as- clinic pure WO3 exhibit the lattice spacing of 0.39 nm, which
synthesized samples were investigated by TEM measurements, can be attributed to a lattice fringe of the (002) crystal planes of
and pure WO3 and 2 at% Pd-WO3 were further characterized by WO3. In Fig. 2d, the lattice spacing values about 0.37, 0.376 and
TEM and HR-TEM. Fig. 2a–f shows the typical TEM observations 0.267 nm can be clearly assigned to reections from the (200),
of WO3 nanosheets. The low-magnication TEM images (Fig. 2a (020) and (202) planes of orthorhombic WO3, respectively.27,28
and b) illustrate that the product primarily possesses a sheet- Moreover, HR-TEM images in Fig. 2e for monoclinic 2 at% Pd-
like morphology according to the shallow contrast, and WO3 display a lattice spacing of 0.384 nm, which is attributed to
because the sheet shape of pure WO3 and 2 at% Pd-WO3 is the lattice fringe of the (002) crystal planes of WO3. In Fig. 2f,
irregular. Fig. 2a indicates a single nanosheets of a dimension the space between neighboring lattice planes is 0.267 nm,
within 100 nm and a thickness of about 20 nm. A typical high- whereas the space between neighboring lattice planes along
Fig. 2 Typical TEM images of as-prepared WO3 nanosheets (a) pure WO3 and (b) 2 at% doped Pd; (c–g) HRTEM images corresponding to their
individual nanosheets section.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 28439–28450 | 28441

View Article Online
RSC Advances
Paper
other directions are 0.377 and 0.376 nm. These spacings
Meanwhile, an angle of 90ꢀ between the (020) and (200)
correspond to the (020) and (200) planes of monoclinic WO3, planes was also observed in the high-resolution TEM images,
respectively. As shown in Fig. 2g, some aggregated palladium which are in great agreement with the values of the lattice
particles are observed, which have lattice spacing values of spacings and planes based on the XRD (Fig. 1). At the same
about 0.23 nm, which almost exactly corresponds to the (111) time, palladium doping also has no effect on the crystal
planes of Pd0 (JCPDS no. 87-0641). This is consistent with the structure.
XRD results (Fig. 1).
To obtain the surface compositions and chemical states of
the as-obtained WO3 nanosheets, X-ray photoelectron
Fig. 3 Survey scan XPS spectrum of pure WO3 and 2 at% Pd-WO3 nanosheets (a), high resolution XPS spectra of pure WO3 and 2 at% Pd-WO3
nanosheets in the vicinity of W 4f (b), XPS core-level O 1s (inset) spectra of the pure WO3 (c), 2 at% Pd-WO3 (d) and (e) Pd 3d spectra XPS for 2 at%
Pd-WO3.
28442 | RSC Adv., 2019, 9, 28439–28450
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
spectroscopy (XPS) was used to investigate the elements and Pd 3d5/2. The difference in the splitting energy between the
present, as shown in Fig. 3. The XPS survey spectrum in Fig. 3a two peaks is 5.3 eV, which again shows the presence of metallic
showed the presence of W, O and C in the WO3 nanosheets, and palladium (Pd0).35 However, no peaks were detected, indicating
peaks for other elements were not observed, demonstrating that the only metallic phase exists in the material. Therefore, it
highly pure WO3 nanosheets. The full-range XPS spectrum of 2 is veried that there is a palladium peak in the diffraction
at% Pd-WO3 (Fig. 3a) indicated that W, Pd, O and C were the spectrum in Fig. 1, which is consistent with the XRD (Fig. 1).
primary elements in the products, and no impurities were
To further study the detailed surface area and pore struc-
present, which was also conrmed by the XRD results. Fig. 3b–e tures of two-dimensional WO3 nanosheets, nitrogen adsorption
revealed more specic information on the chemical states of and desorption tests were carried out on the samples, and the
these elements from the high-resolution XPS spectra of W 4f, O results are shown in Fig. 4. The pure WO3 and 2 at% Pd-WO3
1s and Pd 3d. The W 4f spectrum (Fig. 3b) contained two peaks samples both show a typical type-IV isotherms with a typical H3
located at binding energies of 37.8 eV and 35.8 eV in pure WO3, type hysteresis loop, indicating that the material is a meso-
which are related to the presence of W6+ in the WO3 crystal porous or macro-porous material containing narrow slits
lattice and correspond to W 4f5/2 and W 4f7/2, respectively. In formed by a layered structure, as shown in Fig. 4a and b. The
the W 4f spectrum surveyed for 2 at% Pd-WO3 samples (Fig. 3b), BET specic surface area in Fig. 4 was calculated to be 39.47 m2
two peaks are observed at binding energies of 37.2 eV and gꢂ1 for 2 at% Pd-WO3 nanosheets, which is slightly higher than
35.2 eV, which correspond to W 4f5/2 and W 4f7/2, respectively.29 that of pure WO3 (25.69 m2 gꢂ1). In the illustration of Fig. 4a and
As can be seen from Fig. 3b, the W 4f peak shied towards b, the pore-size distribution curves of the pure WO3 and 2 at%
a higher binding energy aer doping, with an offset of 0.6 eV. Pd-WO3 samples show that the synthesized products have
The 0.6 eV binding energy difference conrms that palladium mesoporous or microporous structures. Fig. 4 shows that, the
was doped into the lattice, and the W5+ state existed in the pure WO3 nanosheets have an average pore width of 32 nm,
sample, indicating the presence of sub-stoichiometric WO3ꢂx while the 2 at% Pd-WO3 nanosheets have an average pore width
and oxygen vacancies.30–32 O 1s curves of the two samples could of 33 nm. The 2 at% Pd-WO3 samples obviously exhibited larger
be differentiated into three tted peaks, suggesting the different surface areas compared with the pure WO3. Accordingly, the
states of oxygen species on the surface of the samples: OL main reason for the increase in specic surface area is due to
(lattice oxygen species), OV (oxygen-vacancy) and the OC (surface the denser pore size distribution aer doping. Meanwhile the
chemisorbed oxygen species). In the O 1s XPS spectra for pure increase in the specic surface area provides more active
WO3 (Fig. 3c), the binding energies of 533.23 eV and 530.41 eV centers for gas sensor reactions, which should increase the
are attributed to the surface chemisorbed oxygen species and sensitivity of samples.
lattice oxygen in WO3, respectively. In Fig. 3d, the O 1s yields
have four peaks at 529.77, 529.99, 530.13 and 530.58 eV.33,34 The
rst two peaks are attributed to lattice oxygen, while the peak at
Gas sensing characteristics
530.13 eV ascribes to surface chemisorbed oxygen species. It can
be observed from Fig. 3c and d that the relative percentages of
OV and OC of the 2 at% Pd-WO3 composite were obviously
higher compared with that of as-prepared pure WO3. The Pd 3d
core level XPS spectrum in Fig. 3e indicates two peaks at
340.7 eV and 335.4 eV, which respectively correspond to Pd 3d3/2
The gas sensing results of the six sensors constructed with the
as-produced pure WO3, and 1 at%, 2 at%, 3 at%, 5 at% and 8
at% Pd-WO3 are presented in this section. In this work the
baseline resistances for pure and Pd loaded sensors is shown in
Fig. S1.† As chemical resistive sensors, the gas sensing perfor-
mance of these sensors was signicantly inuenced by the
Fig. 4 Typical nitrogen adsorption–desorption isotherms and the corresponding pore-size distribution of the pure WO3 nanosheets (a), 2 at%
Pd-WO3 nanosheets (b).
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 28439–28450 | 28443

View Article Online
RSC Advances
Paper
working temperature. The presence of ionized oxygen species response value of 2 at% Pd-WO3 was more than three times that
(O2ꢂ, Oꢂ and O2ꢂ) on the surfaces of semiconductors is closely of pure WO3. The sensitivity value S ¼ (DR/Ra) ꢁ 100% of Fig. 5a
related to the temperature, which changes of sensor resistance. was shown in Fig. S2.† The result showed that the response of
Thus, the optimum operating temperature of the as-produced Pd-WO3 sensors to 100 ppm acetone vapor increased with the
gas sensor using WO3 nanosheets was rst determined. The temperature increased from 150 to 300 ꢀC, and gradually
sensing response towards 100 ppm acetone is shown in Fig. 5a, decreased when the temperature further increased from 350 ꢀC.
including the working temperature curve of each sensor. From The 2 at% Pd-WO3 sensor displayed the best response
ꢀ
Fig. 5a, as the temperature increased from 150 to 300 C, the compared with others, which is consistent with the results of
response of pure WO3 and Pd-WO3 nanosheets to 100 ppm the operating temperature curve in Fig. 5a.
acetone vapor increases, and gradually decreased when the
It is reported that acetone detection has great practical
temperature further increased from 350 ꢀC. It can be clearly signicance.36 Meanwhile, as the environmental factor, the
observed that the response aer doping is higher than that of humidity has a great inuence on the WO3 sensor performance.
pure WO3. These results indicated that both the Pd doping According to the humidity control method,37 MgCl2, NaBr, NaCl
concentration and the working temperature greatly inuenced standard saturated solution were prepared to control the rela-
the response property, and the optimum operating of all tive humidity (33%, 59%, 75%) in the gas chamber. At the same
sensors was 300 ꢀC. As can be seen from the operating time, we used 2 at% Pd-WO3 to test the response value toward
temperature curve, the sensor based on 2 at% Pd-WO3 had the 100 ppm acetone at different humidity, which was showed in
highest response among the sensors, with a response of about Fig. 5b. Fig. 5b shows that the response of the 2 at% Pd-WO3
ꢀ
107.29 at 300 C, while it was about 29.472 for pure WO3. The sensor decreases as the humidity increases. At the same
Fig. 5 (a) Responses of the sensors are based on pure WO3, 1 at%, 2 at%, 3 at%, 5 at% and 8 at% Pd-WO3 at different working temperature toward
100 ppm acetone; (b) responses curves of the sensor are based on 2 at% Pd-WO3 at 33%, 59%, 75% relative humidity toward 100 ppm acetone; (c)
response of the six sensors toward various test gases with a concentration of 100 ppm at their optimal temperatures (300 ꢀC); (d) the error bars of
the sensors at different working temperature toward 100 ppm acetone.
28444 | RSC Adv., 2019, 9, 28439–28450
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
Table 1 A comparison of the gas-sensing characteristics on various sensing materials toward acetone
Sensing materials
Acetone
Operating temperature
Response (Ra/Rg)
Detection limit
Reference
WO3–C3N4 nanosheet
WO3 nanoplates
100 ppm
200 ppm
200 ppm
20 ppm
100 ppm
100 ppm
100 ppm
340 ꢀ
300 ꢀ
300 ꢀ
300 ꢀ
340 ꢀ
350 ꢀ
300 ꢀ
C
C
C
C
C
C
C
35
12
4.4
6.43
12.5
12.7
107.29
500 ppb
10 ppm
5 ppm
250 ppb
500 ppb
800 ppb
50 ppb
38
39
40
41
42
43
WO3–NiO nanorods
WO3–Cu hollow bers
Ag–WO3 nanosheets
La2O3–WO3 nanobers
2 at% Pd-WO3 nanosheets
This work
humidity, the optimum operating temperature of 300 ꢀC has the responses for all tested VOCs compared with the pure WO3.
best response toward 100 ppm acetone. Obviously, the 2 at% Pd-WO3 nanosheets displays the best
In addition, a good acetone sensor should be able to identify response to acetone (107.29), and negligible cross-responses to
acetone vapor in the atmosphere, and the selectivity properties ethanol, formaldehyde, methanol, toluene, aqua ammonia and
of sensors are shown in Fig. 5c. An obvious variation is observed n-hexane. In contrast, the pure WO3 nanosheets sensor showed
in the response of pure WO3, 1 at%, 2 at%, 3 at%, 5 at% and 8 a lower response and selectivity to all tested VOCs. Fig. 5c shows
at% Pd-WO3 sensors toward 100 ppm of acetone, methanol, that both the response value and the selectivity have a greatly
ethanol, aqua ammonia formaldehyde, toluene, and n-hexane improved by doping with palladium. The 2 at% doping amount
at 300 ꢀC. The 2 at% Pd-WO3 samples displayed enhanced shows the best selectivity for acetone vapor and the response is
Fig. 6 The dynamical response–recovery curves of the pure WO3 (a) and 2 at% Pd-WO3 (b) sensors to acetone with the increasing concentration
at their respective optimal temperature (300 ꢀC). (c) The response of pure WO3, 1 at%, 2 at%, 3 at%, 5 at% and 8 at% Pd-WO3 sensors to acetone
increasing concentration at their respective optimal temperature (300 ꢀC); (d) the error bars curve of pure WO3, 1 at%, 2 at%, 3 at%, 5 at% and 8
at% Pd-WO3 sensors to acetone increasing concentration at their optimum operating temperature (300 ꢀC).
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 28439–28450 | 28445

View Article Online
RSC Advances
Paper
Fig. 7 The actual response transient of the pure WO3 (a) and 2 at% Pd-WO3 (b) sensor to 100 ppm acetone vapor at their respective optimal
temperature (300 ꢀC), the response time and recovery time of the pure WO3 (c) and 2 at% Pd-WO3 (d) sensor toward acetone increasing
concentration at their respective optimal temperature (300 ꢀC).
more than three times that of pure WO3. Thus, it can be at%, 3 at%, 5 at% and 8 at% Pd-WO3 at different working
concluded that the Pd-WO3 is a potential sensing material for temperature toward 100 ppm acetone, which was shown in the
the detection of various reducing gases, especially acetone. The Fig. 5d. From Fig. 5d, as the temperature increased from 150 to
error bars curve of the sensors were based on pure WO3, 1 at%, 2 350 ꢀC, the error of the curve became smaller and error of curve
Fig. 8 The response–recovery curve of pure WO3 (a) and 2 at% Pd-WO3 (b) sensor to 100 ppm acetone at their respective optimal temperature
(300 ꢀC) was placed for 3 days, 1 week, 2 weeks, 1 month.
28446 | RSC Adv., 2019, 9, 28439–28450
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
was at a minimum at the optimum operating temperature of increased, the responses of all sensors were enhanced, which
300 ꢀC, revealing that the Pd-WO3 sensor possessed good clearly shows that there are signicant differences between the
stability at optimum operating temperature. The comparison of responses. Remarkably, the 2 at% Pd-WO3 sensor has a signi-
acetone sensing performance between this work and previously cantly higher response value in the full acetone concentration
reported results is showed in Table 1. Obviously, the 2 at% Pd- range than the other, which is consistent with the results of the
WO3 sensor displayed good performance, especially its low operating temperature curve in Fig. 5a. The error bars curve of
detection limit and high selectivity.
pure WO3, 1 at%, 2 at%, 3 at%, 5 at% and 8 at% Pd-WO3 sensors
To measure the current detection ability, the dynamic showed that the error of the sensors increased in acetone
response characteristic of the as-obtained nanosheets sensor increasing concentration at their optimum operating tempera-
toward various concentrations of acetone vapor was tested at ture of 300 ꢀC conditions, which was shown in the Fig. 6d.
300 ꢀC. The dynamic response–recovery characteristics of the
To quickly detect the gas, the response–recovery time of the
pure WO3 sensor is presented in Fig. 6a with an acetone sensors were explored. The transient dynamic response of the
concentration sequence of 0.2, 0.5, 1, 5, 10, 50, 100, 200 and pure WO3 and 2 at% Pd-WO3 sensors in Fig. 7a and b shows that
500 ppm. The inset shows dynamic response curves with an the response–recovery time of pure WO3 and 2 at% Pd-WO3
acetone concentration sequence of 0.2, 0.5, 1 and 5 ppm. The sensors are relatively fast. The corresponding response–recovery
corresponding response value are 1.24, 1.27, 2.5, 3.7, 4.9, 15.8, time is 2/10 s and 1/9 s toward 100 ppm acetone at their optimal
ꢀ
29.472, 47.7 and 112.5, respectively. Therefore, the lowest temperature (300 C). The response–recovery times of the pure
detection limit of the pure WO3 sensor is 200 ppb. Fig. 6b shows WO3 and 2 at% Pd-WO3 sensors at increasing concentrations of
the response of 2 at% Pd-WO3 samples to an increasing acetone vapor are also displayed in Fig. 8c and d. The response–
concentration of acetone at their optimum operating tempera- recovery time of the pure WO3 sensor is 9/10, 8/10, 7/13, 3/8, 2/9,
ture (300 ꢀC). When the concentration of acetone was 0.05, 0.1, 2/9, and 2/10 s when exposed to acetone gas at 1–500 ppm. In
0.2, 0.5, 1, 5, 10, 50, 100, 200 and 500 ppm, the response was addition, the response–recovery time of the 2 at% Pd-WO3
1.16, 1.2, 1.68, 2.08, 2.18, 11.21, 23.59, 78.86, 107.29, 128.52 and sensor is 10/12, 9/11, 8/10, 2/9, 1/9, 2/8, and 2/8 s exposed to
214.62, respectively. Obviously, the latter sensor shows a greatly acetone gas at 1–500 ppm. A comparable investigated of the
enhanced detection limit for acetone. In addition, the 2 at% Pd- sensing characteristics of the response–recovery time between
WO3 sensor has a higher sensitivity and more rapid response to the sensors in this work and other previously-reported acetone
acetone than the pure WO3 sensor at 300 ꢀC. The lowest sensors is given in Table 2. It was evident that the 2 at% Pd-WO3
detection limit of 2 at% Pd-WO3 sensor reached 50 ppb, which it sensor has a fast response and recovery.
lower than that of pure WO3. The response of pure WO3 and all
Stability and reproducibility are also vital parameters of
doped Pd-WO3-based gas sensors for variable acetone concen- gas sensors. To survey the stability and reproducibility of the
tration from 0.05 to 500 ppm at their optimal operating sensor response, 5 reversible cycles consisting of response
temperatures are shown in Fig. 6c. The inset of Fig. 6c shows the and recovery processes for the pure WO3 and 2 at% Pd-WO3
response to lower-concentrations of acetone between 0.05– sensor to 100 ppm acetone vapor are displayed in Fig. 8a and
10 ppm. In this gure, with the acetone concentration b. It was veried that the sensor revealed a relatively
Table 2 A comparison of the response–recovery times of sensors based on various sensing materials toward acetone
Sensing materials
Acetone
Operating temperature
Response time
Recovery time
Reference
Cactus-like WO3–SnO2 nanocomposite
Co3O4/ZnCo2O4 composite hollow
nanostructures
MoO3–WO3 nanobers
WO3 plate
Fe2O3–CuO nanorod
Ag–WO3 nanosheets
Au–ZnO yolk–shell
600 ppm
100 ppm
360 ꢀC
255 ꢀC
14 s
41 s
16 s
47 s
44
45
100 ppm
200 ppm
100 ppm
100 ppm
10 ppm
10 ppm
50 ppm
20 ppm
375 ꢀC
307 ꢀC
240 ꢀC
340 ꢀC
280 ꢀC
200 ꢀC
280 ꢀC
140 ꢀC
17 s
10 s
149 s
28 s
15 s
21 s
2 s
50 s
26 s
133 s
38 s
12 s
6 s
46
47
48
42
49
50
51
52
LaFeO3
Flower-like porous ZnO
Triple-shelled ZnO/ZnFe2O4
microspheres
Three-dimensional hierarchical SnO2
Hierarchical-out hierarchical Fe2O3
nanorod
Bamboo ra-like Co3O4
Pt–In2O3
Hierarchical Fe2O3
23 s
12.8 s
5.2 s
200 ppm
100 ppm
325 ꢀC
220 ꢀC
—
8 s
—
19 s
53
54
200 ppm
1.56 ppm
100 ppm
100 ppm
180 ꢀC
150 ꢀC
220 ꢀC
300 ꢀC
32 s
—
8 s
1 s
35 s
—
19 s
9 s
55
56
57
2 at% Pd-WO3 nanosheets
This work
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 28439–28450 | 28447

View Article Online
RSC Advances
Paper
ꢂ
O2 (ads) + eꢂ 4 O2
(2)
(3)
invertible characteristic at each cycle. In this gure, the
sensors were shown to display excellent stability and revers-
ibility in each cycle. In Fig. 8a, the pure WO3 sensor response
toward 100 ppm acetone vapor for each cycle were 29.472,
30.175, 28.08, 29.399, and 30.515, which is less than a ve
percent change in the response. However, Fig. 8b show that,
the 2 at% Pd-WO3 sensor responses toward 100 ppm acetone
vapor for each cycle were 107.29, 108.748, 108.502, 106.495
and 108.984, which is less than a two change. Furthermore,
the pure WO3 and 2 at% Pd-WO3 sensor stabilities were also
inspected for 1 month, as shown in Fig. 8a and b. The sensor
response value varied and uctuated within an extremely
narrow range, and deviations in the sensor response values
can be regulated lower than 5%. These results demonstrate
that the sensor have favorable stabilities when used to detect
acetone vapor.
O2 + eꢂ 4 2Oꢂ
ꢂ
ꢂ
CH3COCH3 (gas) + 3Oꢂ / CH3COCH2 + OHꢂ + eꢂ (4)
CH3COCH3 (gas) + OHꢂ / CH3CHO + CH3Oꢂ
CH3COCH3 (gas) + Oꢂ / CH3CO+ + CH3Oꢂ + eꢂ
CH3O+ / CH3+ + CO
(5)
(6)
(7)
(8)
CO + Oꢂ / CO2 + eꢂ
The main reasons for the improvement in the gas sensi-
tivity of the Pd-WO3 nanosheets are that the. Pd-WO3
contains a lamellar structure with a large specic surface
area, which is conducive to gas diffusion. According to the
catalytic spill-over effect, when exposed to air, Pd nano-
particles can dissociate O2 into Oꢂ, and then Oꢂ overow
adsorbs onto the WO3 surface.63 Then, Pd reduces the acti-
vation energy required for the reaction which shortens the
response–recovery time and improve the sensitivity. It forms
CH3COCH2ꢂ when exposed to acetone, which and diffuses to
the WO3 surface via the same process and enhances. The
sensitivity to acetone gas is in the doped sensor compared
with pure WO3. This phenomenon may occur due to the
removal of the depletion layer by a spill-over effect, which
improves the gas sensitive response. The presence of Pd
nanoparticles promotes the transfer and regeneration of the
electrons and oxygen species consumed in the acetone gas
reaction, and reduces the response recovery time of the
sensor. The fast response–recovery time is greatly ascribed
to the morphology and structure of materials, which provide
more surface-active sites to promote interaction with
acetone.64
Gas sensing mechanism
As is known, the sensing of n-type semiconductors occurs via
a the surface-controlled mechanism, which is determined by
the adjusting of the resistance produced by the desorption
and adsorption of acetone gas molecules on the surface of the
sensor.58,59 W6+ is an n-type semiconductor that acts as an
electron donor when exposed to air, where oxygen molecules
capture electrons from the WO3 conduction band and act as
electron acceptors. An electron depletion layer and chemi-
sorption of oxygen ions (O2ꢂ, Oꢂ and O2ꢂ) occur on the
surface of the sensing material, eqn (1)–(3).58,60,61 When
exposed to acetone, acetone molecules react with the adsor-
bed oxygenꢂions and are oxidized into intermediate products
CH3COCH2 and OHꢂ, eqn (4).62 Electrons are simulta-
neously released back to the surface of the material via eqn
(5)–(8),62 which reduces the resistance and width of the
depletion layer (Fig. 9).
O2 (gas) 4 O2 (ads)
(1)
Fig. 9 Schematic of sensing mechanisms of pure WO3 and Pd-WO3 sensor.
28448 | RSC Adv., 2019, 9, 28439–28450
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
12 Z. G. Xue, Z. X. Cheng, J. Xu, Q. Xiang, W. H. Wang and
J. Q. Xu, ACS Appl. Mater. Interfaces, 2017, 9, 41559–41567.
13 F. L. Meng, N. N. Hou, Z. Jin, B. Sun, W. Q. Li, X. H. Xiao,
C. Wang, M. Q. Li and J. H. Liu, Sens. Actuators, B, 2015,
219, 209–217.
Conclusions
In this paper, pure WO3 nanosheets and a series of Pd-WO3
nanosheets doped with different amounts of Pd were synthe-
sized through one-step hydrothermal method. The experi-
mental results showed that the gas sensor doped with 2 at% Pd
(Pd-WO3) had the best gas sensitivity to acetone and also had
a fast response, high sensitivity, high selectivity and low
detection limit. Therefore, it has potential application prospects
for acetone detection. The 2 at% Pd-WO3 gas sensor had a fast
response toward 100 ppm acetone with a sensitivity 3.64 times
higher than that of the pure WO3 sensor. Moreover, the sensor
has a low detection limit (50 ppb) and excellent selectivity to
acetone (Sacetone/Sethanol ¼ 5.06). The excellent gas sensitivity of
this material was mainly attributed to the high catalytic activity
and overow effect of the Pd nanoparticles, which provided
more active sites for this sensitive material.
¨
14 M. Bartsch and M. Niederberge, ChemPlusChem, 2017, 82,
42–59.
15 T. T. Zhou, T. Zhang, J. N. Deng, R. Zhang, Z. Luo and
L. L. Wang, Sens. Actuators, B, 2017, 242, 369–377.
16 W. T. Koo, S. Yu, S. J. Choi, J. S. Jang, J. Y. Cheong and
I. D. Kim, ACS Appl. Mater. Interfaces, 2017, 9, 8201–8210.
17 D. L. Chen, X. X. Hou, T. Li, L. Yin, B. B. Fan, H. L. Wang,
X. J. Li, H. L. Xu, H. X. Lu, R. Zhang and J. Sun, Sens.
Actuators, B, 2011, 153, 373–381.
18 M. ul Haq, Z. Wen, Z. Y. Zhang, S. Khan, Z. Lou, Z. Z. Ye and
L. P. Zhu, Sci. Rep., 2018, 8, 1705.
19 X. Zhang, X. Y. Li, H. C. Ye, S. M. Chen, H. X. Jiu, D. B. Liu,
Y. Lin, W. Ye, C. M. Wang and Q. Xu, J. Am. Chem. Soc., 2016,
138, 8928–8935.
Conflicts of interest
20 L. Xiao, S. M. Shu and S. T. Liu, Sens. Actuators, B, 2015, 221,
120–126.
21 H. Y. Liu, W. Yang, M. X. Wang, L. Xiao and S. T. Liu, Sens.
Actuators, B, 2016, 236, 490–498.
There no conicts to declare.
22 W. Yang, L. Feng, S. H. He, L. Y. Liu and S. T. Liu, ACS Appl.
Mater. Interfaces, 2018, 10, 27131–27140.
23 Z. Q. Hua, M. Yuasa, T. Kida, N. Yamazoe and K. Shimanoe,
Thin Solid Films, 2013, 548, 677–682.
24 P. Rai, J. W. Yoon, C. H. Kwak and J. H. Lee, J. Mater. Chem. A,
2016, 4, 264–269.
25 N. Ma, K. Suematsu, M. Yuasa, T. Kida and K. Shimanoe, ACS
Appl. Mater. Interfaces, 2015, 7, 5863–5869.
Acknowledgements
This work was nancially supported by the National Natural
Science Foundation of China (Grant No. 21471120) and the
International Cooperation Project of Hubei Province
(2012IHA00201), Educational Commission of Hubei Province of
China (T201306).
26 S. H. Xiao, B. Liu, R. Zhou, Z. W. Liu, Q. H. Li and
T. H. Wang, Sens. Actuators, B, 2018, 254, 966–972.
References
1 H. Haick, Y. Y. Broza, P. Mochalski, V. Ruzsanyi and 27 Y. Li, Z. Q. Hua, Y. Wu, Y. Zeng, Z. L. Qiu, X. M. Tian,
A. Amann, Chem. Soc. Rev., 2014, 43, 1423–1449.
M. J. Wang and E. P. Li, Sens. Actuators, B, 2018, 265, 249–
2 A. Amann, B. de Lacy Costello, W. Miekisch, J. Schubert,
256.
B. Buszewski, J. Pleil, N. Ratcliffe and T. Risby, J. Breath 28 F. Li, S. P. Ruan, N. Zhang, Y. Y. Yin, S. J. Guo, Y. Chen,
Res., 2014, 8, 034001.
3 M. Righettoni, A. Tricoli and S. E. Pratsinis, Anal. Chem.,
2010, 82, 3581–3587.
H. E. Zhang and C. N. Li, Sens. Actuators, B, 2018, 265,
355–364.
29 J. Morar, F. J. Himpsel, G. Hughes and J. L. Jordan, J. Vac. Sci.
Technol., A, 1985, 3, 1477–1480.
4 S. J. Choi, F. Fuchs, R. Demadrille, B. Grevin, B. H. Jang,
S. J. Lee, J. H. Lee, H. L. Tuller and I. D. Kim, ACS Appl. 30 Y. Liu, J. Yu and P. T. Lai, Int. J. Hydrogen Energy, 2014, 39,
Mater. Interfaces, 2014, 6, 9061–9070.
10313–10319.
5 B. Buszewski, A. Ulanowska, T. Ligor, N. Denderz and 31 K. F. Schoch, IEEE Electr. Insul. Mag., 1998, 14, 49.
A. Amann, Biomed. Chromatogr., 2009, 23, 551–556.
6 B. Ibrahim, M. Basanta, P. Cadden, D. Singh, D. Douce,
A. Woodcock and S. J. Fowler, Thorax, 2011, 66, 804–809.
7 Y. Y. Broza and H. Haick, Nanomedicine, 2013, 8, 785–806.
8 L. L. Wang, D. Chen, K. Jiang and G. Z. Shen, Chem. Soc. Rev.,
2017, 46, 6764–6815.
32 M. Ahsan, T. Tesfamichael, M. Ionescu, J. Bell and N. Motta,
Sens. Actuators, B, 2012, 162, 14–21.
33 J. Han, D. S. Zhang, P. Maitarad, L. Y. Shi, S. X. Cai, H. R. Li,
L. Huang and J. P. Zhang, Catal. Sci. Technol., 2014, 5, 1077–
1083.
34 P. Mach and M. Kocian, J. Phys. Chem. C, 2013, 117, 9999–
10006.
9 Y. H. Zhu, Y. Zhao, J. H. Ma, X. W. Cheng, J. Xie, P. C. Xu,
H. Q. Liu, H. P. Liu, H. J. Zhang and M. H. Wu, J. Am. 35 N. Chen, D. Y. Deng, Y. X. Li, X. Liu, X. X. Xing, X. C. Xiao and
Chem. Soc., 2017, 139, 10365–10373. Y. D. Wang, Sci. Rep., 2017, 7, 7692.
10 Y. B. Shen, H. S. Bi, T. T. Li, X. X. Zhong, X. X. Chen, A. F. Fan 36 M. Righettoni, A. Tricoli, S. Gass, A. Schmid, A. Amann and
and D. Z. Wei, Appl. Surf. Sci., 2018, 434, 922–931. S. E. Pratsinis, Anal. Chim. Acta, 2012, 738, 69–75.
11 J. J. Shi, Z. X. Cheng, L. P. Gao, Y. Zhang, J. Q. Xu and 37 L. Y. Liu, S. M. Shu, G. Z. Zhang and S. T. Liu, ACS Appl. Nano
H. B. Zhao, Sens. Actuators, B, 2016, 230, 736–745.
Mater., 2018, 1, 31–37.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 28439–28450 | 28449

View Article Online
RSC Advances
Paper
38 D. Wang, S. M. Huang, H. J. Li, A. Y. Chen, P. Wang, J. Yang, 52 X. Z. Song, L. Qiao, K. M. Sun, Z. Q. Tan, W. Ma, X. L. Kang,
X. Y. Wang and J. H. Yang, Sens. Actuators, B, 2019, 282, 961–
971.
F. F. Sun, T. Huang and X. F. Wang, Sens. Actuators, B, 2018,
256, 374–382.
39 H. L. Zhang, Z. F. Liu, J. Q. Yang, W. Guo, L. J. Zhu and 53 Y. X. Li, Z. Guo, Y. Su, X. B. Jin, X. H. Tang, J. R. Huang,
W. J. Zheng, Mater. Res. Bull., 2014, 57, 260–267.
40 S. Choi, J. W. Lee, W. S. Lee, C. Lee and W. I. Lee, J. Korean
Phys. Soc., 2017, 71, 487–493.
41 X. Bai, H. Ji, P. Gao, Y. Zhang and X. Sun, Sens. Actuators, B,
2014, 193, 100–106.
X. J. Huang, M. Q. Li and J. H. Liu, ACS Sens., 2016, 2, 102–
110.
54 J. F. Tan and X. T. Huang, Sens. Actuators, B, 2016, 237, 159–
166.
55 S. M. Wang, J. Cao, W. Cui, L. L. Fan, X. F. Li and D. J. Li, J.
Alloys Compd., 2018, 758, 45–53.
42 M. Yin, L. Yu and S. Liu, Mater. Lett., 2017, 186, 66–69.
43 C. H. Feng, C. Wang, P. F. Cheng, X. Li, B. Wang, Y. H. Guan, 56 M. Karmaoui, S. G. Leonardi, M. Latino, D. M. Tobaldi,
J. Ma, H. Zhang, Y. F. Sun and P. Sun, Sens. Actuators, B,
2015, 221, 434–442.
N. Donato, R. C. Pullar, M. P. Seabra, J. A. Labrincha and
G. Neri, Sens. Actuators, B, 2016, 230, 697–705.
44 L. Zhu, W. Zeng and Y. Li, Mater. Lett., 2018, 231, 5–7.
45 F. D. Qu, T. Thomas, B. X. Zhang, X. X. Zhou, S. D. Zhang,
57 H. Wang, L. Yan, S. Li, Y. Li, L. Liu, L. T. Du, H. J. Duan and
Y. L. Cheng, Appl. Phys. A, 2018, 124, 212.
S. P. Ruan and M. H. Yang, Sens. Actuators, B, 2018, 273, 58 N. Barsan, D. Koziej and W. Weimar, Sens. Actuators, B, 2007,
1202–1210. 121, 18–35.
46 J. N. Zhang, H. Lu, H. B. Lu, G. Li, J. Z. Gao, Z. B. Yang, 59 M. D'Arienzo, L. Armelao, C. M. Mari, S. Polizzi, R. Ruffo,
Y. H. Tian, M. Zhang, C. L. Wang and Z. He, J. Alloys
Compd., 2019, 779, 531–542.
R. Scotti and F. Morazzoni, J. Am. Chem. Soc., 2011, 133,
5296–5304.
47 S. Y. Liu, F. Zhang, H. Li, T. Chen and Y. D. Wang, Sens. 60 G. Eranna, B. C. Joshi, D. P. Runthala and R. P. Gupta, Crit.
Actuators, B, 2012, 162, 259–268. Rev. Solid State Mater. Sci., 2004, 29, 111–188.
48 S. Park, H. Kheel, G. J. Sun, T. Ko, W. I. Lee and C. Lee, J. 61 H. Geistlinger, Sens. Actuators, B, 1993, 17, 47–60.
Nanomater., 2015, 16, 398.
49 F. B. Gu, H. H. Chen, D. M. Han and Z. H. Wang, RSC Adv.,
2016, 6, 29727–29733.
50 H. Zhang, H. W. Qin, C. Y. Gao, G. J. Zhou, Y. P. Chen and
J. F. Hu, Sensors, 2018, 18, 1990.
51 X. S. Wang, Y. Li, L. Liu, L. Y. Wang, H. Wang and X. X. Guo,
J. Porous Mater., 2016, 24, 463–468.
62 S. B. Upadhyay, R. K. Mishra and P. P. Sahay, Sens. Actuators,
B, 2015, 209, 368–376.
63 B. X. Xiao, S. I. Song, P. Wang, Q. Zhao, M. Y. Chuai and
M. Z. Zhang, Sens. Actuators, B, 2017, 241, 489–497.
64 C. B. Zhai, Q. Zhao, K. K. Gu, D. J. Xing and M. Z. Zhang, J.
Alloys Compd., 2019, 784, 660–667.
28450 | RSC Adv., 2019, 9, 28439–28450
This journal is © The Royal Society of Chemistry 2019