- Email: [email protected]
Hierarchical porous hollow SnO2 nanofiber sensing electrode for high performance potentiometric H2 sensor
Accepted Manuscript Title: Hierarchical Porous Hollow SnO2 Nanofiber Sensing Electrode for High Performance Potentiometric H2 Sensor Authors: Jianxin Yi, Hong Zhang, Zuobin Zhang, Dongdong Chen PII: DOI: Reference:
S0925-4005(18)30790-1 https://doi.org/10.1016/j.snb.2018.04.086 SNB 24562
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
14-12-2017 23-3-2018 18-4-2018
Please cite this article as: Jianxin Yi, Hong Zhang, Zuobin Zhang, Dongdong Chen, Hierarchical Porous Hollow SnO2 Nanofiber Sensing Electrode for High Performance Potentiometric H2 Sensor, Sensors and Actuators B: Chemical https://doi.org/10.1016/j.snb.2018.04.086 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Hierarchical Porous Hollow SnO2 Nanofiber Sensing Electrode for High Performance Potentiometric H2 Sensor
IP T
Jianxin Yi*, Hong Zhang, Zuobin Zhang, Dongdong Chen
State Key Laboratory of Fire Science, Department of Safety Science and Engineering, University
SC R
of Science and Technology of China, Hefei, Anhui, 230026, P.R. China *
U
Corresponding author. E-mail address: [email protected] . Tel.: +86 551 63607817
N
Highlights
Mixed-potential sensor of SnO2 nanofiber sensing electrode was obtained.
The sensor showed high response, fast kinetics, and excellent selectivity to H2.
Performance was discussed based on gas diffusion and heterogeneous reaction.
Superior sensing was ascribed to the highly porous 3D hierarchical architecture.
Results highlight importance of morphology to mixed potential sensors.
CC E
PT
ED
M
A
ABSTRACT
A
A planar mixed-potential sensor was prepared based on SnO2 sensing electrode of hierarchical
porous
hollow
nanofibers.
The
electrode
was
featured
by
three-dimensional scaffold architecture with high porosity, large pore size, and excellent pore interconnectivity. A response value of -289.1 mV and response time of 5 s were achieved at 450 ºC for 1000 ppm H2, which were 5 and 2.6 times better than 1
those of a similar sensor with sensing electrode of SnO2 nanoparticles, respectively. The response values exhibited a linear or logarithmic dependence on the H2 concentration for below or above 140 ppm, respectively, corresponding to a change from diffusion- to reaction-controlled kinetics. Power-law concentration dependence
IP T
was observed for the response time, which also exhibited a similar transition at 140 ppm. Moreover, excellent selectivity, long-term stability, and repeatability were also
SC R
achieved. The morphology-dependent sensing behavior was discussed in terms of the
diffusion-reaction process and Thiele modulus. These results highlight the importance
U
of morphology to mixed-potential gas sensors, and show that hierarchical nanofiber
N
electrode is desirable for achieving high gas sensing performance.
A
KEYWORDS: Hierarchical nanostructure, mixed-potential, gas sensor, morphology,
ED
1. Introduction
M
diffusion-reaction.
PT
In recent years, gas sensors have been developed for gas detection in a diversity
CC E
of housing, industrial, and environmental applications, such as health and disease diagnosis, pollution monitoring, and leak and fire detection [1-3]. For the detection
A
of H2, which is not only an important raw material and a clean fuel but also highly flammable and explosive, gas sensors have to fulfill a number of requirements, such as high sensitivity, short response time, and low cross-sensitivity. Nevertheless, there is still a performance gap between the current sensors and these practical requirements [4]. Development of high performance H2 sensors is vitally needed. 2
Mixed potentiometric sensors have attracted great interest due to a number of features such as high temperature resistance, strong adaptability, low cost, simple structure, and real-time monitoring [5-7]. This kind of sensor generally consists of a solid electrolyte such as yttria-stabilized zirconia, a sensing electrode (SE), and a
IP T
reference electrode (RE). In the case of H2 detection, upon exposure of the sensor to H2-containing air, both O2 and H2 diffuse through the porous electrodes, and reach
SC R
the triple phase boundary (TPB), i.e., interface between gas, electrode, and solid
electrolyte. At the TPB, reactions of cathodic O2 reduction and anodic H2 oxidation
U
simultaneously take place. The mixed potential generated at the dissimilar electrodes
N
due to the different kinetics of these electrochemical reactions gives rise to an
A
electric potential difference, enabling sensing response to the analyte gas, H2 [8-10].
M
To achieve high sensing performance, it is essential that the anodic H2
ED
oxidation reaction is facilitated at the TPB of SE [6, 8]. The reaction rate is not only governed by the intrinsic activity of the SE material [10, 11], but also greatly
PT
dependent on the length [12] and the H2 concentration available at the TPB [13].
CC E
Large TPB length provides more active sites for the electrochemical reactions, and thus leads to high sensor response. Previous research has largely focused and made
A
great progress on the engineering of TPB. A variety of fabrication techniques, such as chemical corrosion of the electrolyte surface, double-tape casting, pore-forming and laser fabrication methods, have been developed to achieve enlarged TPB [7, 14-16]. On the other hand, part of H2 is consumed prior to reaching the TPB during gas diffusion through the electrode, due to the heterogeneous reaction of H2 3
oxidation [13, 17]. As a result, the heterogeneous reaction shall be depressed to achieve as high H2 concentration as possible at the TPB, which can be accomplished by using materials of low heterogeneous activity and by optimizing the electrode microstructure for fast gas transport.
IP T
Recently, hierarchical nanostructured materials have been widely studied in various fields due to the peculiar properties arising from the large surface/volume
SC R
ratio. Three-dimensional architectures constructed by these materials possess a number of merits, such as high porosity and fast gas transport. Such microstructural
U
engineering has been successfully employed to enhance the performance for
N
semiconductor gas sensors [18-20], solid oxide fuel cell cathodes [21, 22], and
A
catalysts [23], and may also be desirable for mixed potential gas sensors.
M
Nonetheless, so far there is very limited research on mixed potential sensors with
ED
nanostructured morphologies [24-27]. The response behavior and its relation with the hierarchical structure and gas transport have not yet been well understood.
PT
Herein, porous and hollow SnO2 nanofibers were used for the SE of a planar
CC E
mixed-potential sensor with a three-dimensional scaffold. To highlight the importance of electrode morphology and advantage of the hierarchical nanofibers, a
A
similar sensor based on SE of SnO2 nanoparticles was also prepared for comparison. Both sensors were based on the same solid electrolyte and used the same RE in order to confine their difference to only the SE morphology. The gas sensing characteristics of the sensors, especially the response sensitivity and sensing dynamics, were systematically examined at various temperatures and gas 4
concentrations. The sensing behavior and performance were discussed in relation to the electrode microstructure and the diffusion-reaction process. The sensor of nanofiber SE exhibited superior H2 sensing performance, which was attributed to the unique hierarchical morphology and electrode framework, resulting in faster gas
IP T
transport and higher H2 concentration at the TPB.
SC R
2. Experimental 2.1 Synthesis of nanofibers
U
All reagents were of analytical grade and purchased from Sinopharm Chemical
N
Reagent Co., Ltd., China, unless otherwise stated. SnO2 nanofibers were synthesized
A
via electrospinning on a home-built setup. Typically, 0.8 g SnCl2.2H2O was
M
dissolved in a mixture of 5.6 ml anhydrous ethanol and 4.7 ml N,
ED
N-dimethylformamide (DMF) under stirring. 0.8 g polyvinyl pyrrolidone (PVP, Mw=1.3106) was then added into the solution under stirring until a transparent
PT
solution was formed. The precursor was then transferred into a plastic syringe
CC E
mounted on a syringe pump (LSP01-1A). Electrospinning was conducted at a voltage of 15 kV at an injection rate of 0.4 ml/h. A grounded stainless steel foil was
A
used as the collector of the electrospun nanofibers. The as-spun nanofiber mats were then peeled off from the collector and dried at 80 C, and finally calcined at 600 ºC for 3 h to obtain SnO2 nanofibers.
2.2 Sensor fabrication and Characterization An YSZ (8 mol% yttria-stabilized zirconia, Hefei Kejing Co., Ltd., China) disk 5
of 13 mm×13 mm size and 0.25 mm thickness was used as the solid electrolyte. Two circular Pt pastes (Sino-Platinum Co., Ltd., China) were coated on both sides of the YSZ disk and sintered at 800 ºC for 10min, serving as RE and counter electrode (CE), respectively. Concerning the SE, two different pastes were prepared by mixing
IP T
and grinding the SnO2 nanofibers or SnO2 nanoparticles (Hefei Quantum Quelle Nano Sci. Tech. Co., Ltd., China) with an organic binder (α-terpineol and ethyl
SC R
cellulose). The obtained pastes were successively coated on the same side as the RE of the YSZ disk, and then sintered at 600 ºC for 3h. Each SE and RE had a diameter
U
of 2.2 mm, and was spaced 3.1 mm apart, while the circular CE had a larger
N
diameter of 5.8 mm. Pt wires were connected to the electrodes with some high
A
temperature silver paste (DAD-87, Shanghai Research Institute of Synthetic Resins,
M
China) as the current collector. The obtained sensors of nanofibers and nanoparticles,
ED
having different SE morphology but identical solid electrolyte and RE, were denoted as NFs sensor and NPs sensor, respectively.
PT
Phase structure of the samples was identified by X-ray powder diffraction
CC E
(XRD, Rigaku TTR-III) using Cu Kα radiation. Micro-structural analysis was conducted with scanning electron microscopy (FE-SEM, JEOL JSM-6700F)
A
operated at an accelerating voltage of 20 kV, and transmission electron microscopy (TEM, JEM-2010) operated at 200 kV. Specific surface area of nanofibers and nanoparticles was measured by Trista II 3020M analyzer. X-ray photoelectron spectroscopy (XPS) was performed on an ESCLAB 250 spectrometer using Al Kα as the exciting source, wherein the binding energy of C1s at 284.8 eV was used as a 6
reference for energy calibration. X-ray fluorescence spectrometry (XRF-1800, Shimadzu) was carried out for quantitative elemental analysis. 2.3 Sensor test The sensing characteristics of sensors were measured using a home-built sensor
IP T
test system, similar to that described previously [28]. Sample gas was obtained by mixing dry air and a stream of standard gas. Each standard gas contained one
SC R
analyte of H2, NO2, CH4, C3H8, CO, and NH3, and was balanced with N2. Gas flow rates were adjusted by the mass flow controllers (MFC, CS200, Beijing Sevenstar
U
Electronics, China), and the total flow rate was fixed at 200 ml/min. The open
N
circuit electric potential (V) of the sensors was measured with Agilent 34972A
A
electrometer. A two-electrode configuration was adopted for response measurements,
M
wherein SE and RE were connected to the positive and negative terminal of the
ED
electrometer, respectively. Response of the sensor was defined as the difference of electric potential in sample gas and in air, ΔV = Vgas – Vair. Response/recovery time
PT
of the sensor was defined as the time when the V reached 90% of the response upon
CC E
H2 feeding/removal. Polarization (current–voltage) curves were obtained by means of the potentiodynamic method (CHI604E, Chenhua, China) using a three-electrode
A
configuration under a scan rate of 0.5 mV/s.
3. Results and discussion 3.1 Materials Properties and Microstructure As shown in Figure 1, the crystal structure of SnO2 nanofibers (NFs) and 7
nanoparticles (NPs) are both indexed as a single-phase tetragonal rutile structure (JCPDS No.: 41-1445). Compared with the sharp and intense peaks for the nanofibers, the peaks were significantly broadened for the nanoparticles, suggesting smaller grain size. According to the Debye-Scherrer equation, the grain sizes of the
IP T
nanofibers and nanoparticles were estimated to be 14.4 nm and 4.6 nm, respectively. The surface elemental composition and chemical status of the as-prepared NFs and
SC R
NPs were further investigated by XPS, and the corresponding results are presented
in Figure 2. Regardless of the presence of carbon likely from the ambient, the
U
samples consisted of only Sn and O, and no impurity was detected. The two strong
N
Sn 3d peaks centered at 495.0 and 486.6 eV corresponded to the Sn 3d3/2 and Sn3d5/2,
A
respectively, indicating that Sn existed in the form of Sn4+ in both samples [29].
M
XRF analysis revealed 99.2 at.% Sn content on the metal basis for the NPs,
ED
confirming a high purity of the commercial powders. Figure 3 presents SEM and TEM images of SnO2 NFs, NPs, and the different
PT
sensing electrodes. It can be clearly seen that the SnO2 nanofibers were highly
CC E
porous with numerous pores on the wall of the fiber and had a hollow core (Figure 3a). The fibers were composed of nanoparticles of 10-20 nm, and had a typical
A
length of 0.5-2 m and a uniform outer diameter of ~120 nm. The size of the pores on the wall of the fiber was close to that of the particles. The highly porous hollow NFs were randomly but evenly stacked upon each other in the sintered 14.5μm-thick NFs SE, establishing a three-dimensional scaffold architecture with a high degree of submicron-sized void space among them (Figure 3b, 3c). In contrast to the NFs, the 8
pristine NPs in the powder form had much smaller grain sizes of ~2-5 nm. Loose agglomeration of the NPs into assemblies of size up to hundreds of nanometers was also observed, which contained numerous voids of several nanometers size among the particles. In the sintered NPs SE, secondary particles formed by aggregation of
IP T
the nanoparticles were found, the size of which varied significantly in a broad range from tens of nanometers up to ~1 m. As a result of the wide particle size
SC R
distribution, the NP SE appeared more compact with much lower porosity and generally much smaller pores when compared with the NFs SE, despite the presence
U
of some submicron pores among adjacent larger secondary particles. This
N
morphology is expected to significantly increase the resistance for gas transport and
A
lead to smaller gas diffusivity. The film thickness of the NPs SE was 17.2 μm, close
M
to that of the NFs SE. Furthermore, the BET surface area of the pristine NFs and
ED
NPs were determined to be 14.1 m2/g and 93.7 m2/g, respectively, agreeing well with the observed smaller particle size of the NPs.
PT
3.2 Sensing performance
CC E
Figures 4a-b compare the response (ΔV, in absolute values) as a function of H2 concentration for the NFs and NPs sensors at 450 °C and 550 °C, respectively. It can
A
be seen that the NFs sensor exhibited markedly higher response than the NPs sensor, which were 10 and 5 times higher for 100 ppm and 1000 ppm H2 at 450 °C, respectively. Furthermore, the response of the NFs sensor varied more pronouncedly with the H2 concentration especially for lower concentrations, indicating that it was more sensitive to the concentration changes (i.e., had larger sensitivity). 9
Figures 5a-b show more clearly the variation of response values with H2 concentration for the NFs sensor at 400-550 ºC. Clearly, the response increased, i.e., became more negative, with increasing H2 concentration and with decreasing temperature. High response of -371.1 mV and -289.1mV was observed at 400 ºC
IP T
and 450 ºC for 1000 ppm H2, respectively. In the lower H2 concentration range of 40-140 ppm, the response varied linearly with H2 concentration at 450-550 ºC
SC R
(Figure 5a). The sensitivity, which is represented by the slope of the linear fitting,
increased with decreasing temperature. In contrast, at a lower temperature of 400 ºC,
U
the sensor response exhibited logarithmic concentration dependence with a slope of
N
-419.6 mV/decade. Similar logarithmic dependence was observed in the higher H2
A
concentration range of 140-2000 ppm for all temperatures tested (Figure 5b). The
M
sensitivity (slope) was close to each other at 400-500 ºC, but became much smaller
ED
at a higher temperature, 550 ºC. The largest sensitivity of -158.6 mV/decade for 140-2000 ppm was obtained at 450 ºC. The different concentration dependence of
PT
the response values of NFs sensor is indicative of different sensing mechanisms
CC E
according to the mixed-potential theory [30, 31]. The linear variation of the response with H2 concentration suggests linear oxygen reduction kinetics and H2 oxidation
A
kinetics rate-limited by the diffusional mass transport of H2. With respect to the logarithmic concentration dependence, the reactions of anodic hydrogen oxidation and/or cathodic oxygen reduction most likely follow the Tafel-type kinetics. In order to verify the sensing mechanism, the polarization curves of the NFs sensor were measured at 450 ºC. As shown in Figure 6, the cathodic current 10
measured in air varied linearly with the applied potential. This behavior manifested that cathodic oxygen reduction followed the linear approximation of the Butler-Volmer equation at low overpotentials [8, 30], consistent with the linear concentration dependence of the response in Figure 5a. From the intersections
IP T
of the anodic plots with the cathodic one, the V can be estimated. The estimated V values for 60 ppm and 100 ppm hydrogen were -28 mV and -97 mV, respectively,
SC R
in good agreement with the corresponding measured response values of -29.1 mV and -98.2 mV in Figure 5a. These results confirmed that the operation of NFs sensor
U
followed the mixed-potential mechanism.
N
Figure 7 shows variation of response/recovery time for the NFs and NPs
A
sensors with H2 concentration at 450 ºC. The response time decreased quickly with
M
the increasing H2 concentration at low concentrations and slowly at high
ED
concentrations (Figure 7a). The NFs sensor exhibited distinctly smaller response time than the NPs sensor, which were 5 s and 13 s at 1000 ppm, respectively,
PT
corresponding to a 2.6-fold reduction. Similar behavior was also observed for the
CC E
recovery time, despite the much larger values (Figure 7b). The decrease of the recovery time with concentration was not so pronounced as the response time, and
A
became slow above 100 ppm. Again, the NFs sensor showed distinctly shorter recovery time than the NPs sensor. A value of 38 s was obtained for the NFs sensor at 100 ppm, 2.5 times shorter than that of the NPs sensor. As shown in Figure 8, the response time of the NFs sensor decreased with increasing temperature or with increasing H2 concentration. Linear relation was 11
observed in the log-log plots of response time versus H2 concentration at various temperatures, indicating power-law dependence. A change of the dependence occurred at 140 ppm H2 in the temperature range of 450-550 ºC. The exponents for the higher concentration range varied only slightly within -0.8~-1.0, whereas smaller
IP T
exponents of -0.3~-0.5 were found for lower concentrations below 140 ppm H2. In contrast, at the lower temperature, 400 ºC, the change of the concentration
SC R
dependence was not observed, and the same power-law relation held for the entire
concentration range of 40-2000 ppm with a single exponent of -0.8. The
U
concentration dependence of the response time highly resembled that for the
N
response values, which may be ascribed to the different sensing mechanisms at low
A
and high H2 concentrations.
M
Figure 9 displays the cross-sensitivities of the sensors to 100 ppm various gases
ED
at 450 ºC. The NFs sensor was most responsive to H2, and also exhibited a minor response to NH3 and CO. Negligible response to the rest of the interfering gases,
PT
CH4, C3H8, and NO2, was observed. The response to the most interfering gas, CO,
CC E
was only -13.3 mV, over 7 times smaller than that for H2. By contrast, the NPs sensor was confronted with strong interference from NH3, the response to which was
A
comparable to that for H2. In addition, a modest response to CO with a reduced value of -3.2 mV was also found. Therefore, more selective H2 sensing was achieved for the NFs sensor. For practical applications, stable and repeatable operation of gas sensors is of crucial importance. Figure 10a presents the response values of NFs sensor for 100 12
ppm, 200 ppm, and 300 ppm H2 during a 33-day measurement at 450 ºC. One can see clearly that the values remained almost constant within this period, demonstrating excellent long-time stability of the NFs sensor. Figure 10b shows that during seven successive runs at 450 ºC in 100 ppm H2, both the baseline and
IP T
response transients of the NFs sensor were stable and reproduced very well. Figure 10c compares the response values of two different NFs sensors at 450 ºC.
SC R
Apparently they were very close to each other at each concentration tested and a
discrepancy within 10% was found. Therefore, the sensing characteristics of the NFs
U
sensor can be well repeated not only for a single sample, but also between different
N
sensors.
A
3.3 Relation of sensing performance and diffusion-reaction process
M
The above results showed evidently that electrode morphology plays an
ED
important role in determining the gas sensing performance of SnO2/YSZ/Pt mixed potential sensor. The NFs sensor of porous hollow nanofibers exhibited much superior
PT
H2 sensing performance.
CC E
It has been known that heterogeneous reaction takes place on the SE surface during diffusion of hydrogen through the porous SE:
H 2 (g) O2 (g) H 2O( g )
(1)
A
At the TPB, cathodic oxygen reduction reaction and anodic reaction take place simultaneously to generate the mixed potential: O2 (g) 4e 2O2
(2)
H 2 (g) O2 H 2O( g ) 2e
(3) 13
A good SE should have high electrochemical activity for the anodic reaction (3), for which the reaction rate is a function of both the reaction rate constant and concentration, e.g. R=k’Ceff for a first-order reaction. Note that Ceff is the actual (effective) concentration at TPB, which is generally smaller than the starting
IP T
concentration, C0, due to reaction (1). Reaction (1) inside the SE consumes H2 and is unfavorable to achieving high response values. In particular, insufficient H2 would
SC R
reach the TPB when the H2 concentration is low, giving rise to diffusion-controlled kinetics. The observed linear concentration dependence of the sensor response for
U
below 140 ppm at 450-550 ºC in Figure 5a agrees well with this mechanism.
N
For a planar thick film electrode used in this work, a concentration profile is
A
established inside the electrode due to reaction (1) according to the heterogeneous
M
catalysis theory [32]. Assuming a first-order reaction kinetics, the H2 concentration
C0 cosh(( L x) k / D )
(4)
cosh( L k / D )
PT
C ( x)
ED
under steady state can be calculated by:
where L is the film thickness, x the distance from the electrode surface into the
CC E
electrode, and D gas diffusion coefficient, k rate constant for reaction (1). For mixed potential sensors, only H2 that reaches the TPB is effective for generating the mixed
A
potential and the sensor signals. This effective H2 concentration can be obtained from Equation (4) at x=L: Ceff C ( L)
C0
(5)
cosh( L k / D )
Equation (5) can be rewritten to obtain the normalized effective concentration at TPB, 14
c
Ceff C0
1 cosh( )
(6)
where
L k/D
(7)
IP T
is the so called Thiele modulus for isothermal first-order reaction, which has been widely used to represent the relative importance of diffusion and reaction rates in
SC R
porous catalysts. In principle, a larger value of Thiele modulus accords with relatively slow gas diffusion. Figure 11 shows variation of the normalized effective concentration c at TPB as a function of . Specifically, at <0.4, the concentration
U
reduction is insignificant with c>0.92, and gas diffusion is fast. As value increases,
M
primarily rate-limited by gas diffusion.
A
N
c decreases significantly. At >3, c is smaller than 0.1, indicating that the process is
According to Equation (7), small Thiele modulus value can be achieved by
ED
reducing the electrode thickness, improving gas transport, and using materials of low
PT
activity for reaction (1). Both k and D are known to strongly correlate with the materials morphology and electrode microstructure. Concerning the sensors
CC E
investigated in the present work, their only difference is the materials morphology and electrode microstructure of the SE, since they shared the same solid electrolyte and
A
RE and had similar SE thickness. As schematically illustrated in Figure 12a, the NFs SE was characterized by a 3D scaffold architecture formed by random stacking of highly porous and hollow nanofibers upon each other. The macropores among the nanofibers and those along the hollow axis appeared to dominate the overall porosity, through which gas diffusion was expected to proceed via the fast molecular diffusion. 15
Moreover, these macropores were highly inter-connected, partly through the numerous pores on the fiber walls, allowing for more facile gas transport. The high porosity, large pores, high degree of pore interconnectivity, and low degree of channel tortuosity are highly beneficial to achieving large gas diffusivity. By great contrast, in
IP T
the NPs SE (Figure 12b), nano-sized pores within and among the secondary particles dominated the overall porosity, and thus Knudsen diffusion is expected to play an
SC R
important role, which is much slower than the molecular diffusion. In such structure,
poor pore interconnectivity and tortuous gas transport pathway would be expected in
U
addition to the observed compactness. These would lead to poorer gas transport in the
N
NPs SE. On the other hand, the smaller particle size and much larger BET surface
A
area of the nanoparticles were in principle associated with higher surface reactivity,
M
and thus resulted in larger k. Therefore, the much superior sensing response for the
ED
NFs sensor relative to the NPs sensor as observed in Figure 4 can be accounted for by its smaller (larger D and smaller k), resulting from the unique 3D hierarchical
PT
electrode microstructure.
CC E
For an irreversible reaction of order n (n<1) the generalized Thiele modulus is used [32]:
n 1 k n 1 C0 2 D
(8)
A
L
When n=1, Equation (8) becomes Equation (7) for first-order reaction, and is independent of concentration. For a reaction of order -1
diffusion- or reaction-controlled behavior of the response values for NFs sensor at 450-550 ºC in Figure 5. That only reaction-controlled behavior was observed in the whole concentration range at the lower temperature of 400 ºC can be interpreted by the temperature dependence of . As k decreases faster than D with temperature
IP T
decreasing, reaction rate becomes slower than gas diffusion and dominates the process at lower temperatures. The two-segmental behavior of the response value at 400 ºC
SC R
(Figure 5b) is also consistent with the variation of with concentration.
Similar to the case of response value, the diffusion-reaction process in the SE
U
may also exert some impact on the response kinetics. Nevertheless, this important
N
aspect has received little attention and has not been addressed by the existing theory
A
of mixed-potential gas sensors [8, 30]. In the present work, the response time for the
M
SnO2 NFs sensor exhibited power-law dependence on H2 concentration. Similar
ED
phenomenon has also been observed for semiconductor gas sensors [33, 34], which was explained by a non-linear diffusion-reaction model developed based on
PT
Freundlich adsorption isotherm. This model could in principle be extended to
CC E
mixed-potential gas sensors, given the same diffusion-reaction process in their porous electrode as in the semiconductor sensors. According to the model, a modified response time is defined as [34]
A
L2
k m1 C0 D
(9)
where m is a constant. An m value close to 1 suggests a diffusion-controlled process, whilst reaction-controlled process dominates as m approaches zero. The m values of the NFs sensor at 450-550 ºC were 0.5~0.7 and 0~0.2 (Figure 7), corresponding to 17
diffusion- and reaction-controlled kinetics for below and above 140 ppm H2, respectively. At 400 ºC, no transition was observed, and the m value of 0.2 indicated reaction-controlled kinetics for the entire concentration range. This behavior of the response kinetics is in good accordance with that of the response values in Figure 5
IP T
and as discussed above.
is in fact related with the generalized Thiele modulus:
2 2 mn C0 n 1
SC R
If Eqs. (8) and (9) are combined, one can immediately see that the response time
(10)
U
Therefore, the magnitude of response time is largely governed by , and small may
N
accord with fast response kinetics. The much faster response kinetics of the NFs
A
sensor relative to the NPs sensor in Figure 6a can thus be ascribed again to the faster
M
gas transport and smaller k in the 3D nanofiber SE.
ED
The sensing performance of mixed-potential gas sensors is jointly governed by the kinetics of the electrochemical and heterogeneous reactions. As the kinetics of
PT
reactions (1) and (3) differ for various gases, different response characteristics would
CC E
be expected, accounting for the sensor selectivity. The high H2 response for the NFs sensor suggests that SnO2 has a high activity for the electrochemical H2 oxidation reaction (3) while the heterogeneous H2 oxidation reaction (1) did not significantly
A
take place. The greatly reduced H2 and CO response for the NPs sensor can be ascribed to the slow gas transport in the electrode, which facilitated the heterogeneous reaction (1). The observation in Figure 9 that the response for the other gases was quite small and did not significantly vary with the electrode morphology suggests that 18
most likely both the electrochemical and heterogeneous activities to these gases are low for SnO2. These results also show that the sensor selectivity can be tuned by manipulating the electrode microstructure. The present work showed that electrode morphology was of crucial importance to
IP T
the gas sensing performance of SnO2/YSZ/Pt mixed potential sensor. Superior sensing performance, particularly high sensitivity and fast response kinetics, could be
SC R
achieved based on hierarchical porous hollow nanofibers, benefiting from the fast gas transport and reduced heterogeneous reaction. Note that although this work was
U
mainly focused on H2 sensing, the results obtained herein could in principle be
A
N
extended to other mixed-potential gas sensors for different gases.
M
4. Conclusion
ED
A high performance planar mixed-potential H2 sensor was obtained based on hierarchical porous hollow SnO2 nanofiber sensing electrode. A high response value
PT
of -289.1 mV and short response time of 5 s at 450 ºC for 1000 ppm H2 were obtained.
CC E
The response values were linearly or logarithmically dependent on H2 concentration for low and high concentrations, respectively, indicating a change from diffusion- to reaction-controlled kinetics. A power-law relationship between the response time and
A
concentration was observed, which also exhibited a similar transition at 140 ppm. Moreover, excellent selectivity to H2 and repeatable and stable operation were achieved. The sensing performance was highly dependent on the electrode morphology, which closely correlated with the diffusion-reaction process and Thiele 19
modulus. The highly porous three-dimensional electrode architecture of nanofibers favors small Thiele modulus and is advantageous for mixed potential gas sensors.
Acknowledgements
IP T
Finance support by the Natural Science Foundation of China (No. U1432108), and the Fundamental Research Funds for the Central Universities (No. WK2320000034) is
A
CC E
PT
ED
M
A
N
U
SC R
gratefully acknowledged.
20
References [1] S. Bagchi, S. Sengupta, S. Mandal, Development and Characterization of Carbonic Anhydrase Based CO2 Biosensor for Primary Diagnosis of Respiratory Health, Sensors, 15423 (2016) 1. [2] G. Gregis, J.-B. Sanchez, I. Bezverkhyy, W. Guy, F. Berger, V. Fierro, et al.,
IP T
Detection and quantification of lung cancer biomarkers by a micro-analytical device using a single metal oxide-based gas sensor, Sensors and Actuators B: Chemical, 255 (2018) 391-400.
SC R
[3] R. Baron, J. Saffell, Amperometric Gas Sensors as a Low Cost Emerging
Technology Platform for Air Quality Monitoring Applications: A Review, ACS sensors, 2 (2017) 1553-1566.
U
[4] L. Boon-Brett, J. Bousek, G. Black, P. Moretto, P. Castello, T. Hübert, et al.,
N
Identifying performance gaps in hydrogen safety sensor technology for automotive
A
and stationary applications, International Journal of Hydrogen Energy, 35 (2010)
M
373-84.
[5] J.W. Fergus, A review of electrolyte and electrode materials for high temperature
(2008) 1034-1041.
ED
electrochemical CO2 and SO2 gas sensors, Sensors and Actuators B: Chemical, 134
PT
[6] S.A. Anggraini, M. Breedon, H. Ikeda, N. Miura, Insight into the aging effect on enhancement of hydrogen-sensing characteristics of a zirconia-based sensor utilizing
CC E
a Zn–Ta–O-based sensing electrode, ACS applied materials & interfaces, 5 (2013) 12099-12106.
[7] Y. Guan, C. Li, X. Cheng, B. Wang, R. Sun, X. Liang, et al., Highly sensitive
A
mixed-potential-type NO2 sensor with YSZ processed using femtosecond laser direct writing technology, Sensors and Actuators B: Chemical, 198 (2014) 110-113. [8] N. Miura, T. Sato, S.A. Anggraini, H. Ikeda, S. Zhuiykov, A review of mixed-potential type zirconia-based gas sensors, Ionics, 20 (2014) 901-925. [9] S.A. Anggraini, M. Breedon, N. Miura, Zn–Ta-based oxide as a hydrogen sensitive electrode material for zirconia-based electrochemical gas sensors, Sensors 21
and Actuators B: Chemical, 187 (2013) 58-64. [10] Y. Li, X. Li, J. Wang, Y. Jun, Z. Tang, A cobalt tungstate compound sensing electrode for hydrogen detection based upon mixed-potential type sensors, RSC Advances, 7 (2017) 2919-2925. [11] Z. Tang, X. Li, J. Yang, J. Yu, J. Wang, Z. Tang, Mixed potential hydrogen sensor using ZnWO4 sensing electrode, Sensors and Actuators B: Chemical, 195 (2014)
IP T
520-525.
[12] X. Liang, S. Yang, J. Li, H. Zhang, Q. Diao, W. Zhao, et al., Mixed-potential-type
SC R
zirconia-based NO2 sensor with high-performance three-phase boundary, Sensors and Actuators B: Chemical, 158 (2011) 1-8.
[13] L.P. Martin, A.Q. Pham, R.S. Glass, Effect of Cr2O3 electrode morphology on the
U
nitric oxide response of a stabilized zirconia sensor, Sensors and Actuators B:
N
Chemical, 96 (2003) 53-60.
A
[14] B. Wang, F. Liu, X. Yang, Y. Guan, C. Ma, X. Hao, et al., Fabrication of well-ordered three-phase boundary with nanostructure pore array for mixed
M
potential-type zirconia-based NO2 sensor, ACS applied materials & interfaces, 8
ED
(2016) 16752-16760.
[15] R. Sun, Y. Guan, X. Cheng, Y. Guan, X. Liang, J. Ma, et al., High performance three-phase boundary obtained by sand blasting technology for mixed-potential-type
PT
zirconia-based NO2 sensors, Sensors and Actuators B: Chemical, 210 (2015) 91-95. [16] X. Cheng, C. Wang, B. Wang, R. Sun, Y. Guan, Y. Sun, et al.,
CC E
Mixed-potential-type YSZ-based sensor with nano-structured NiO and porous TPB processed with pore-formers using coating technique, Sensors and Actuators B: Chemical, 221 (2015) 1321-1329.
A
[17] M. Yamaguchi, S.A. Anggraini, Y. Fujio, T. Sato, M. Breedon, N. Miura, Stabilized zirconia-based sensor utilizing SnO2-based sensing electrode with an integrated Cr2O3 catalyst layer for sensitive and selective detection of hydrogen, International Journal of Hydrogen Energy, 38 (2013) 305-312. [18] L. Zhang, Y. Yin, Hierarchically mesoporous SnO2 nanosheets: hydrothermal synthesis and highly ethanol-sensitive properties operated at low temperature, Sensors 22
and Actuators B: Chemical, 185 (2013) 594-601. [19] T. Yu, X. Cheng, X. Zhang, L. Sui, Y. Xu, S. Gao, et al., Highly sensitive H2S detection sensors at low temperature based on hierarchically structured NiO porous nanowall arrays, Journal of Materials Chemistry A, 3 (2015) 11991-11999. [20] Q. Wang, X. Li, F. Liu, Y. Sun, C. Wang, X. Li, et al., Three-dimensional flake-flower Co/Sn oxide composite and its excellent ethanol sensing properties,
IP T
Sensors and Actuators B: Chemical, 230 (2016) 17-24.
[21] M. Zhi, S. Lee, N. Miller, N.H. Menzler, N. Wu, An intermediate-temperature
SC R
solid oxide fuel cell with electrospun nanofiber cathode, Energy & Environmental Science, 5 (2012) 7066-7071.
[22] Y. Jeon, J.-h. Myung, S.-h. Hyun, Y.-g. Shul, J.T. Irvine, Corn-cob like nanofibres
U
as cathode catalysts for an effective microstructure design in solid oxide fuel cells,
N
Journal of Materials Chemistry A, 5 (2017) 3966-3973.
[23] Z. Wang, M. Li, C. Liang, L. Fan, J. Han, Y. Xiong, Effect of morphology on the
A
oxygen evolution reaction for La0.8Sr0.2Co0.2Fe0.8O3−δelectrochemical catalyst in
M
alkaline media, RSC Adv, 6 (2016) 69251-69256.
ED
[24] L. Dai, Y. Liu, W. Meng, G. Yang, H. Zhou, Z. He, et al., Ammonia sensing characteristics of La10Si5MgO26 -based sensors using In2O3 sensing electrode with different morphologies and CuO reference electrode, Sensors and Actuators B:
PT
Chemical, 228 (2016) 716-724.
[25] Y. Liu, L. Dai, H. Hou, Y. Li, W. Meng, L. Wang, Mixed-potential type NO2
CC E
sensor based on La10Si6O27 electrolyte and WO3 sensing electrode with different morphologies, Ceramics International, 42 (2016) 9712-9716. [26] W. Meng, L. Dai, J. Zhu, Y. Li, W. Meng, H. Zhou, et al., A novel mixed potential
A
NH3 sensor based on TiO2@WO3 core–shell composite sensing electrode, Electrochimica Acta, 193 (2016) 302-310. [27] F. Liu, Y. Guan, H. Sun, X. Xu, R. Sun, X. Liang, et al., YSZ-based NO2 sensor utilizing hierarchical In2O3 electrode, Sensors and Actuators B: Chemical, 222 (2016) 698-706. [28] H. Zhang, J. Yi, X. Jiang, Fast Response, Highly Sensitive and Selective 23
Mixed-Potential H2 Sensor Based on (La, Sr)(Cr, Fe) O3-δ Perovskite Sensing Electrode, ACS Applied Materials & Interfaces, 9 (2017) 17218-17225. [29] Y.-H. Choi, S.-H. Hong, H2 sensing properties in highly oriented SnO2 thin films, Sensors and Actuators B: Chemical, 125 (2007) 504-509. [30] F.H. Garzon, R. Mukundan, E.L. Brosha, Solid-state mixed potential gas sensors: theory, experiments and challenges, Solid State Ionics, 136 (2000) 633-638.
IP T
[31] S.A. Anggraini, M. Breedon, N. Miura, Sensing characteristics of aged
zirconia-based hydrogen sensor utilizing Zn–Ta-based oxide sensing-electrode,
SC R
Electrochemistry Communications, 31 (2013) 133-136.
[32] M.E. Davis, R.J. Davis, Fundamentals of chemical reaction engineering, Courier Corporation, 2012.
U
[33] S.J. Choi, S. Chattopadhyay, J.J. Kim, S.J. Kim, H.L. Tuller, G.C. Rutledge, et
N
al., Coaxial electrospinning of WO3 nanotubes functionalized with bio-inspired Pd
A
catalysts and their superior hydrogen sensing performance, Nanoscale, 8 (2016) 9159-9166.
M
[34] J.W. Gardner, A non-linear diffusion-reaction model of electrical conduction in
A
CC E
PT
ED
semiconductor gas sensors, Sensors and Actuators B: Chemical, 1 (1990) 166-170.
24
Author biographies Jianxin Yi is an associate professor of State Key Laboratory of Fire Science, University of Science and Technology of China (USTC). He received both his B.S. degree in Chemical Physics and PhD degree in Materials Science from USTC. He
IP T
then worked as a postdoctoral researcher at University of Twente and at RWTH Aachen University during 2006-2011. Before joining USTC in 2012, he worked as an
SC R
associate professor of Materials science at Huazhong University of Science and
Technology. His current research is focused on gas sensors, functional nanostructured
N
U
materials, and mixed-conducting oxides.
A
Hong Zhang is a master student at University of Science and Technology of China.
M
Her current research is on mixed potential gas sensors of nanostructured sensing
ED
electrode.
PT
Zuobin Zhang is a PhD degree candidate at University of Science and Technology of
CC E
China. His research is focused on gas sensors based on mixed potential theory.
Dongdong Chen is a master student at University of Science and Technology of
A
China. His research is mainly focused on semiconductor gas sensor.
25
< Figure Captions> Figure 1. XRD patterns of SnO2 nanofibers and nanoparticles. Figure 2. (a) Survey XPS spectrum and (b) high resolution spectra for Sn 3d of the as-prepared SnO2 nanofibers and nanoparticles.
SC R
IP T
Figure 3. (a-c) SEM images for (a) pristine SnO2 nanofibers and (b) surface and (c) cross-section of the NFs SE; inset in (a) is the TEM image of a nanofiber. (d) TEM image of pristine SnO2 nanoparticles, (e, f) SEM images for (e) surface and (f) cross-section of NPs SE. Figure 4. Response of NFs sensor and NPs sensor as a function of hydrogen concentration at (a) 450°C and (b) 550°C. Absolute values are used for the sake of comparison.
N
U
Figure 5. Response values of NFs sensor as a function of (a) hydrogen concentration and (b) logarithmic concentrations.
A
Figure 6. Polarization curves for the NFs sensor at 450 °C.
M
Figure 7. Variation of (a) response time and (b) recovery time for NFs sensor and NPs sensor with hydrogen concentration at 450°C.
ED
Figure 8. Log-log plots of response time of NFs sensor versus hydrogen concentration at 450 °C and 550 °C. Inset shows the data at 400 °C.
PT
Figure 9. Cross-sensitivities of NFs sensor and NP sensor to 100ppm various gases at 450°C.
CC E
Figure10. (a) Long-term stability, (b) continuous dynamic response curves in 100 ppm hydrogen, and (c) response comparison of two different samples at 450°C for NFs sensor.
A
Figure 11. Variation of the normalized effective concentration c as a function of . Figure12. Schematic illustrations for the gas transport in electrodes of different morphology. (a) NFs SE of nanofibers, (b) NPs SE of nanoparticles.
26
IP T SC R
A
CC E
Figure 2.
PT
ED
M
A
N
U
Figure 1.
27
IP T SC R
CC E
PT
ED
M
A
N
U
Figure 3.
A
Figure 4.
28
IP T
A
CC E
Figure 6.
PT
ED
M
A
N
U
SC R
Figure 5.
29
IP T
CC E
PT
ED
M
A
N
U
SC R
Figure 7.
A
Figure 8.
30
A ED
PT
CC E
IP T
SC R
U
N
A
M
Figure 9.
31
A ED
PT
CC E
Figure10.
32
IP T
SC R
U
N
A
M
IP T
A
CC E
PT
ED
M
A
N
U
SC R
Figure 11.
Figure12.
33
S0925-4005(18)30790-1 https://doi.org/10.1016/j.snb.2018.04.086 SNB 24562
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
14-12-2017 23-3-2018 18-4-2018
Please cite this article as: Jianxin Yi, Hong Zhang, Zuobin Zhang, Dongdong Chen, Hierarchical Porous Hollow SnO2 Nanofiber Sensing Electrode for High Performance Potentiometric H2 Sensor, Sensors and Actuators B: Chemical https://doi.org/10.1016/j.snb.2018.04.086 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Hierarchical Porous Hollow SnO2 Nanofiber Sensing Electrode for High Performance Potentiometric H2 Sensor
IP T
Jianxin Yi*, Hong Zhang, Zuobin Zhang, Dongdong Chen
State Key Laboratory of Fire Science, Department of Safety Science and Engineering, University
SC R
of Science and Technology of China, Hefei, Anhui, 230026, P.R. China *
U
Corresponding author. E-mail address: [email protected] . Tel.: +86 551 63607817
N
Highlights
Mixed-potential sensor of SnO2 nanofiber sensing electrode was obtained.
The sensor showed high response, fast kinetics, and excellent selectivity to H2.
Performance was discussed based on gas diffusion and heterogeneous reaction.
Superior sensing was ascribed to the highly porous 3D hierarchical architecture.
Results highlight importance of morphology to mixed potential sensors.
CC E
PT
ED
M
A
ABSTRACT
A
A planar mixed-potential sensor was prepared based on SnO2 sensing electrode of hierarchical
porous
hollow
nanofibers.
The
electrode
was
featured
by
three-dimensional scaffold architecture with high porosity, large pore size, and excellent pore interconnectivity. A response value of -289.1 mV and response time of 5 s were achieved at 450 ºC for 1000 ppm H2, which were 5 and 2.6 times better than 1
those of a similar sensor with sensing electrode of SnO2 nanoparticles, respectively. The response values exhibited a linear or logarithmic dependence on the H2 concentration for below or above 140 ppm, respectively, corresponding to a change from diffusion- to reaction-controlled kinetics. Power-law concentration dependence
IP T
was observed for the response time, which also exhibited a similar transition at 140 ppm. Moreover, excellent selectivity, long-term stability, and repeatability were also
SC R
achieved. The morphology-dependent sensing behavior was discussed in terms of the
diffusion-reaction process and Thiele modulus. These results highlight the importance
U
of morphology to mixed-potential gas sensors, and show that hierarchical nanofiber
N
electrode is desirable for achieving high gas sensing performance.
A
KEYWORDS: Hierarchical nanostructure, mixed-potential, gas sensor, morphology,
ED
1. Introduction
M
diffusion-reaction.
PT
In recent years, gas sensors have been developed for gas detection in a diversity
CC E
of housing, industrial, and environmental applications, such as health and disease diagnosis, pollution monitoring, and leak and fire detection [1-3]. For the detection
A
of H2, which is not only an important raw material and a clean fuel but also highly flammable and explosive, gas sensors have to fulfill a number of requirements, such as high sensitivity, short response time, and low cross-sensitivity. Nevertheless, there is still a performance gap between the current sensors and these practical requirements [4]. Development of high performance H2 sensors is vitally needed. 2
Mixed potentiometric sensors have attracted great interest due to a number of features such as high temperature resistance, strong adaptability, low cost, simple structure, and real-time monitoring [5-7]. This kind of sensor generally consists of a solid electrolyte such as yttria-stabilized zirconia, a sensing electrode (SE), and a
IP T
reference electrode (RE). In the case of H2 detection, upon exposure of the sensor to H2-containing air, both O2 and H2 diffuse through the porous electrodes, and reach
SC R
the triple phase boundary (TPB), i.e., interface between gas, electrode, and solid
electrolyte. At the TPB, reactions of cathodic O2 reduction and anodic H2 oxidation
U
simultaneously take place. The mixed potential generated at the dissimilar electrodes
N
due to the different kinetics of these electrochemical reactions gives rise to an
A
electric potential difference, enabling sensing response to the analyte gas, H2 [8-10].
M
To achieve high sensing performance, it is essential that the anodic H2
ED
oxidation reaction is facilitated at the TPB of SE [6, 8]. The reaction rate is not only governed by the intrinsic activity of the SE material [10, 11], but also greatly
PT
dependent on the length [12] and the H2 concentration available at the TPB [13].
CC E
Large TPB length provides more active sites for the electrochemical reactions, and thus leads to high sensor response. Previous research has largely focused and made
A
great progress on the engineering of TPB. A variety of fabrication techniques, such as chemical corrosion of the electrolyte surface, double-tape casting, pore-forming and laser fabrication methods, have been developed to achieve enlarged TPB [7, 14-16]. On the other hand, part of H2 is consumed prior to reaching the TPB during gas diffusion through the electrode, due to the heterogeneous reaction of H2 3
oxidation [13, 17]. As a result, the heterogeneous reaction shall be depressed to achieve as high H2 concentration as possible at the TPB, which can be accomplished by using materials of low heterogeneous activity and by optimizing the electrode microstructure for fast gas transport.
IP T
Recently, hierarchical nanostructured materials have been widely studied in various fields due to the peculiar properties arising from the large surface/volume
SC R
ratio. Three-dimensional architectures constructed by these materials possess a number of merits, such as high porosity and fast gas transport. Such microstructural
U
engineering has been successfully employed to enhance the performance for
N
semiconductor gas sensors [18-20], solid oxide fuel cell cathodes [21, 22], and
A
catalysts [23], and may also be desirable for mixed potential gas sensors.
M
Nonetheless, so far there is very limited research on mixed potential sensors with
ED
nanostructured morphologies [24-27]. The response behavior and its relation with the hierarchical structure and gas transport have not yet been well understood.
PT
Herein, porous and hollow SnO2 nanofibers were used for the SE of a planar
CC E
mixed-potential sensor with a three-dimensional scaffold. To highlight the importance of electrode morphology and advantage of the hierarchical nanofibers, a
A
similar sensor based on SE of SnO2 nanoparticles was also prepared for comparison. Both sensors were based on the same solid electrolyte and used the same RE in order to confine their difference to only the SE morphology. The gas sensing characteristics of the sensors, especially the response sensitivity and sensing dynamics, were systematically examined at various temperatures and gas 4
concentrations. The sensing behavior and performance were discussed in relation to the electrode microstructure and the diffusion-reaction process. The sensor of nanofiber SE exhibited superior H2 sensing performance, which was attributed to the unique hierarchical morphology and electrode framework, resulting in faster gas
IP T
transport and higher H2 concentration at the TPB.
SC R
2. Experimental 2.1 Synthesis of nanofibers
U
All reagents were of analytical grade and purchased from Sinopharm Chemical
N
Reagent Co., Ltd., China, unless otherwise stated. SnO2 nanofibers were synthesized
A
via electrospinning on a home-built setup. Typically, 0.8 g SnCl2.2H2O was
M
dissolved in a mixture of 5.6 ml anhydrous ethanol and 4.7 ml N,
ED
N-dimethylformamide (DMF) under stirring. 0.8 g polyvinyl pyrrolidone (PVP, Mw=1.3106) was then added into the solution under stirring until a transparent
PT
solution was formed. The precursor was then transferred into a plastic syringe
CC E
mounted on a syringe pump (LSP01-1A). Electrospinning was conducted at a voltage of 15 kV at an injection rate of 0.4 ml/h. A grounded stainless steel foil was
A
used as the collector of the electrospun nanofibers. The as-spun nanofiber mats were then peeled off from the collector and dried at 80 C, and finally calcined at 600 ºC for 3 h to obtain SnO2 nanofibers.
2.2 Sensor fabrication and Characterization An YSZ (8 mol% yttria-stabilized zirconia, Hefei Kejing Co., Ltd., China) disk 5
of 13 mm×13 mm size and 0.25 mm thickness was used as the solid electrolyte. Two circular Pt pastes (Sino-Platinum Co., Ltd., China) were coated on both sides of the YSZ disk and sintered at 800 ºC for 10min, serving as RE and counter electrode (CE), respectively. Concerning the SE, two different pastes were prepared by mixing
IP T
and grinding the SnO2 nanofibers or SnO2 nanoparticles (Hefei Quantum Quelle Nano Sci. Tech. Co., Ltd., China) with an organic binder (α-terpineol and ethyl
SC R
cellulose). The obtained pastes were successively coated on the same side as the RE of the YSZ disk, and then sintered at 600 ºC for 3h. Each SE and RE had a diameter
U
of 2.2 mm, and was spaced 3.1 mm apart, while the circular CE had a larger
N
diameter of 5.8 mm. Pt wires were connected to the electrodes with some high
A
temperature silver paste (DAD-87, Shanghai Research Institute of Synthetic Resins,
M
China) as the current collector. The obtained sensors of nanofibers and nanoparticles,
ED
having different SE morphology but identical solid electrolyte and RE, were denoted as NFs sensor and NPs sensor, respectively.
PT
Phase structure of the samples was identified by X-ray powder diffraction
CC E
(XRD, Rigaku TTR-III) using Cu Kα radiation. Micro-structural analysis was conducted with scanning electron microscopy (FE-SEM, JEOL JSM-6700F)
A
operated at an accelerating voltage of 20 kV, and transmission electron microscopy (TEM, JEM-2010) operated at 200 kV. Specific surface area of nanofibers and nanoparticles was measured by Trista II 3020M analyzer. X-ray photoelectron spectroscopy (XPS) was performed on an ESCLAB 250 spectrometer using Al Kα as the exciting source, wherein the binding energy of C1s at 284.8 eV was used as a 6
reference for energy calibration. X-ray fluorescence spectrometry (XRF-1800, Shimadzu) was carried out for quantitative elemental analysis. 2.3 Sensor test The sensing characteristics of sensors were measured using a home-built sensor
IP T
test system, similar to that described previously [28]. Sample gas was obtained by mixing dry air and a stream of standard gas. Each standard gas contained one
SC R
analyte of H2, NO2, CH4, C3H8, CO, and NH3, and was balanced with N2. Gas flow rates were adjusted by the mass flow controllers (MFC, CS200, Beijing Sevenstar
U
Electronics, China), and the total flow rate was fixed at 200 ml/min. The open
N
circuit electric potential (V) of the sensors was measured with Agilent 34972A
A
electrometer. A two-electrode configuration was adopted for response measurements,
M
wherein SE and RE were connected to the positive and negative terminal of the
ED
electrometer, respectively. Response of the sensor was defined as the difference of electric potential in sample gas and in air, ΔV = Vgas – Vair. Response/recovery time
PT
of the sensor was defined as the time when the V reached 90% of the response upon
CC E
H2 feeding/removal. Polarization (current–voltage) curves were obtained by means of the potentiodynamic method (CHI604E, Chenhua, China) using a three-electrode
A
configuration under a scan rate of 0.5 mV/s.
3. Results and discussion 3.1 Materials Properties and Microstructure As shown in Figure 1, the crystal structure of SnO2 nanofibers (NFs) and 7
nanoparticles (NPs) are both indexed as a single-phase tetragonal rutile structure (JCPDS No.: 41-1445). Compared with the sharp and intense peaks for the nanofibers, the peaks were significantly broadened for the nanoparticles, suggesting smaller grain size. According to the Debye-Scherrer equation, the grain sizes of the
IP T
nanofibers and nanoparticles were estimated to be 14.4 nm and 4.6 nm, respectively. The surface elemental composition and chemical status of the as-prepared NFs and
SC R
NPs were further investigated by XPS, and the corresponding results are presented
in Figure 2. Regardless of the presence of carbon likely from the ambient, the
U
samples consisted of only Sn and O, and no impurity was detected. The two strong
N
Sn 3d peaks centered at 495.0 and 486.6 eV corresponded to the Sn 3d3/2 and Sn3d5/2,
A
respectively, indicating that Sn existed in the form of Sn4+ in both samples [29].
M
XRF analysis revealed 99.2 at.% Sn content on the metal basis for the NPs,
ED
confirming a high purity of the commercial powders. Figure 3 presents SEM and TEM images of SnO2 NFs, NPs, and the different
PT
sensing electrodes. It can be clearly seen that the SnO2 nanofibers were highly
CC E
porous with numerous pores on the wall of the fiber and had a hollow core (Figure 3a). The fibers were composed of nanoparticles of 10-20 nm, and had a typical
A
length of 0.5-2 m and a uniform outer diameter of ~120 nm. The size of the pores on the wall of the fiber was close to that of the particles. The highly porous hollow NFs were randomly but evenly stacked upon each other in the sintered 14.5μm-thick NFs SE, establishing a three-dimensional scaffold architecture with a high degree of submicron-sized void space among them (Figure 3b, 3c). In contrast to the NFs, the 8
pristine NPs in the powder form had much smaller grain sizes of ~2-5 nm. Loose agglomeration of the NPs into assemblies of size up to hundreds of nanometers was also observed, which contained numerous voids of several nanometers size among the particles. In the sintered NPs SE, secondary particles formed by aggregation of
IP T
the nanoparticles were found, the size of which varied significantly in a broad range from tens of nanometers up to ~1 m. As a result of the wide particle size
SC R
distribution, the NP SE appeared more compact with much lower porosity and generally much smaller pores when compared with the NFs SE, despite the presence
U
of some submicron pores among adjacent larger secondary particles. This
N
morphology is expected to significantly increase the resistance for gas transport and
A
lead to smaller gas diffusivity. The film thickness of the NPs SE was 17.2 μm, close
M
to that of the NFs SE. Furthermore, the BET surface area of the pristine NFs and
ED
NPs were determined to be 14.1 m2/g and 93.7 m2/g, respectively, agreeing well with the observed smaller particle size of the NPs.
PT
3.2 Sensing performance
CC E
Figures 4a-b compare the response (ΔV, in absolute values) as a function of H2 concentration for the NFs and NPs sensors at 450 °C and 550 °C, respectively. It can
A
be seen that the NFs sensor exhibited markedly higher response than the NPs sensor, which were 10 and 5 times higher for 100 ppm and 1000 ppm H2 at 450 °C, respectively. Furthermore, the response of the NFs sensor varied more pronouncedly with the H2 concentration especially for lower concentrations, indicating that it was more sensitive to the concentration changes (i.e., had larger sensitivity). 9
Figures 5a-b show more clearly the variation of response values with H2 concentration for the NFs sensor at 400-550 ºC. Clearly, the response increased, i.e., became more negative, with increasing H2 concentration and with decreasing temperature. High response of -371.1 mV and -289.1mV was observed at 400 ºC
IP T
and 450 ºC for 1000 ppm H2, respectively. In the lower H2 concentration range of 40-140 ppm, the response varied linearly with H2 concentration at 450-550 ºC
SC R
(Figure 5a). The sensitivity, which is represented by the slope of the linear fitting,
increased with decreasing temperature. In contrast, at a lower temperature of 400 ºC,
U
the sensor response exhibited logarithmic concentration dependence with a slope of
N
-419.6 mV/decade. Similar logarithmic dependence was observed in the higher H2
A
concentration range of 140-2000 ppm for all temperatures tested (Figure 5b). The
M
sensitivity (slope) was close to each other at 400-500 ºC, but became much smaller
ED
at a higher temperature, 550 ºC. The largest sensitivity of -158.6 mV/decade for 140-2000 ppm was obtained at 450 ºC. The different concentration dependence of
PT
the response values of NFs sensor is indicative of different sensing mechanisms
CC E
according to the mixed-potential theory [30, 31]. The linear variation of the response with H2 concentration suggests linear oxygen reduction kinetics and H2 oxidation
A
kinetics rate-limited by the diffusional mass transport of H2. With respect to the logarithmic concentration dependence, the reactions of anodic hydrogen oxidation and/or cathodic oxygen reduction most likely follow the Tafel-type kinetics. In order to verify the sensing mechanism, the polarization curves of the NFs sensor were measured at 450 ºC. As shown in Figure 6, the cathodic current 10
measured in air varied linearly with the applied potential. This behavior manifested that cathodic oxygen reduction followed the linear approximation of the Butler-Volmer equation at low overpotentials [8, 30], consistent with the linear concentration dependence of the response in Figure 5a. From the intersections
IP T
of the anodic plots with the cathodic one, the V can be estimated. The estimated V values for 60 ppm and 100 ppm hydrogen were -28 mV and -97 mV, respectively,
SC R
in good agreement with the corresponding measured response values of -29.1 mV and -98.2 mV in Figure 5a. These results confirmed that the operation of NFs sensor
U
followed the mixed-potential mechanism.
N
Figure 7 shows variation of response/recovery time for the NFs and NPs
A
sensors with H2 concentration at 450 ºC. The response time decreased quickly with
M
the increasing H2 concentration at low concentrations and slowly at high
ED
concentrations (Figure 7a). The NFs sensor exhibited distinctly smaller response time than the NPs sensor, which were 5 s and 13 s at 1000 ppm, respectively,
PT
corresponding to a 2.6-fold reduction. Similar behavior was also observed for the
CC E
recovery time, despite the much larger values (Figure 7b). The decrease of the recovery time with concentration was not so pronounced as the response time, and
A
became slow above 100 ppm. Again, the NFs sensor showed distinctly shorter recovery time than the NPs sensor. A value of 38 s was obtained for the NFs sensor at 100 ppm, 2.5 times shorter than that of the NPs sensor. As shown in Figure 8, the response time of the NFs sensor decreased with increasing temperature or with increasing H2 concentration. Linear relation was 11
observed in the log-log plots of response time versus H2 concentration at various temperatures, indicating power-law dependence. A change of the dependence occurred at 140 ppm H2 in the temperature range of 450-550 ºC. The exponents for the higher concentration range varied only slightly within -0.8~-1.0, whereas smaller
IP T
exponents of -0.3~-0.5 were found for lower concentrations below 140 ppm H2. In contrast, at the lower temperature, 400 ºC, the change of the concentration
SC R
dependence was not observed, and the same power-law relation held for the entire
concentration range of 40-2000 ppm with a single exponent of -0.8. The
U
concentration dependence of the response time highly resembled that for the
N
response values, which may be ascribed to the different sensing mechanisms at low
A
and high H2 concentrations.
M
Figure 9 displays the cross-sensitivities of the sensors to 100 ppm various gases
ED
at 450 ºC. The NFs sensor was most responsive to H2, and also exhibited a minor response to NH3 and CO. Negligible response to the rest of the interfering gases,
PT
CH4, C3H8, and NO2, was observed. The response to the most interfering gas, CO,
CC E
was only -13.3 mV, over 7 times smaller than that for H2. By contrast, the NPs sensor was confronted with strong interference from NH3, the response to which was
A
comparable to that for H2. In addition, a modest response to CO with a reduced value of -3.2 mV was also found. Therefore, more selective H2 sensing was achieved for the NFs sensor. For practical applications, stable and repeatable operation of gas sensors is of crucial importance. Figure 10a presents the response values of NFs sensor for 100 12
ppm, 200 ppm, and 300 ppm H2 during a 33-day measurement at 450 ºC. One can see clearly that the values remained almost constant within this period, demonstrating excellent long-time stability of the NFs sensor. Figure 10b shows that during seven successive runs at 450 ºC in 100 ppm H2, both the baseline and
IP T
response transients of the NFs sensor were stable and reproduced very well. Figure 10c compares the response values of two different NFs sensors at 450 ºC.
SC R
Apparently they were very close to each other at each concentration tested and a
discrepancy within 10% was found. Therefore, the sensing characteristics of the NFs
U
sensor can be well repeated not only for a single sample, but also between different
N
sensors.
A
3.3 Relation of sensing performance and diffusion-reaction process
M
The above results showed evidently that electrode morphology plays an
ED
important role in determining the gas sensing performance of SnO2/YSZ/Pt mixed potential sensor. The NFs sensor of porous hollow nanofibers exhibited much superior
PT
H2 sensing performance.
CC E
It has been known that heterogeneous reaction takes place on the SE surface during diffusion of hydrogen through the porous SE:
H 2 (g) O2 (g) H 2O( g )
(1)
A
At the TPB, cathodic oxygen reduction reaction and anodic reaction take place simultaneously to generate the mixed potential: O2 (g) 4e 2O2
(2)
H 2 (g) O2 H 2O( g ) 2e
(3) 13
A good SE should have high electrochemical activity for the anodic reaction (3), for which the reaction rate is a function of both the reaction rate constant and concentration, e.g. R=k’Ceff for a first-order reaction. Note that Ceff is the actual (effective) concentration at TPB, which is generally smaller than the starting
IP T
concentration, C0, due to reaction (1). Reaction (1) inside the SE consumes H2 and is unfavorable to achieving high response values. In particular, insufficient H2 would
SC R
reach the TPB when the H2 concentration is low, giving rise to diffusion-controlled kinetics. The observed linear concentration dependence of the sensor response for
U
below 140 ppm at 450-550 ºC in Figure 5a agrees well with this mechanism.
N
For a planar thick film electrode used in this work, a concentration profile is
A
established inside the electrode due to reaction (1) according to the heterogeneous
M
catalysis theory [32]. Assuming a first-order reaction kinetics, the H2 concentration
C0 cosh(( L x) k / D )
(4)
cosh( L k / D )
PT
C ( x)
ED
under steady state can be calculated by:
where L is the film thickness, x the distance from the electrode surface into the
CC E
electrode, and D gas diffusion coefficient, k rate constant for reaction (1). For mixed potential sensors, only H2 that reaches the TPB is effective for generating the mixed
A
potential and the sensor signals. This effective H2 concentration can be obtained from Equation (4) at x=L: Ceff C ( L)
C0
(5)
cosh( L k / D )
Equation (5) can be rewritten to obtain the normalized effective concentration at TPB, 14
c
Ceff C0
1 cosh( )
(6)
where
L k/D
(7)
IP T
is the so called Thiele modulus for isothermal first-order reaction, which has been widely used to represent the relative importance of diffusion and reaction rates in
SC R
porous catalysts. In principle, a larger value of Thiele modulus accords with relatively slow gas diffusion. Figure 11 shows variation of the normalized effective concentration c at TPB as a function of . Specifically, at <0.4, the concentration
U
reduction is insignificant with c>0.92, and gas diffusion is fast. As value increases,
M
primarily rate-limited by gas diffusion.
A
N
c decreases significantly. At >3, c is smaller than 0.1, indicating that the process is
According to Equation (7), small Thiele modulus value can be achieved by
ED
reducing the electrode thickness, improving gas transport, and using materials of low
PT
activity for reaction (1). Both k and D are known to strongly correlate with the materials morphology and electrode microstructure. Concerning the sensors
CC E
investigated in the present work, their only difference is the materials morphology and electrode microstructure of the SE, since they shared the same solid electrolyte and
A
RE and had similar SE thickness. As schematically illustrated in Figure 12a, the NFs SE was characterized by a 3D scaffold architecture formed by random stacking of highly porous and hollow nanofibers upon each other. The macropores among the nanofibers and those along the hollow axis appeared to dominate the overall porosity, through which gas diffusion was expected to proceed via the fast molecular diffusion. 15
Moreover, these macropores were highly inter-connected, partly through the numerous pores on the fiber walls, allowing for more facile gas transport. The high porosity, large pores, high degree of pore interconnectivity, and low degree of channel tortuosity are highly beneficial to achieving large gas diffusivity. By great contrast, in
IP T
the NPs SE (Figure 12b), nano-sized pores within and among the secondary particles dominated the overall porosity, and thus Knudsen diffusion is expected to play an
SC R
important role, which is much slower than the molecular diffusion. In such structure,
poor pore interconnectivity and tortuous gas transport pathway would be expected in
U
addition to the observed compactness. These would lead to poorer gas transport in the
N
NPs SE. On the other hand, the smaller particle size and much larger BET surface
A
area of the nanoparticles were in principle associated with higher surface reactivity,
M
and thus resulted in larger k. Therefore, the much superior sensing response for the
ED
NFs sensor relative to the NPs sensor as observed in Figure 4 can be accounted for by its smaller (larger D and smaller k), resulting from the unique 3D hierarchical
PT
electrode microstructure.
CC E
For an irreversible reaction of order n (n<1) the generalized Thiele modulus is used [32]:
n 1 k n 1 C0 2 D
(8)
A
L
When n=1, Equation (8) becomes Equation (7) for first-order reaction, and is independent of concentration. For a reaction of order -1
diffusion- or reaction-controlled behavior of the response values for NFs sensor at 450-550 ºC in Figure 5. That only reaction-controlled behavior was observed in the whole concentration range at the lower temperature of 400 ºC can be interpreted by the temperature dependence of . As k decreases faster than D with temperature
IP T
decreasing, reaction rate becomes slower than gas diffusion and dominates the process at lower temperatures. The two-segmental behavior of the response value at 400 ºC
SC R
(Figure 5b) is also consistent with the variation of with concentration.
Similar to the case of response value, the diffusion-reaction process in the SE
U
may also exert some impact on the response kinetics. Nevertheless, this important
N
aspect has received little attention and has not been addressed by the existing theory
A
of mixed-potential gas sensors [8, 30]. In the present work, the response time for the
M
SnO2 NFs sensor exhibited power-law dependence on H2 concentration. Similar
ED
phenomenon has also been observed for semiconductor gas sensors [33, 34], which was explained by a non-linear diffusion-reaction model developed based on
PT
Freundlich adsorption isotherm. This model could in principle be extended to
CC E
mixed-potential gas sensors, given the same diffusion-reaction process in their porous electrode as in the semiconductor sensors. According to the model, a modified response time is defined as [34]
A
L2
k m1 C0 D
(9)
where m is a constant. An m value close to 1 suggests a diffusion-controlled process, whilst reaction-controlled process dominates as m approaches zero. The m values of the NFs sensor at 450-550 ºC were 0.5~0.7 and 0~0.2 (Figure 7), corresponding to 17
diffusion- and reaction-controlled kinetics for below and above 140 ppm H2, respectively. At 400 ºC, no transition was observed, and the m value of 0.2 indicated reaction-controlled kinetics for the entire concentration range. This behavior of the response kinetics is in good accordance with that of the response values in Figure 5
IP T
and as discussed above.
is in fact related with the generalized Thiele modulus:
2 2 mn C0 n 1
SC R
If Eqs. (8) and (9) are combined, one can immediately see that the response time
(10)
U
Therefore, the magnitude of response time is largely governed by , and small may
N
accord with fast response kinetics. The much faster response kinetics of the NFs
A
sensor relative to the NPs sensor in Figure 6a can thus be ascribed again to the faster
M
gas transport and smaller k in the 3D nanofiber SE.
ED
The sensing performance of mixed-potential gas sensors is jointly governed by the kinetics of the electrochemical and heterogeneous reactions. As the kinetics of
PT
reactions (1) and (3) differ for various gases, different response characteristics would
CC E
be expected, accounting for the sensor selectivity. The high H2 response for the NFs sensor suggests that SnO2 has a high activity for the electrochemical H2 oxidation reaction (3) while the heterogeneous H2 oxidation reaction (1) did not significantly
A
take place. The greatly reduced H2 and CO response for the NPs sensor can be ascribed to the slow gas transport in the electrode, which facilitated the heterogeneous reaction (1). The observation in Figure 9 that the response for the other gases was quite small and did not significantly vary with the electrode morphology suggests that 18
most likely both the electrochemical and heterogeneous activities to these gases are low for SnO2. These results also show that the sensor selectivity can be tuned by manipulating the electrode microstructure. The present work showed that electrode morphology was of crucial importance to
IP T
the gas sensing performance of SnO2/YSZ/Pt mixed potential sensor. Superior sensing performance, particularly high sensitivity and fast response kinetics, could be
SC R
achieved based on hierarchical porous hollow nanofibers, benefiting from the fast gas transport and reduced heterogeneous reaction. Note that although this work was
U
mainly focused on H2 sensing, the results obtained herein could in principle be
A
N
extended to other mixed-potential gas sensors for different gases.
M
4. Conclusion
ED
A high performance planar mixed-potential H2 sensor was obtained based on hierarchical porous hollow SnO2 nanofiber sensing electrode. A high response value
PT
of -289.1 mV and short response time of 5 s at 450 ºC for 1000 ppm H2 were obtained.
CC E
The response values were linearly or logarithmically dependent on H2 concentration for low and high concentrations, respectively, indicating a change from diffusion- to reaction-controlled kinetics. A power-law relationship between the response time and
A
concentration was observed, which also exhibited a similar transition at 140 ppm. Moreover, excellent selectivity to H2 and repeatable and stable operation were achieved. The sensing performance was highly dependent on the electrode morphology, which closely correlated with the diffusion-reaction process and Thiele 19
modulus. The highly porous three-dimensional electrode architecture of nanofibers favors small Thiele modulus and is advantageous for mixed potential gas sensors.
Acknowledgements
IP T
Finance support by the Natural Science Foundation of China (No. U1432108), and the Fundamental Research Funds for the Central Universities (No. WK2320000034) is
A
CC E
PT
ED
M
A
N
U
SC R
gratefully acknowledged.
20
References [1] S. Bagchi, S. Sengupta, S. Mandal, Development and Characterization of Carbonic Anhydrase Based CO2 Biosensor for Primary Diagnosis of Respiratory Health, Sensors, 15423 (2016) 1. [2] G. Gregis, J.-B. Sanchez, I. Bezverkhyy, W. Guy, F. Berger, V. Fierro, et al.,
IP T
Detection and quantification of lung cancer biomarkers by a micro-analytical device using a single metal oxide-based gas sensor, Sensors and Actuators B: Chemical, 255 (2018) 391-400.
SC R
[3] R. Baron, J. Saffell, Amperometric Gas Sensors as a Low Cost Emerging
Technology Platform for Air Quality Monitoring Applications: A Review, ACS sensors, 2 (2017) 1553-1566.
U
[4] L. Boon-Brett, J. Bousek, G. Black, P. Moretto, P. Castello, T. Hübert, et al.,
N
Identifying performance gaps in hydrogen safety sensor technology for automotive
A
and stationary applications, International Journal of Hydrogen Energy, 35 (2010)
M
373-84.
[5] J.W. Fergus, A review of electrolyte and electrode materials for high temperature
(2008) 1034-1041.
ED
electrochemical CO2 and SO2 gas sensors, Sensors and Actuators B: Chemical, 134
PT
[6] S.A. Anggraini, M. Breedon, H. Ikeda, N. Miura, Insight into the aging effect on enhancement of hydrogen-sensing characteristics of a zirconia-based sensor utilizing
CC E
a Zn–Ta–O-based sensing electrode, ACS applied materials & interfaces, 5 (2013) 12099-12106.
[7] Y. Guan, C. Li, X. Cheng, B. Wang, R. Sun, X. Liang, et al., Highly sensitive
A
mixed-potential-type NO2 sensor with YSZ processed using femtosecond laser direct writing technology, Sensors and Actuators B: Chemical, 198 (2014) 110-113. [8] N. Miura, T. Sato, S.A. Anggraini, H. Ikeda, S. Zhuiykov, A review of mixed-potential type zirconia-based gas sensors, Ionics, 20 (2014) 901-925. [9] S.A. Anggraini, M. Breedon, N. Miura, Zn–Ta-based oxide as a hydrogen sensitive electrode material for zirconia-based electrochemical gas sensors, Sensors 21
and Actuators B: Chemical, 187 (2013) 58-64. [10] Y. Li, X. Li, J. Wang, Y. Jun, Z. Tang, A cobalt tungstate compound sensing electrode for hydrogen detection based upon mixed-potential type sensors, RSC Advances, 7 (2017) 2919-2925. [11] Z. Tang, X. Li, J. Yang, J. Yu, J. Wang, Z. Tang, Mixed potential hydrogen sensor using ZnWO4 sensing electrode, Sensors and Actuators B: Chemical, 195 (2014)
IP T
520-525.
[12] X. Liang, S. Yang, J. Li, H. Zhang, Q. Diao, W. Zhao, et al., Mixed-potential-type
SC R
zirconia-based NO2 sensor with high-performance three-phase boundary, Sensors and Actuators B: Chemical, 158 (2011) 1-8.
[13] L.P. Martin, A.Q. Pham, R.S. Glass, Effect of Cr2O3 electrode morphology on the
U
nitric oxide response of a stabilized zirconia sensor, Sensors and Actuators B:
N
Chemical, 96 (2003) 53-60.
A
[14] B. Wang, F. Liu, X. Yang, Y. Guan, C. Ma, X. Hao, et al., Fabrication of well-ordered three-phase boundary with nanostructure pore array for mixed
M
potential-type zirconia-based NO2 sensor, ACS applied materials & interfaces, 8
ED
(2016) 16752-16760.
[15] R. Sun, Y. Guan, X. Cheng, Y. Guan, X. Liang, J. Ma, et al., High performance three-phase boundary obtained by sand blasting technology for mixed-potential-type
PT
zirconia-based NO2 sensors, Sensors and Actuators B: Chemical, 210 (2015) 91-95. [16] X. Cheng, C. Wang, B. Wang, R. Sun, Y. Guan, Y. Sun, et al.,
CC E
Mixed-potential-type YSZ-based sensor with nano-structured NiO and porous TPB processed with pore-formers using coating technique, Sensors and Actuators B: Chemical, 221 (2015) 1321-1329.
A
[17] M. Yamaguchi, S.A. Anggraini, Y. Fujio, T. Sato, M. Breedon, N. Miura, Stabilized zirconia-based sensor utilizing SnO2-based sensing electrode with an integrated Cr2O3 catalyst layer for sensitive and selective detection of hydrogen, International Journal of Hydrogen Energy, 38 (2013) 305-312. [18] L. Zhang, Y. Yin, Hierarchically mesoporous SnO2 nanosheets: hydrothermal synthesis and highly ethanol-sensitive properties operated at low temperature, Sensors 22
and Actuators B: Chemical, 185 (2013) 594-601. [19] T. Yu, X. Cheng, X. Zhang, L. Sui, Y. Xu, S. Gao, et al., Highly sensitive H2S detection sensors at low temperature based on hierarchically structured NiO porous nanowall arrays, Journal of Materials Chemistry A, 3 (2015) 11991-11999. [20] Q. Wang, X. Li, F. Liu, Y. Sun, C. Wang, X. Li, et al., Three-dimensional flake-flower Co/Sn oxide composite and its excellent ethanol sensing properties,
IP T
Sensors and Actuators B: Chemical, 230 (2016) 17-24.
[21] M. Zhi, S. Lee, N. Miller, N.H. Menzler, N. Wu, An intermediate-temperature
SC R
solid oxide fuel cell with electrospun nanofiber cathode, Energy & Environmental Science, 5 (2012) 7066-7071.
[22] Y. Jeon, J.-h. Myung, S.-h. Hyun, Y.-g. Shul, J.T. Irvine, Corn-cob like nanofibres
U
as cathode catalysts for an effective microstructure design in solid oxide fuel cells,
N
Journal of Materials Chemistry A, 5 (2017) 3966-3973.
[23] Z. Wang, M. Li, C. Liang, L. Fan, J. Han, Y. Xiong, Effect of morphology on the
A
oxygen evolution reaction for La0.8Sr0.2Co0.2Fe0.8O3−δelectrochemical catalyst in
M
alkaline media, RSC Adv, 6 (2016) 69251-69256.
ED
[24] L. Dai, Y. Liu, W. Meng, G. Yang, H. Zhou, Z. He, et al., Ammonia sensing characteristics of La10Si5MgO26 -based sensors using In2O3 sensing electrode with different morphologies and CuO reference electrode, Sensors and Actuators B:
PT
Chemical, 228 (2016) 716-724.
[25] Y. Liu, L. Dai, H. Hou, Y. Li, W. Meng, L. Wang, Mixed-potential type NO2
CC E
sensor based on La10Si6O27 electrolyte and WO3 sensing electrode with different morphologies, Ceramics International, 42 (2016) 9712-9716. [26] W. Meng, L. Dai, J. Zhu, Y. Li, W. Meng, H. Zhou, et al., A novel mixed potential
A
NH3 sensor based on TiO2@WO3 core–shell composite sensing electrode, Electrochimica Acta, 193 (2016) 302-310. [27] F. Liu, Y. Guan, H. Sun, X. Xu, R. Sun, X. Liang, et al., YSZ-based NO2 sensor utilizing hierarchical In2O3 electrode, Sensors and Actuators B: Chemical, 222 (2016) 698-706. [28] H. Zhang, J. Yi, X. Jiang, Fast Response, Highly Sensitive and Selective 23
Mixed-Potential H2 Sensor Based on (La, Sr)(Cr, Fe) O3-δ Perovskite Sensing Electrode, ACS Applied Materials & Interfaces, 9 (2017) 17218-17225. [29] Y.-H. Choi, S.-H. Hong, H2 sensing properties in highly oriented SnO2 thin films, Sensors and Actuators B: Chemical, 125 (2007) 504-509. [30] F.H. Garzon, R. Mukundan, E.L. Brosha, Solid-state mixed potential gas sensors: theory, experiments and challenges, Solid State Ionics, 136 (2000) 633-638.
IP T
[31] S.A. Anggraini, M. Breedon, N. Miura, Sensing characteristics of aged
zirconia-based hydrogen sensor utilizing Zn–Ta-based oxide sensing-electrode,
SC R
Electrochemistry Communications, 31 (2013) 133-136.
[32] M.E. Davis, R.J. Davis, Fundamentals of chemical reaction engineering, Courier Corporation, 2012.
U
[33] S.J. Choi, S. Chattopadhyay, J.J. Kim, S.J. Kim, H.L. Tuller, G.C. Rutledge, et
N
al., Coaxial electrospinning of WO3 nanotubes functionalized with bio-inspired Pd
A
catalysts and their superior hydrogen sensing performance, Nanoscale, 8 (2016) 9159-9166.
M
[34] J.W. Gardner, A non-linear diffusion-reaction model of electrical conduction in
A
CC E
PT
ED
semiconductor gas sensors, Sensors and Actuators B: Chemical, 1 (1990) 166-170.
24
Author biographies Jianxin Yi is an associate professor of State Key Laboratory of Fire Science, University of Science and Technology of China (USTC). He received both his B.S. degree in Chemical Physics and PhD degree in Materials Science from USTC. He
IP T
then worked as a postdoctoral researcher at University of Twente and at RWTH Aachen University during 2006-2011. Before joining USTC in 2012, he worked as an
SC R
associate professor of Materials science at Huazhong University of Science and
Technology. His current research is focused on gas sensors, functional nanostructured
N
U
materials, and mixed-conducting oxides.
A
Hong Zhang is a master student at University of Science and Technology of China.
M
Her current research is on mixed potential gas sensors of nanostructured sensing
ED
electrode.
PT
Zuobin Zhang is a PhD degree candidate at University of Science and Technology of
CC E
China. His research is focused on gas sensors based on mixed potential theory.
Dongdong Chen is a master student at University of Science and Technology of
A
China. His research is mainly focused on semiconductor gas sensor.
25
< Figure Captions> Figure 1. XRD patterns of SnO2 nanofibers and nanoparticles. Figure 2. (a) Survey XPS spectrum and (b) high resolution spectra for Sn 3d of the as-prepared SnO2 nanofibers and nanoparticles.
SC R
IP T
Figure 3. (a-c) SEM images for (a) pristine SnO2 nanofibers and (b) surface and (c) cross-section of the NFs SE; inset in (a) is the TEM image of a nanofiber. (d) TEM image of pristine SnO2 nanoparticles, (e, f) SEM images for (e) surface and (f) cross-section of NPs SE. Figure 4. Response of NFs sensor and NPs sensor as a function of hydrogen concentration at (a) 450°C and (b) 550°C. Absolute values are used for the sake of comparison.
N
U
Figure 5. Response values of NFs sensor as a function of (a) hydrogen concentration and (b) logarithmic concentrations.
A
Figure 6. Polarization curves for the NFs sensor at 450 °C.
M
Figure 7. Variation of (a) response time and (b) recovery time for NFs sensor and NPs sensor with hydrogen concentration at 450°C.
ED
Figure 8. Log-log plots of response time of NFs sensor versus hydrogen concentration at 450 °C and 550 °C. Inset shows the data at 400 °C.
PT
Figure 9. Cross-sensitivities of NFs sensor and NP sensor to 100ppm various gases at 450°C.
CC E
Figure10. (a) Long-term stability, (b) continuous dynamic response curves in 100 ppm hydrogen, and (c) response comparison of two different samples at 450°C for NFs sensor.
A
Figure 11. Variation of the normalized effective concentration c as a function of . Figure12. Schematic illustrations for the gas transport in electrodes of different morphology. (a) NFs SE of nanofibers, (b) NPs SE of nanoparticles.
26
IP T SC R
A
CC E
Figure 2.
PT
ED
M
A
N
U
Figure 1.
27
IP T SC R
CC E
PT
ED
M
A
N
U
Figure 3.
A
Figure 4.
28
IP T
A
CC E
Figure 6.
PT
ED
M
A
N
U
SC R
Figure 5.
29
IP T
CC E
PT
ED
M
A
N
U
SC R
Figure 7.
A
Figure 8.
30
A ED
PT
CC E
IP T
SC R
U
N
A
M
Figure 9.
31
A ED
PT
CC E
Figure10.
32
IP T
SC R
U
N
A
M
IP T
A
CC E
PT
ED
M
A
N
U
SC R
Figure 11.
Figure12.
33