Highly Sensitive and Selective Hydrogen Gas Sensor ...

5 downloads 0 Views 7MB Size Report
Oct 14, 2017 - Manjula, P.; Satyanarayana, L.; Swarnalatha, Y.; Manorama, S.V. Raman and MASNMR studies to support the mechanism of low temperature ...
sensors Article

Highly Sensitive and Selective Hydrogen Gas Sensor Using the Mesoporous SnO2 Modified Layers Niuzi Xue 1 , Qinyi Zhang 1, *, Shunping Zhang 2 , Pan Zong 1 and Feng Yang 1 1 2

*

School of Material Science and Engineering, Wuhan University of Technology, Wuhan 430070, China; [email protected] (N.X.); [email protected] (P.Z.); [email protected] (F.Y.) Department of Materials Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, China; [email protected] Correspondence: [email protected]; Tel.: +86-181-6226-2798

Received: 7 September 2017; Accepted: 10 October 2017; Published: 14 October 2017

Abstract: It is important to improve the sensitivities and selectivities of metal oxide semiconductor (MOS) gas sensors when they are used to monitor the state of hydrogen in aerospace industry and electronic field. In this paper, the ordered mesoporous SnO2 (m-SnO2 ) powders were prepared by sol-gel method, and the morphology and structure were characterized by X-ray diffraction analysis (XRD), transmission electron microscope (TEM) and Brunauer–Emmett–Teller (BET). The gas sensors were fabricated using m-SnO2 as the modified layers on the surface of commercial SnO2 (c-SnO2 ) by screen printing technology, and tested for gas sensing towards ethanol, benzene and hydrogen with operating temperatures ranging from 200 ◦ C to 400 ◦ C. Higher sensitivity was achieved by using the modified m-SnO2 layers on the c-SnO2 gas sensor, and it was found that the S(c/m2) sensor exhibited the highest response (Ra/Rg = 22.2) to 1000 ppm hydrogen at 400 ◦ C. In this paper, the mechanism of the sensitivity and selectivity improvement of the gas sensors is also discussed. Keywords: gas sensor; mesoporous SnO2 ; hydrogen; sensitivity; selectivity

1. Introduction As one of the most important clean energies, H2 is widely used in various fields such as fuel cell vehicles, aerospace industry, petrochemical industry, and electronic field [1–3]. In consideration of the leakage in the applications of H2 whose explosive limit is very low, it is essential to monitor the state of hydrogen. Gas sensor is one of the most effective detectors [4,5]. Great emphasis is being given to metal oxide semiconductors (MOS), including ZnO [6], WO3 [7], TiO2 [8], In2 O3 [9] and SnO2 [10], as gas sensing materials for a long time. Among various MOS gas sensors, SnO2 -based gas sensors are widely used because of their low cost, high sensitivity and long-term stability [11]. However, poor selectivity to various gases restricts their applications. Gas sensing performances of SnO2 , especially the selectivity to H2 , can be improved by applying doping [12–14], catalyst [15–17], filtering membranes [18–20], etc. For example, Inyawilert et al. studied the films of SnO2 nanoparticles doped with 0.1~2 wt.% rhodium (Rh). It showed that the Rh-doped SnO2 sensor presented high H2 selectivity against NO2 , SO2 , C2 H4 , C3 H6 O, CH4 , H2 S and CO [13]. Liewhiran et al. reported that Pd-catalyzed SnO2 sensor (0.2 wt.% Pd/ SnO2 , 10 µm in thickness) showed ultra-high response to H2 [15]. It was found in the work of Tournier et al. that SiO2 filter film deposited on the SnO2 film is highly selective to hydrogen [18]. Filtering membranes such as SiO2 , Al2 O3 , Fe2 O3 , etc. work as molecular sieves. They are useful to improve the selectivity of gas sensors. Gas sensing performances of SnO2 gas sensors can be highly improved by using mesoporous material because of its high specific surface area (SSA) [21]. Japanese researchers fabricated nano-SnO2 powders coated by mesoporous SnO2 (m-SnO2 ), and this kind of SnO2 films highly increased the responses to H2 [21]. It was found in the work of Pijolat et al. that thin SiO2 films deposited on the Sensors 2017, 17, 2351; doi:10.3390/s17102351

www.mdpi.com/journal/sensors

Sensors 2017, 17, 2351

2 of 17

SnO2 thick films could improve the selectivity to H2 [22]. Dhawale et al. synthesized mesoporous ZnO thin films which showed high selectivity towards liquefied petroleum [23]. The aim of the present study was to improve the selectivity and sensitivity simultaneously in hydrogen detection. The uses of such mesoporous materials enable enhancement of the adsorption and reaction of test gas because of the high specific surface area. On the other hand, mesoporous materials are potential molecular sieves for gas sensors to improve their selectivity because of the mesoporous structure. In this paper, the m-SnO2 powders were synthesized with a simple and low cost sol-gel method. The sensors were fabricated using commercial SnO2 (c-SnO2 ) films as the basic layer and the m-SnO2 films as the modified layers by screen printing method. Their sensing performances were tested with hydrogen, ethanol and benzene. The relationships between selectivity and the thickness of the films were studied. The present study aims to develop a low cost and highly sensitive and selective hydrogen gas sensor. 2. Materials and Methods 2.1. Preparation of m-SnO2 Powders Employing Na2 SnO3 ·4H2 O as the Sn source, n-cetylpyridinium chloride (C16 PyCl) as the template and trimethylbenzene (C6 H3 (CH3 )3 ) as the surfactant, m-SnO2 powders were prepared in a similar way to that reported previously [21]. The typical preparation manner was as follows. C16 PyCl was added to the deionized water at 2.6 wt.%, while Na2 SnO3 ·4H2 O was dissolved in the deionized water at 3.6 wt.%. In this case, Na2 SnO3 ·4H2 O aqueous was mixed with the C16 PyCl solution at a molar ratio [C16 PyCl]/[Na2 SnO3 ·4H2 O] = 2.0. Then, trimethylbenzene was added to the solution at a molar ratio [C6 H3 (CH3 )3 ]/[Na2 SnO3 ·4H2 O] = 2.5. The pH of the mixture was then adjusted to 10 with an aqueous 35 wt.% HCl solution. The resultant emulsion solution was aged for 2 days at 25 ◦ C. After suction filtration with deionized water and drying, the resultant solid products were treated with a 0.1 M aqueous phosphoric acid (PA) solution for 2 h with magnetic stirrers. Then, it was filtered off, washed and dried at 60 ◦ C for 12 h. Eventually, the solid was calcined at 600 ◦ C for 5 h in air. After calcination, the powders were subjected to mechanical grinding with an agate mortar. The crystal phases of the m-SnO2 powders were characterized via X-ray diffraction analysis (XRD, D8 Adwance, Bruker, Karlsruhe, Germany). The specific surface area and pore size distribution were measured by the Brunauer–Emmett–Teller (BET) method using a N2 adsorption isotherm (BET, ASAP 2020, Micromeritics, Norcross, GA, USA). Morphology of the m-SnO2 powders was observed by a transmission electron microscope (TEM, JEM2100F STEM/EDS, JEOL, Tokyo, Japan) and the morphology of the commercial SnO2 (c-SnO2 ) powders was observed by a scanning electron microscope (SEM, JSM-IT300, JEOL, Tokyo, Japan). 2.2. Fabrication of SnO2 Sensors Pastes of the c-SnO2 powders and the as-prepared m-SnO2 powders were applied on a substrate (30 mm × 6 mm × 0.625 mm), on which interdigitated Pt electrodes had been printed with mechanically automated screen printing technology, as shown in Figure 1. The thick film gas sensors were fabricated using screen printing technology. For the first layer, the c-SnO2 powders were mixed with the printing oil (YY-1010, Wuhan Huachuang Ruike Tech. Co. LTD, Wuhan, China) at the mass ratio of 1:1 as the paste. Furthermore, to improve the stability of the gas sensors, the frit of PbO, B2 O3 , and SiO2 (mass ratio [PbO]/[B2 O3 ]/[SiO2 ] = 45/35/20) was added into the c-SnO2 powders at the level of 2 wt.%. The substrates were treated with drying at room temperature for 10 min and 50 ◦ C for 1 h when the pastes were printed on them. For modified layer, the paste was mixed with the m-SnO2 powders and the printing oil at the same mass ratio of 1:1. To prepare more modified layers, simply repeat the printing step above. Eventually, the gas sensors were dried at 50 ◦ C for 1 h and calcined at 650 ◦ C for 2 h. The different fabricated gas sensors are listed in Table 1.

Sensors 2017, 17, 2351

3 of 17

Sensors 2017, 17, 2351

3 of 17

were dried at 50 °C for 1 h and calcined at 650 °C for 2 h. The different fabricated gas sensors are listed dried in Table 1. °C for 1 h and calcined at 650 °C for 2 h. The different fabricated gas sensors are were at 50 Sensors 2017, 17, 2351 3 of 17 listed in Table 1.

Figure 1. The substrate structure of the gas sensor. Figure 1. The substrate structure of the gas sensor. Figure 1. The substrate structure thesensors. gas sensor. Table 1. The structure of theof gas Table 1. The structure of the gas sensors.

Basic Layer Modified Layer Table 1. The structure of the gas sensors. Sample Description Material Material Basic Layer Modified Layer Sample Description S(c) c-SnO2 Basic Layer Material / Modified Layer one layerDescription of c-SnO2 Material Sample Material Material S(m) m-SnO one S(c) c-SnO22 / one layer layer of of m-SnO c-SnO22 / layer of layer c-SnO S(c/m1) c-SnO22 c-SnO2 m-SnO one layer of one c-SnO 2one and one S(m)S(c) m-SnO / 2 layer of m-SnO 2 2 of m-SnO2 S(m) m-SnO2 / one layer of m-SnO2 S(c/m2) and ofofm-SnO m-SnO S(c/m1) c-SnO2 c-SnO2 m-SnO2 m-SnO2one one layer layer of of c-SnO c-SnO and2two one layer of S(c/m1) one layer of22c-SnO and layers one layer m-SnO222 S(c/m2) one layer of22c-SnO and two layers ofm-SnO m-SnO2 2 2 2onelayer 2two S(c/m3) and layers of m-SnO S(c/m2) c-SnO2 c-SnO2 m-SnO2 m-SnOone layerof ofc-SnO c-SnO andthree layers of S(c/m3) c-SnO2 m-SnO2 one layer of c-SnO2 and three layers of m-SnO2 S(c/m3) c-SnO2 m-SnO2 one layer of c-SnO2 and three layers of m-SnO2 The surface morphology of the prepared gas sensors was observed by a scanning electron The ofPlus, the sensors observed by aa scanning microscope (SEM,morphology Zeiss Utralof Cari Zeissgas AG, Jena, was Germany). The of the The surface surface morphology the prepared prepared gas sensors was observed bycross-sections scanning electron electron microscope (SEM, Zeiss Utral Plus, Cari Zeiss AG, Jena, Germany). The cross-sections of the different different SnO 2 films were observed by a scanning electron microscope (SEM, S-4800, HITACHI, microscope (SEM, Zeiss Utral Plus, Cari Zeiss AG, Jena, Germany). The cross-sections of the SnO were observed a scanning microscope S-4800,(SEM, HITACHI, Tokyo, Japan). 2 films Tokyo, Japan). different SnO 2 films wereby observed by electron a scanning electron(SEM, microscope S-4800, HITACHI, Tokyo, Japan). 2.3. 2.3. Measurement Measurement of of Sensing Sensing Performance Performance 2.3. Measurement of Sensingmeasured Performance The by abycommercial SD-101 gas sensing performance testing testing device The gas gassensors sensorswere were measured a commercial SD-101 gas sensing performance (Wuhan Huachuang Ruike Tech. Co. LTD, Wuhan, China) which can be used with four gas sensors device (Wuhan Huachuang Tech.byCo. Wuhan,SD-101 China)gas which can beperformance used with four gas The gas sensors were Ruike measured a LTD, commercial sensing testing to test their gas sensing performance simultaneously (Figure 2). The operating temperature can be sensors(Wuhan to test their gas sensing performance simultaneously (Figure 2). The temperature device Huachuang Ruike Tech. Co. LTD, Wuhan, China) which can operating be used with four gas controlled via adjusting the power of the heater coil by a microprocessor. The operating temperature can be controlled the power ofsimultaneously the heater coil(Figure by a microprocessor. The operating sensors to test theirvia gasadjusting sensing performance 2). The operating temperature of the gas sensors isgas in the rangeisof room temperature to 450 ◦ C. temperature of the sensors in the range of room temperature to 450 °C. can be controlled via adjusting the power of the heater coil by a microprocessor. The operating temperature of the gas sensors is in the range of room temperature to 450 °C.

Figure 2. The SD-101 gas sensing performance testing device. Figure 2. The SD-101 gas sensing performance testing device. Figure 2. The SD-101 gas sensing performance testing device.

The prepared gas sensors were measured to sense 1000 ppm H2 with dynamic method and 10 ppm ethanol and benzene with static method at the temperature of 200 ◦ C, 250 ◦ C, 300 ◦ C, 350 ◦ C and 400 ◦ C. In the process of dynamic measurement, the SD-101 gas sensing performance testing device

Sensors 2017, 17, 2351 Sensors 2017, 17, 2351

4 of 17 4 of 17

sensors were measured SensorsThe 2017,prepared 17, 2351 gas The prepared gas sensors were measured

to sense 1000 ppm H2 with dynamic method and 4 of10 17 to sense 1000 ppm H2 with dynamic method and 10 ppm ethanol and benzene with static method at the temperature of 200 °C, 250 °C, 300 °C, 350 °C ppm ethanol and benzene with static method at the temperature of 200 °C, 250 °C, 300 °C, 350 °C and 400 °C. In the process of dynamic measurement, the SD-101 gas sensing performance testing and 400 °C. In the process of dynamic measurement, the SD-101 gas sensing performance testing was placed a cylinder of 60 mm which is made methacrylate (PMMA). device was in placed in a cylinder ofin 60diameter, mm in diameter, whichof is polymethyl made of polymethyl methacrylate device was placed in a cylinder of 60 mm in diameter, which is made of polymethyl methacrylate The testing gas flowchart is shown in Figure 3. (PMMA). The testing gas flowchart is shown in Figure 3. (PMMA). The testing gas flowchart is shown in Figure 3.

Figure 3. Gas sensing testing device. Figure Figure 3. 3. Gas Gas sensing sensing testing testing device. device.

The synthetic air, whose flow rate was set as 250 mL/min, consisted of N2 and O2 at the volume The synthetic synthetic air, air, whose whoseflow flowrate ratewas wasset setas as250 250mL/min, mL/min, consisted consisted of of N N22 and and O O22 at the volume ratio of 4:1. To match with the synthetic air, the volume ratio of the 1000 ppm H2 in N2 and O2 was ratio of 4:1. 4:1. To To match match with with the the synthetic synthetic air, air,the thevolume volumeratio ratioof ofthe the1000 1000ppm ppmHH2 2ininNN2 2and andO O22 was also set as 4:1 with the flow rate of 200 mL/min and 50 mL/min, respectively. During the testing also set as as 4:1 4:1 with with the the flow flowrate rateofof200 200mL/min mL/min and and 50 50mL/min, mL/min, respectively. respectively. During the testing process, the synthetic air was replenished by adjusting the four-way valve. The four-way valve is process, byby adjusting thethe four-way valve. TheThe four-way valvevalve is first process, the thesynthetic syntheticair airwas wasreplenished replenished adjusting four-way valve. four-way is first turned to let the hydrogen in when the response was stabilized. When the response was turned to let the hydrogen in when in thewhen response stabilized. the response was response stabilized,was the first turned to let the hydrogen the was response was When stabilized. When the stabilized, the four-way valve is turned to lead the synthetic air to go through the cylinder until the four-way is turnedvalve to lead synthetic airthe to synthetic go through until sensorsuntil recover stabilized,valve the four-way is the turned to lead airthe to cylinder go through thethe cylinder the sensors recover from the hydrogen. The response transients of the gas sensors to 1000◦ ppm H2 at from therecover hydrogen. transients of the transients gas sensors 1000 H2 at is shown sensors fromThe theresponse hydrogen. The response ofto the gasppm sensors to 400 1000Cppm H2 at 400 °C is shown in Figure 4. It is obvious that all the gas sensors exhibit stable and quick response. in 4. It is in obvious the gas sensors quickstable response. In theresponse. process 400Figure °C is shown Figurethat 4. Itall is obvious that all exhibit the gas stable sensorsand exhibit and quick In the process of dynamic measurement, the SD-101 gas sensing performance testing device was of measurement, SD-101 gas sensing performance testing device was placeddevice in a cubic In dynamic the process of dynamicthe measurement, the SD-101 gas sensing performance testing was placed in a cubic evaporated cavity for 50 L. During the testing process, the corresponding evaporated 50 L. During the testing the the corresponding quantities of the organic placed in acavity cubicfor evaporated cavity for 50 process, L. During testing process, the corresponding quantities of the organic solution (ethanol and benzene) were injected by a micro-injector on a solution (ethanol benzene) were(ethanol injected by a heating in the evaporated quantities of the and organic solution anda micro-injector benzene) wereoninjected bypanel a micro-injector on a heating panel in the evaporated cavity, when the gas sensors responses to air stabilized. When the cavity, responses air stabilized. When the response test gas stabilized, heatingwhen panelthe in gas the sensors evaporated cavity,towhen the gas sensors responses to to airthe stabilized. When the response to the test gas stabilized, the cubic testing cavity was opened for recovery. the cubic testing cavity openedthe forcubic recovery. response to the test gas was stabilized, testing cavity was opened for recovery.

Figure 4. Response transients of the gas sensors to 1000 ppm H2 at 400 °C. Figure 4. 4. Response Response transients transients of of the the gas gas sensors sensors to to 1000 1000 ppm ppm H H2 at at 400 400 ◦°C. Figure C. 2

The response is defined as Ra/Rg, where Ra and Rg are the sensor resistances in air and in the The response is defined as Ra/Rg, where Ra and Rg are the sensor resistances in air and in the The respectively. response is defined as Ra/Rg, Ra anddefined Rg are the sensor and in the test gas, The response timewhere is generally as the timeresistances necessaryin forairachieving a test gas, respectively. The response time is generally defined as the time necessary for achieving a test gas, respectively. The response time is generally defined as the time necessary for achieving a

Sensors 2017, 17, 2351 Sensors 2017, 17, 2351

5 of 17 5 of 17

90% resistance change to the steady-state value. The recovery time is defined as the time for sensor 90% resistance change to the steady-state value. The recovery time is defined as the time for sensor resistance to reach 90% of air resistance. resistance to reach 90% of air resistance.

3. Results and Discussion 3. Results and Discussion

3.1. Characterization of the c-SnO2 Powders and the m-SnO2 Powders 3.1. Characterization of the c-SnO2 Powders and the m-SnO2 Powders

Figure 5 shows XRD patterns of the c-SnO2 powders (Figure 5a) and the m-SnO2 powders Figure shows XRD patterns of the c-SnO2 powders (Figure 5a) and the m-SnO2 powders (Figure 5b). The5 c-SnO 2 powders have peaks corresponding to the SnO2 crystalline phase (PDF (Figure 5b). The c-SnO2 powders have peaks corresponding to the SnO2 crystalline phase (PDF 41-1445). This implies that the c-SnO2 powders are well-crystallized, and have a tetragonal SnO2 phase. 41-1445). This implies that the c-SnO2 powders are well-crystallized, and have a tetragonal SnO2 The crystallite size of the c-SnO2 , calculated by Scherrer’s equation (Jade), is about 65.5 nm. It is also phase. The crystallite size of the c-SnO2, calculated by Scherrer’s equation (Jade), is about 65.5 nm. It confirmed the SEMbyimage (Figure The XRD pattern of the of m-SnO (Figure 5b) shows 2 powders is alsoby confirmed the SEM image6). (Figure 6). The XRD pattern the m-SnO 2 powders (Figure 5b) that they have peaksmain corresponding to SnO2 crystalline phase. phase. It reveals that the prepared shows thatsome they main have some peaks corresponding to SnO2 crystalline It reveals that the m-SnO powders have low crystallinity. In addition, the ordered mesoporous structure is confirmed prepared m-SnO 2 powders have low crystallinity. In addition, the ordered mesoporous structure is 2 clearly by the TEM observation of the m-SnO 2 powders Figure The pore size clearlyconfirmed by the TEM observation of the m-SnO2 powders in Figure 7. Theinpore size 7. distribution and the distribution andofthe area are of the m-SnO powders in Figure 8. It is2 clear specific surface area thespecific m-SnOsurface shown in2 Figure 8. are It isshown clear that the m-SnO powders 2 powders 2 2 the m-SnO a large SSA ofpore 262.30 m of /g with a small pore size of 2.6 nm. show that a large SSA of2 powders 262.30 mshow /g with a small size 2.6 nm.

FigureFigure 5. XRD of theof c-SnO powders and the (a) c-SnO 5. patterns XRD patterns the 2c-SnO 2 powders andm-SnO the m-SnO 2 powders.: (a) c-SnO 2; (b) andm-SnO (b) 2 . 2 powders.: 2 ; and m-SnO2.

Sensors 2017, 17, 2351

6 of 17

Sensors 2017, 17, 2351

6 of 17

Sensors 2017, 17, 2351 Sensors 2017, 17, 2351

6 of 17 6 of 17

Figure 6. SEM image of the c-SnO2 powders.

Figure 6. SEM image of the c-SnO2 powders. Figure 6. SEM image of the c-SnO2 powders. Figure 6. SEM image of the c-SnO2 powders.

Figure 7. TEM image of the m-SnO2 powders. Figure 7. TEM image of the m-SnO2 powders.

Figure 7. 7.TEM m-SnO2 2powders. powders. Figure TEMimage image of of the m-SnO

Figure 8. Pore size distribution of the m-SnO2 powders. Figure 8. Pore size distribution of the m-SnO2 powders.

Figure 8. Pore size distribution of the m-SnO2 powders. 3.2. Characterization of Gas Sensors Figure 8. Pore size distribution of the m-SnO powders. 2

3.2. Characterization Gas Sensors Figure 9 showsof SEM images of the surface morphology of the gas sensors. It was found that 3.2. Characterization ofthe Gas Sensors the basic layers of the sensor were dense and morphology have flat surfaces Moreover, the Figure 9 shows theS(c) SEM images of the surface of the(see gas Figure sensors.9a). It was found that 3.2. Characterization of Gas Sensors Figure 9 shows the SEM images of the surface morphology of the gas sensors. It was found that the basic layers of the S(c) sensor were dense and have flat surfaces (see Figure 9a). Moreover, the the basic layers of theSEM S(c) images sensor were dense and morphology have flat surfaces (seegas Figure 9a). Moreover, the that Figure 9 shows the of the surface of the sensors. It was found

the basic layers of the S(c) sensor were dense and have flat surfaces (see Figure 9a). Moreover, the

Sensors 2017, 17, 2351

7 of 17

Sensors 2017, 17, 2351

7 of 17

S(c) sensor appeared to have a particle size of approximately dozens of nanometers, despite a small appeared to have a particle size of approximately dozens of nanometers, despite a small quantityS(c) of sensor lager particles. The calcination resulted in some sintered macropores with a size of several quantity of lager particles. The calcination resulted in some sintered macropores with a size of hundred nanometers. In contrast, the film of the S(m) sensor show rough and loosened surfaces, as several hundred nanometers. In contrast, the film of the S(m) sensor show rough and loosened shown in Figureas9b. It is obvious sensor filmS(m) showed (100–200 nm) than surfaces, shown in Figure that 9b. Itthe is S(m) obvious that the sensorlager film particles showed lager particles that of the S(c) sensor film due to the agglomerations of the particles. The agglomerations (100–200 nm) than that of the S(c) sensor film due to the agglomerations of the particles. The of the of distinct the m-SnO 2 were extremely distinct from those the of c-SnO 2. Furthermore, the sensor m-SnO2agglomerations were extremely from those of c-SnO calcination of the S(m) 2 . Furthermore, calcination of the S(m) sensor resulted in lager sintered macropores. The surface morphology of the resulted in lager sintered macropores. The surface morphology of the other sensors is similar to those other sensors is similar to those of the S(m) sensor (see Figure 9c–e) because of the same printing of the S(m) sensor (see Figure 9c–e) because of the same printing materials and printing process. materials and printing process.

Figure 9. Cont.

Sensors 2017, 17, 2351

8 of 17

Sensors 2017, 2351 Sensors 2017, 17,17, 2351

8 of817 of 17

Figure 9. SEM images of the surfaces of the gas sensors: (a) S(c); (b) S(m); (c) S(c/m1); (d) S(c/m2);

Figure 9. SEM images of the surfaces of the gas sensors: (a) S(c); (b) S(m); (c) S(c/m1); (d) S(c/m2); and and (e) S(c/m3). Figure 9. SEM images of the surfaces of the gas sensors: (a) S(c); (b) S(m); (c) S(c/m1); (d) S(c/m2); (e) S(c/m3). andThe (e) S(c/m3). SEM images of the cross-sections of the gas sensors are shown in Figure 10. The

cross-sectional morphology of the S(c) sensor shows that the calcined c-SnO2 was more compact

The SEM images of the of cross-sections of the gas sensors are shown inshown Figure in 10.Figure The cross-sectional ThetheSEM images the5 μm cross-sections of the sensors are 10. The with thickness of about (see Figure 10a), butgas the modified layers of the m-SnO 2 showed morphology of the S(c) sensor shows that the calcined c-SnO was more compact with the thickness of 2 cross-sectional morphology of the S(c) sensor shows that the calcined c-SnO 2 was more compact relatively loosened morphology (see Figure 10b). It is apparent that there is an obvious stratification aboutwith 5between µm 10a), modified layers of the m-SnO showed relatively loosened the (see thickness of 2about 5layer μmthe (see Figure 10a), the modified layers of the 2 showed theFigure c-SnO basic but and the m-SnO 2 but modified layer (see 10c).m-SnO In addition, 2Figure fabricated with the same and printing m-SnO 2 modified layers had no relatively loosened morphology Figure 10b). It ismanner, apparent that there is an obvious stratification morphology (see Figure 10b). materials It (see is apparent that there isthe an obvious stratification between the stratification to each other. The thickness of each m-SnO 2 modified layer was confirmed with SEM between the c-SnO 2 basic layer and the m-SnO 2 modified layer (see Figure 10c). In addition, c-SnO2 basic layer and the m-SnO2 modified layer (see Figure 10c). In addition, fabricated with observation to be 10–15 µ m. Thus, the thickness of the m-SnO 2 modified layers oflayers the S(c/m1), fabricated with theabout same materials andthe printing the m-SnO 2 modified had no the same materials and printing manner, m-SnOmanner, 2 modified layers had no stratification to each S(c/m2) and sensors confirmed to m-SnO be 15 µ2 m, 31 µ m layer and 41 µ m, respectively stratification toS(c/m3) each other. Thewere thickness of each modified was confirmed with(see other. The thickness of each m-SnO2 modified layer was confirmed with SEM observation toSEM be about Figure 10c–e). observation to be about 10–15 µ m. Thus, the thickness of the m-SnO2 modified layers of the S(c/m1), 10–15 µm. Thus, the thickness of the m-SnO2 modified layers of the S(c/m1), S(c/m2) and S(c/m3) S(c/m2) and S(c/m3) sensors were confirmed to be 15 µ m, 31 µ m and 41 µ m, respectively (see sensors were confirmed to be 15 µm, 31 µm and 41 µm, respectively (see Figure 10c–e). Figure 10c–e).

Figure 10. Cont.

Sensors 2017, 17, 2351 Sensors 2017, 17, 2351

9 of 17 9 of 17

Figure 10. SEM images of the cross-section of the gas sensors: (a) S(c); (b) S(m); (c) S(c/m1); (d)

Figure 10. SEM images of the cross-section of the gas sensors: (a) S(c); (b) S(m); (c) S(c/m1); (d) S(c/m2); S(c/m2); and (e) S(c/m3). and (e) S(c/m3).

Sensors 2017, 17, 2351 Sensors 2017, 17, 2351

10 of 17 10 of 17

3.3. The Resistance of the Gas Sensors in Air Figure 11 shows the temperature dependence dependence of the resistances of the gas sensors in air. As a semiconductor material,the the resistance of SnO 2 shows decrement trend dueincrease to the ofincrease semiconductor material, resistance of SnO decrement trend due to the carriers of at 2 shows carriers at the condition of thermal excitation, as confirmed in Figure 11. The higher is the operating the condition of thermal excitation, as confirmed in Figure 11. The higher is the operating temperature temperature thework, gas sensors work, theresistances lower is the in air. in which the in gaswhich sensors the lower is the inresistances air.

Figure of the the gas gas sensors sensors in in air. air. Figure 11. 11. Temperature Temperature dependence dependence of of the the resistances resistances of

The values of the resistance in air of the S(c) sensor were slightly decreased due to the frit. The The values of the resistance in air of the S(c) sensor were slightly decreased due to the frit. resistance of the S(m) sensor in air was much higher, which leads to difficult measurement problem The resistance of the S(m) sensor in air was much higher, which leads to difficult measurement in its application. It can be ascribed to the mesoporous structure, which leads to the extreme problem in its application. It can be ascribed to the mesoporous structure, which leads to the extreme decrease of conductive path [21]. However, the resistance of the S(m) sensor in air decreased decrease of conductive path [21]. However, the resistance of the S(m) sensor in air decreased obviously obviously when the operating temperature increased to 400 °C, owing to the condition of thermal when the operating temperature increased to 400 ◦ C, owing to the condition of thermal excitation [24]. excitation [24]. Using the m-SnO2 as the modified layers, the resistances of the (S(c/m1), S(c/m2) and Using the m-SnO2 as the modified layers, the resistances of the (S(c/m1), S(c/m2) and S(c/m3)) sensors S(c/m3)) sensors changed obviously. Since the resistance of SnO2 semiconductors was affected by changed obviously. Since the resistance of SnO2 semiconductors was affected by thermal excitation, thermal excitation, the resistance of the (S(c/m1), S(c/m2) and S(c/m3)) sensors in air decreased the resistance of the (S(c/m1), S(c/m2) and S(c/m3)) sensors in air decreased remarkably when the remarkably when the operating temperature increased to 300 °C. As for the S(c/m1), S(c/m2) and operating temperature increased to 300 ◦ C. As for the S(c/m1), S(c/m2) and S(c/m3) sensors, the S(c/m3) sensors, the resistance of the S(c/m2) sensor in air appeared to be the lowest at all the tested resistance of the S(c/m2) sensor in air appeared to be the lowest at all the tested operating temperature. operating temperature. The value of the resistance in air of the S(c/m2) sensor was 3.7 × 105 Ω at The value of the resistance in air of the S(c/m2) sensor was 3.7 × 105 Ω at 400 ◦ C. The above results 400 °C. The above results demonstrate that both thermal excitation and adsorption affect the demonstrate that both thermal excitation and adsorption affect the resistance of MOS gas sensors in resistance of MOS gas sensors in air. The oxygen adsorbates were considered as the main reason to air. The oxygen adsorbates were considered as the main reason to change the resistance of MOS gas change the resistance of MOS gas sensors in air. The absorbed O2 on the surface− of SnO 2 films sensors in air. The absorbed O2 on the surface of SnO2 films implies the formation of O or O2− , which implies the formation of O− or O2−, which result in a decrease in the quantity of carrier. Thus, the result in a decrease in the quantity of carrier. Thus, the resistance of the sensors fabricated with the resistance of the sensors fabricated with the modified m-SnO2 layers in air increased, due to the modified m-SnO2 layers in air increased, due to the marked improvement of the adsorption capacity of marked improvement of the adsorption capacity of surface oxygen. However, since the increase of surface oxygen. However, since the increase of the modified m-SnO2 layers resulted in larger distance the modified m-SnO2 layers resulted in larger distance for the oxygen adsorbates diffusing to the basic c-SnO2 layer, as well as the diffusion inhibition of mesoporous to oxygen, the S(c/m3) sensor showed lower resistance in air than the S(c/m2) sensor.

Sensors 2017, 17, 2351

11 of 17

Sensors 17, 2351 11 ofof 17 for the2017, oxygen adsorbates diffusing to the basic c-SnO2 layer, as well as the diffusion inhibition mesoporous to oxygen, the S(c/m3) sensor showed lower resistance in air than the S(c/m2) sensor. 3.4. Sensing Responses to the Testing Gas 3.4. Sensing Responses to the Testing Gas The temperature dependence of the responses to ethanol, benzene and hydrogen are depicted in Figure 12. The response of the sensor totoethanol 10 ppmand increased slightly with the The temperature dependence of S(c) the responses ethanol,atbenzene hydrogen are depicted in increasing of operating up toto400 °C, asatshown Figure 12a. However, theincreasing responses Figure 12. The response temperature of the S(c) sensor ethanol 10 ppminincreased slightly with the ◦ C, as to operating ethanol of temperature the S(c) sensor lower than thoseinofFigure the S(c/m1), S(c/m2) and S(c/m3) sensors, due of upwere to 400 shown 12a. However, the responses to ethanol 2/g) to the large specific surface area (262.30 of S(c/m1), the m-SnO 2. In addition, the S(c/m3) sensor of S(c) sensor were lower than thosemof the S(c/m2) and S(c/m3) sensors, dueshowed to the 2 /g) the largest response = 11.4)mto ethanol 300 °C. responsesthe to S(c/m3) benzene sensor of the showed S(c/m1), large specific surface(Ra/Rg area (262.30 of theatm-SnO In addition, 2 . The ◦ S(c/m2) and S(c/m3)(Ra/Rg sensors=showed similarattendency: theresponses response decreased the largest response 11.4) to aethanol 300 C. The to benzeneatofthe the relatively S(c/m1), low operating temperature to a aslight effect of thermal diffusion. While the reaction between S(c/m2) and S(c/m3) sensorsdue showed similar tendency: the response decreased at the relatively low testing gastemperature and the basic 2 waseffect controlled by gas absorption, to benzene was operating duec-SnO to a slight of thermal diffusion. Whilethe theresponse reaction between testing improved increasing of operating temperature from °C to to 400 °C mainly to the gas and thewith basicthe c-SnO by gas absorption, the 300 response benzene was due improved 2 was controlled ◦ C mainly thermal diffusion. of The S(c/m3)temperature sensor exhibited the◦ largest (Ra/Rg 4.31) to benzene at with the increasing operating from 300 C to 400response due to =the thermal diffusion. ◦ 200 S(c/m3) °C, as shown inexhibited Figure 12b. response to hydrogen from 200 400 (Figure The sensor theThe largest response (Ra/Rg =increased 4.31) to benzene at °C 200to C, as°C shown in 12c), while the S(c/m2) showedincreased the largest response (Ra/Rg 22.2) to 12c), hydrogen °C. It Figure 12b. The responsesensor to hydrogen from 200 ◦ C to 400 ◦ C= (Figure while at the400 S(c/m2) ◦ C.molecular can be showed deducedthe from the response hydrogen(Ra/Rg molecular diffusion that the small dimension of sensor largest = 22.2) to hydrogen at 400 It can be deduced from hydrogen benefits the gas diffusion. gas adsorbing was the response to the hydrogen molecular diffusion thatWhile the small molecular capacity dimension of enhanced, hydrogen benefits the gas hydrogenWhile was highly improved (see was Figure 12c). All of the above resultswas confirmed that the diffusion. gas adsorbing capacity enhanced, the response to hydrogen highly improved modified of the m-SnO 2 contribute to improving themodified responselayers of theofS(c) sensor 2sufficiently. (see Figurelayers 12c). All of the above results confirmed that the the m-SnO contribute Moreover, thethe magnitude enhancement not directlythe proportional of to improving responseofofthe theresponse S(c) sensor sufficiently.is Moreover, magnitudetoofthe theamount response the m-SnO2. is not directly proportional to the amount of the m-SnO2 . enhancement

Figure 12. Cont.

Sensors 2017, 17, 2351 Sensors 2017, 17, 2351

12 of 17 12 of 17

Figure 12. Temperature Temperaturedependence dependenceofofthe theresponses responsesofof sensors: ethanol at ppm; 10 ppm; thethe gasgas sensors: (a) (a) ethanol at 10 (b) (b) benzene at 10 ppm; and (c) hydrogen at 1000 ppm. benzene at 10 ppm; and (c) hydrogen at 1000 ppm.

To further further investigate investigate the the effects effects of of the the modified modified m-SnO m-SnO22 layers, layers, the the evolutions evolutions of the response response To of the ◦ versus thickness of the modified films of the gas sensors at 400 °C are depicted in Figure It is versus thickness of the modified films of the gas sensors at 400 C are depicted in Figure 13. It13. is clear clear that the response of the S(c/m1), S(c/m2) and S(c/m3) sensors to ethanol, benzene and hydrogen were all improved to a certain extent in comparison to the S(c) sensor response, which means higher gas sensitivities. Especially, the response of the S(c/m2) sensor to hydrogen appeared

Sensors 2017, 17, 2351

13 of 17

Sensors 2017, 17, 2351 that the response

13 of 17 of the S(c/m1), S(c/m2) and S(c/m3) sensors to ethanol, benzene and hydrogen were all improved to a certain extent in comparison to the S(c) sensor response, which means higher to by 11.4 times, compared to benzene and ethanol of 2.03 and 2.18 times, gasimprove sensitivities. Especially, the response of the S(c/m2) sensorimprovements to hydrogen appeared to improve by respectively. 11.4 times, compared to benzene and ethanol improvements of 2.03 and 2.18 times, respectively.

Figure 13. Evolutions Evolutionsofofthe theresponse response versus thickness of modified the modified the gas sensors Figure 13. versus thickness of the filmsfilms of theofgas sensors at 400 ◦at C. 400 The thickness of the modified films the gas sensors: S(c),S(c/m1), 0 μm; S(c/m1), μm; S(c/m2), The°C. thickness of the modified films with thewith gas sensors: S(c), 0 µm; 15 µm; 15 S(c/m2), 31 µm; 31 μm; and S(c/m3), and S(c/m3), 41 µm.41 μm.

3.5. 3.5. The The Response Response and and Recovery Recovery Times Times of of the the Gas Gas Sensors Sensors Table shows the response and sensors to Table 22 shows the response and recovery recovery times times of of the the gas gas sensors to ethanol, ethanol, benzene benzene and and hydrogen. The response and recovery times of some of the gas sensors were difficult to summarize hydrogen. The response and recovery times of some of the gas sensors were difficult to summarize due due tolower the lower response low operating temperature. The response of the S(c) to sensor to to the response at lowatoperating temperature. The response time oftime the S(c) sensor ethanol ethanol was markedly short from 350 °C (response time = 115 s) to 400 °C (response time = 54 s), ◦ ◦ was markedly short from 350 C (response time = 115 s) to 400 C (response time = 54 s), while the while values oftime response time toathydrogen at 400 350 ◦°C and 400 were s and 89 s, However, respectively. valuesthe of response to hydrogen 350 ◦ C and C were 80 s°C and 89 s,80 respectively. the However, the response times benzene were hard to summarize because ofmolecular the relatively large response times of benzene wereof hard to summarize because of the relatively large dimension molecular dimension which ledtotobenzene. the low In response In theS(c/m2) case of and the (0.65–0.68 nm), which(0.65–0.68 led to the nm), low response the casetoofbenzene. the S(c/m1), S(c/m1), S(c/m2) the andresponse S(c/m3) times sensors, the response thelayers thicker S(c/m3) sensors, increased with the times thickerincreased modified with m-SnO formodified all of the 2 m-SnO 2 layers for all of the tested gases at 350 °C and 400 °C. In addition, the smallest response ◦ ◦ tested gases at 350 C and 400 C. In addition, the smallest response time appeared to be 74 s with time appeared to be 74 s with the S(c/m1) sensors to hydrogen at 400 °C. It can be ascribed to the ◦ the S(c/m1) sensors to hydrogen at 400 C. It can be ascribed to the large specific surface area and the large specific surface the molecular at high temperature, of the which lead to molecular diffusion at area high and temperature, both ofdiffusion which lead to easy gas diffusionboth inside mesopores. easy gas diffusion inside the longer mesopores. All gas sensors showed recovery times to ethanol and benzene in comparison to hydrogen All gas sensors showed longer recoveryS(c/m2) times and to ethanol benzene intocomparison to ◦ ◦ from 350 C to 400 C. Especially, the S(c/m1), S(c/m3)and sensors tended show a longer hydrogen from(>500 350 s) °Ctotohydrogen 400 °C. Especially, the S(c/m1), S(c/m2) and (600 >600 403 478 293

S(c) 14 26 37 -

S(m) 90 271 201 -

Recovery Time (s) S(c/m1) S(c/m2) 292 519 281 355 178 182 246 235 467 467 >600 >600 60 158 14 12 89 63 207 103

S(c/m3) >600 349 142 226 462 >600 >600 13 78 115

Sensors 2017, 17, 2351

14 of 17

Table 2. The response and recovery times of the gas sensors to ethanol, benzene and hydrogen. Gas

Temperature (◦ C)

Ethanol

200 250 300 350 400

Sensors 2017, 17, 2351 200 250 300 200 350 250 400

Benzene

Hydrogen Hydrogen

300 200 350 250 300 400 350 400

Response Time (s)

Recovery Time (s)

S(c)

S(m)

S(c/m1)

S(c/m2)

126 115 54

19 41 56

286 252 197 127 152

496 277 106 289 210

549 180 87 430 295

>600 360 110 344 411 261 173

>600 486 225 343 422 162 218

>600 >600 403 349 478 146 293

96344 87261 7496

82 343 110 162 82 105

87 349 121 146 87 110 121 110

-

- - -

83 - 131 80 - 144 89 83 34131 80 89

144 34

87 74

110 105

S(c/m3)

S(c)

S(m)

S(c/m1)

S(c/m2)

S(c/m3)

14 26 37

90 271 201

292 281 178 246 467

519 355 182 235 467

>600 349 142 226 462

---

---

>600 60 14 84 89 387 207

>600 158 12 216 63 406 103

>600 14 of 17 >600 13 78233 115404

-194 -165 194 139 165 139

>600 >600 >600 >600

423 84 568 387 423 507 568 507

451 216 >600 406 451 519 >600 519

233442 >600 404 442520 >600 520

>600 >600

3.6. Discussion 3.6. Discussion Among these, the possible gas sensing mechanism of the gas sensors are shown in Figure 14. It is considered that the thepossible adsorption/desorption properties ofgas thesensors mesoporous influenced gas Among these, gas sensing mechanism of the are shown in Figurethe 14. It is sensing performances of the gas sensors. Owing to theoflarge specific surface area of m-SnO which considered that the adsorption/desorption properties the mesoporous influenced the gas2,sensing could enhance of thethe adsorption of gas molecules, the S(c/m1) sensors exhibited performances gas sensors. Owing to the large specificS(c/m2) surfaceand areaS(c/m3) of m-SnO 2 , which could higher gas (ethanol, benzene, and hydrogen) responses than those of the S(c) sensor. The responses enhance the adsorption of gas molecules, the S(c/m1) S(c/m2) and S(c/m3) sensors exhibited higher of are lower than of ethanol and than hydrogen ofsensor. its weak reducibility larger gasbenzene (ethanol, benzene, and those hydrogen) responses those because of the S(c) The responses and of benzene size of benzene ring,which is difficult to pass through the m-SnO 2 modified layer. However, the are lower than those of ethanol and hydrogen because of its weak reducibility and larger size of sintered 14)toamong the m-SnO is helpful to adsorb more benzene molecules. benzene macropores ring, which (Figure is difficult pass through the2m-SnO 2 modified layer. However, the sintered This is why (Figure the responses of the and S(c/m3)more sensors to benzene are higher macropores 14) among theS(c/m1), m-SnO2 S(c/m2) is helpful to adsorb benzene molecules. This is than why those of the S(c) sensor to benzene. In contrast, the smaller molecular size of the hydrogen is the responses of the S(c/m1), S(c/m2) and S(c/m3) sensors to benzene are higher than those of the S(c) beneficial to pass In through thethe ordered m-SnO and the sintered macropores. Thus, sensor to benzene. contrast, smallerstructure molecularofsize of the2 hydrogen is beneficial to pass through thickness of the films and the ordered level of mesoporous influenced the gas sensing performance the ordered structure of m-SnO2 and the sintered macropores. Thus, thickness of the films and the of the gas sensors fabricated with the modified layers. In addition, further approaches to ordered level of mesoporous influenced the gas m-SnO sensing2 performance of the gas sensors fabricated control amountm-SnO of sintered macropores, the thickness of the films, surface contact of the films with thethe modified 2 layers. In addition, further approaches to control the amount of sintered and the ordered level of mesoporous effective to films improve the ordered sensing level performance of the macropores, the thickness of the films, would surfacebe contact of the and the of mesoporous gas sensors. would be effective to improve the sensing performance of the gas sensors.

Figure Figure 14. 14. Schematic Schematic drawing drawing of of the the possible possible gas gas sensing sensing mechanism mechanism of of the gas sensors.

Experimental results of ofSnO SnO2sensors sensorshave have been compared with results reported byother the Experimental results been compared with the the results reported by the 2 other researchers on H 2 sensors. Manjula et al. reported the Pd doped m-SnO2 gas sensors. It researchers on H2 sensors. Manjula et al. reported the Pd doped m-SnO2 gas sensors. It showed showed that the 0.25% Pd doped gas sensor response towards 1000 ppm hydrogen at 50 °C is 0.95. The gas sensors showed zero response to ethanol, LPG, NH3 and acetone [25]. Seftel et al. obtained gas sensing material by combining Pt with SnO2 or In2O3 based on SBA-15. The response of the gas sensor based on Pt/SnO2/SBA-15 is about 1.4 to 1000 ppm hydrogen at 350 °C [26]. Although the selectivity of the sensors was improved by doping, the responses of the sensors to hydrogen are no

Sensors 2017, 17, 2351

15 of 17

that the 0.25% Pd doped gas sensor response towards 1000 ppm hydrogen at 50 ◦ C is 0.95. The gas sensors showed zero response to ethanol, LPG, NH3 and acetone [25]. Seftel et al. obtained gas sensing material by combining Pt with SnO2 or In2 O3 based on SBA-15. The response of the gas sensor based on Pt/SnO2 /SBA-15 is about 1.4 to 1000 ppm hydrogen at 350 ◦ C [26]. Although the selectivity of the sensors was improved by doping, the responses of the sensors to hydrogen are no more than 2, which limits the applications of the sensors in hydrogen measurement. We can find a large number of examples of sensitivities improvement of gas sensors to hydrogen when mesoporous structures are employed. Shen et al. reported that the influence of the different morphology of SnO2 nanomaterials on hydrogen sensing properties. They obtained the response of about 2.1 to 1000 ppm hydrogen at 250 ◦ C for nanofilms [27]. Yeow et al. reported the gas sensors based on SnO2 nanospheres with various degrees porosity. The reference (SSA SnO2 = 101.4 m2 /g) gas sensor showed the largest response: 5.2 to 500 ppm hydrogen at 350 ◦ C [28]. Zhao et al. prepared ordered mesoporous SnO2 and mesoporous Pd/SnO2 via nanocasting method using the hexagonal mesoporous SBA-15 as template. The maximum response of the sensor based on the ordered mesoporous SnO2 is 16.4 to 1000 ppm hydrogen at 300 ◦ C [29]. Hayashi et al. prepared SnO2 gas sensors based on various m-SnO2 powders from two kinds of combination of tin source and surfactant template. The largest response of the gas sensors to 1000 ppm hydrogen at 350 ◦ C appeared to be 42 [30]. It is evidential that the responses of the sensors based on the ordered mesoporous SnO2 to hydrogen have been dramatically increased. The one limitation of these studies is that the selectivity of the mesoporous SnO2 has not been studied. Shahabuddin et al. reported the sputter deposited SnO2 thin film gas sensors with 9 nm thin Pt clusters. The Pt/SnO2 sensor shows an improvement in sensing response: 168 towards 500 ppm of hydrogen at 110 ◦ C. The sensor revealed negligible cross sensing signals against acetone, IPA, NO2 , methane, LPG, etc. [10]. Gong et al. reported the mesoporous nanocrystalline SnO2 gas sensor based on the fabricated SnO2 sputtering with Pt thin film. The gas sensor showed the response of about 1.8 to 1000 ppm hydrogen at 250 ◦ C [31]. The sensitivities and selectivity of the SnO2 gas sensors could be significantly improved by sputtering with Pt thin film. However, the method requires an expensive facility and complex sample preparing process. In our work, the ordered mesoporous SnO2 was prepared by simple sol-gel method. The gas sensors were prepared with a simple and low cost screen printing method while the mesoporous SnO2 worked as the modified layers. It was shown that both the sensitivities and the selectivity of the gas sensors to hydrogen were improved. The S(c/m2) sensor showed the largest response 22.2 to 1000 ppm hydrogen at 400 ◦ C. The response to hydrogen is >10 times higher than that of the sensor without the modified layer (the S(c) sensor). Compared with the responses of the S(c) sensor, the responses of the S(c/m2) sensor to benzene and ethanol did not change significantly. 4. Conclusions Ordered mesoporous SnO2 powders were prepared by employing Na2 SnO3 ·4H2 O, C16 PyCl and trimethylbenzene. The specific surface area of the m-SnO2 powder was 262.30 m2 /g after calcination at 600 ◦ C. The gas sensors were fabricated using m-SnO2 films as the modified layers. It was proven that the gassensing performance of the gas sensors could be highly improved, especially to hydrogen, compared with ethanol or benzene gas. In addition, the S(c/m2) sensor exhibited the highest sensitivity (response: Ra/Rg = 22.2) to 1000 ppm hydrogen at 400 ◦ C. The main reason for the high selectivity may be the diffusivity of hydrogen molecules in the ordered mesopores is easier than that of ethanol and benzene molecules. Supplementary Materials: The Supplementary Materials are available online at http://www.mdpi.com/14248220/17/10/2351/s1, Figure S1: The substrate structure of the gas sensor, Figure S2: The SD-101 gas sensing performance testing device, Figure S3: Gas sensing testing device, Figure S4: Response transients of gas sensors to 1000 ppm H2 at 400 ◦ C, Figure S5: XRD patterns of c-SnO2 powders and m-SnO2 powders. (a) c-SnO2 ; and (b) m-SnO2 , Figure S6: SEM image of c-SnO2 powders, Figure S7: TEM image of m-SnO2 powders, Figure S8: Pore size distribution of m-SnO2 powders, Figure S9: SEM images of the surfaces of gas sensors: (a) S(c); (b) S(m);

Sensors 2017, 17, 2351

16 of 17

(c) S(c/m1); (d) S(c/m2); and (e) S(c/m3), Figure S10: SEM images of the cross-section of gas sensors: (a) S(c); (b) S(m); (c) S(c/m1); (d) S(c/m2); and (e) S(c/m3), Figure S11: Temperature dependence of the resistances of the gas sensors in air, Figure S12: Temperature dependence of the responses of the gas sensors (a) ethanol at 10 ppm; (b) benzene at 10 ppm; and (c) hydrogen at 1000 ppm, Figure S13: Evolutions of the response versus thickness of the modified films of the gas sensors at 400 ◦ C. The thickness of the modified films with the gas sensors: S(c), 0 µm; S(c/m1), 15 µm; S(c/m2), 31 µm; and S(c/m3), 41 µm, Figure S14: Schematic drawing of the possible gas sensing mechanism of the gas sensors, Table S1: The structure of the gas sensors, Table S2: the response and recovery times of the gas sensors to ethanol, benzene and hydrogen. Acknowledgments: This work was founded by the Scientific Research Foundation for Returned Scholars, Ministry of Education of China, and the National Science Foundation of China (Grant No. 51502229). Author Contributions: Qinyi Zhang and Niuzi Xue conceived and designed the experiments; Niuzi Xue, Feng Yang and Pan Zong performed the experiments; Niuzi Xue and Qinyi Zhang analyzed the data; Shunping Zhang contributed the test tools; and Niuzi Xue wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.

References 1.

2.

3. 4.

5. 6.

7.

8.

9. 10. 11.

12.

13.

14.

Cao, S.; Klein, K.; Herkel, S.; Sirén, K. Approaches to enhance the energy performance of a zero-energy building integrated with a commercial-scale hydrogen fueled zero-energy vehicle under finnish and german conditions. Energy Convers. Manag. 2017, 142, 153–175. [CrossRef] Andronov, D.Y.; Arseniev, D.G.; Polyanskiy, A.M.; Polyanskiy, V.A.; Yakovlev, Y.A. Application of multichannel diffusion model to analysis of hydrogen measurements in solid. Int. J. Hydrogen Energy 2017, 42, 699–710. [CrossRef] Bockris, J.O.M. On hydrogen futures: Toward a sustainable energy system. Int. J. Hydrogen Energy 2003, 28, 131–133. [CrossRef] Cardozacontrera, M.N.; Romoherrera, J.M.; Ríos, L.A.; Garcíagutiérrez, R.; Zepeda, T.A.; Contreras, O.E. Single ZnO Nanowire-Based Gas Sensors to Detect Low Concentrations of Hydrogen. Sensors 2015, 15, 30539–30544. [CrossRef] [PubMed] Kim, S.; Choi, J.; Jung, M.; Joo, S.; Kim, S. Silicon carbide-based hydrogen gas sensors for high-temperature application. Sensors 2013, 13, 13575–13583. [CrossRef] [PubMed] Kim, H.W.; Kwon, Y.J.; Mirzaei, A.; Kang, S.Y.; Choi, M.S.; Bang, J.H.; Kim, S.S. Synthesis of zinc oxide semiconductors-graphene nanocomposites by microwave irradiation for application to gas sensors. Sens. Actuators B Chem. 2017, 249, 590–601. [CrossRef] Wongchoosuk, C.; Wisitsoraat, A.; Phokharatkul, D.; Tuantranont, A.; Kerdcharoen, T. Multi-Walled Carbon Nanotube-Doped Tungsten Oxide Thin Films for Hydrogen Gas Sensing. Sensors 2010, 10, 7705–7715. [CrossRef] [PubMed] ˇ Krško, O.; Plecenik, T.; Roch, T.; Granˇciˇc, B.; Satrapinskyy, L.; Truchlý, M.; Durina, P.; Gregor, M.; Kúš, P.; Plecenik, A. Flexible highly sensitive hydrogen gas sensor based on a TiO2 thin film on polyimide foil. Sens. Actuators B Chem. 2017, 240, 1058–1065. [CrossRef] Kimura, Y.; Kimura, S.; Kojima, R.; Bitoh, M.; Abe, M.; Niwano, M. Micro-scaled hydrogen gas sensors with patterned anodic titanium oxide nanotube film. Sens. Actuators B Chem. 2013, 177, 1156–1160. [CrossRef] Shahabuddin, M.; Umar, A.; Tomar, M.; Gupta, V. Custom designed metal anchored SnO2 sensor for H2 detection. Int. J. Hydrogen Energy 2017, 42, 4597–4609. [CrossRef] Nguyen, V.T.; Nguyen, V.C.; Nguyen, V.D.; Hoang, S.H.; Hugo, N.; Nguyen, D.H.; Nguyen, V.H. Fabrication of highly sensitive and selective H2 gas sensor based on SnO2 thin film sensitized with microsized Pd islands. J. Hazard. Mater. 2016, 301, 433–442. [PubMed] Ammar, A.H.; Abo-Ghazala, M.S.; Farag, A.A.M.; Abdel-Moniem, N.M.; Farag, E.-S.M. Effect of gas type, pressure and temperature on the electrical characteristics of Al-doped SnO2 thin films deposited by RGTO method for gas sensor application. Vacuum 2013, 94, 30–40. [CrossRef] Inyawilert, K.; Wisitsoraat, A.; Tuantranont, A.; Phanichphant, S.; Liewhiran, C. Ultra-sensitive and highly selective H2 sensors based on FSP-made Rh-substituted SnO2 sensing films. Sens. Actuators B Chem. 2017, 240, 1141–1152. [CrossRef] Bianchetti, M.F.; Arrieta, C.; Walsöe de Reca, N.E. Microstructural study of nanocrystalline pure and doped tin dioxide to be used for resistive gas sensors. Sens. Actuators B Chem. 2015, 217, 113–118. [CrossRef]

Sensors 2017, 17, 2351

15.

16.

17.

18. 19. 20.

21. 22. 23. 24. 25.

26. 27.

28. 29.

30. 31.

17 of 17

Liewhiran, C.; Tamaekong, N.; Wisitsoraat, A.; Tuantranont, A.; Phanichphant, S. Ultra-sensitive H2 sensors based on flame-spray-made Pd-loaded SnO2 sensing films. Sens. Actuators B Chem. 2013, 176, 893–905. [CrossRef] Jang, B.-H.; Landau, O.; Choi, S.-J.; Shin, J.; Rothschild, A.; Kim, I.-D. Selectivity enhancement of SnO2 nanofiber gas sensors by functionalization with Pt nanocatalysts and manipulation of the operation temperature. Sens. Actuators B Chem. 2013, 188, 156–168. [CrossRef] Yamaguchi, M.; Anggraini, S.A.; Fujio, Y.; Sato, T.; Breedon, M.; Miura, N. Stabilized zirconia-based sensor utilizing SnO2 -based sensing electrode with an integrated Cr2 O3 catalyst layer for sensitive and selective detection of hydrogen. Int. J. Hydrogen Energy 2013, 38, 305–312. [CrossRef] Tournier, G.; Pijolat, C. Selective filter for SnO-based gas sensor: Application to hydrogen trace detection. Sens. Actuators B Chem. 2005, 106, 553–562. [CrossRef] Masuzawa, S.; Okazaki, S.; Maru, Y.; Mizutani, T. Catalyst-type-an optical fiber sensor for hydrogen leakage based on fiber bragg gratings. Sens. Actuators B Chem. 2015, 217, 151–157. [CrossRef] Yaqoob, U.; Uddin, A.S.M.I.; Chung, G.-S. Foldable hydrogen sensor using Pd nanocubes dispersed into multiwall carbon nanotubes-reduced graphene oxide network assembled on nylon filter membrane. Sens. Actuators B Chem. 2016, 229, 355–361. [CrossRef] Hyodo, T.; Abe, S.; Shimizu, Y.; Egashira, M. Gas-sensing properties of ordered mesoporous SnO2 and effects of coatings thereof. Sens. Actuators B Chem. 2003, 93, 590–600. [CrossRef] Pijolat, C.; Viricelle, J.P.; Tournier, G.; Montmeat, P. Application of membranes and filtering films for gas sensors improvements. Thin Solid Films 2005, 490, 7–16. [CrossRef] Dhawale, D.S.; Lokhande, C.D. Chemical route to synthesis of mesoporous ZnO thin films and their liquefied petroleum gas sensor performance. J. Alloys Compd. 2011, 509, 10092–10097. [CrossRef] Das, S.; Jayaraman, V. SnO2 : A comprehensive review on structures and gas sensors. Prog. Mater. Sci. 2014, 66, 112–255. [CrossRef] Manjula, P.; Satyanarayana, L.; Swarnalatha, Y.; Manorama, S.V. Raman and MASNMR studies to support the mechanism of low temperature hydrogen sensing by Pd doped mesoporous SnO2 . Sens. Actuators B Chem. 2009, 138, 28–34. [CrossRef] Seftel, E.M.; Cool, P.; Lloyd-Spetz, A.; Lutic, D. Pt-doped semiconductive oxides loaded on mesoporous SBA-15 for gas sensing. Comptes Rendus Chim. 2014, 17, 717–724. [CrossRef] Shen, Y.; Wang, W.; Fan, A.; Wei, D.; Liu, W.; Han, C.; Shen, Y.; Meng, D.; San, X. SnO2 : Highly sensitive hydrogen sensors based on SnO2 nanomaterials with different morphologies. Int. J. Hydrogen Energy 2015, 40, 15773–15779. [CrossRef] Yeow, S.C.; Ong, W.L.; Wong, A.S.W.; Ho, G.W. Template-free synthesis and gas sensing properties of well-controlled porous tin oxide nanospheres. Sens. Actuators B Chem. 2009, 143, 295–301. [CrossRef] Zhao, J.; Wang, W.; Liu, Y.; Ma, J.; Li, X.; Du, Y.; Lu, G. Ordered mesoporous Pd/SnO2 synthesized by a nanocasting route for high hydrogen sensing performance. Sens. Actuators B Chem. 2011, 160, 604–608. [CrossRef] Hayashi, M.; Hyodo, T.; Shimizu, Y.; Egashira, M. Effects of microstructure of mesoporous SnO2 powders on their H2 sensing properties. Sens. Actuators B Chem. 2009, 141, 465–470. [CrossRef] Gong, J.; Chen, Q.; Fei, W.; Seal, S. Micromachined nanocrystalline SnO2 chenical gas sensors for electronic nose. Sens. Actuators B Chem. 2004, 102, 117–125. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).