Highly selective detection of sulfur hexafluoride ...

Carbon 135 (2018) 95e103

Contents lists available at ScienceDirect

Carbon journal homepage: www.elsevier.com/locate/carbon

Highly selective detection of sulfur hexafluoride decomposition components H2S and SOF2 employing sensors based on tin oxide modified reduced graphene oxide Jifeng Chu a, Xiaohua Wang a, Dawei Wang a, Aijun Yang a, *, Pinlei Lv a, Yi Wu a, **, Mingzhe Rong a, Lei Gao b a b

State Key Laboratory of Electrical Insulation for Power Equipment, Xi'an Jiaotong University, Xi'an, 710049, PR China State Grid Changzhi Power Supply Company, Changzhi, 046000, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 March 2018 Received in revised form 12 April 2018 Accepted 14 April 2018 Available online 16 April 2018

In this paper, a high-performance sensor based on tin oxide modified reduced graphene oxide (SnO2rGO) has been prepared via a one-step hydrothermal process for the detection of sulfur hexafluoride (SF6) decomposition components H2S and SOF2. The morphology and nanostructure of the product were characterized by X-ray diffraction (XRD), Raman spectroscopy (RS), scanning electron microscopy (SEM), and energy dispersive spectrometer (EDS). The gas sensing properties of SnO2-rGO nanocomposite were investigated by exposing to various target gases. The experimental results revealed that SnO2-rGO sensor exhibited better responses (34.31% and 3.13%) than those (5.97% and 1.45%) of pristine rGO sensor at the optimal temperature of 125 C in the presentence 100 ppm H2S and 10 ppm SOF2, respectively. Meanwhile, SnO2-rGO sensor emerged as one of the promising candidate materials to select these two gases by yielding opposite response to H2S and SOF2. In addition, uniformity, repeatability, long-term stability of SnO2-rGO sensors were investigated to demonstrate the practicability, and the possible mechanism of SnO2-rGO nanocomposite for sensing H2S and SOF2 was discussed. © 2018 Elsevier Ltd. All rights reserved.

1. Introduction SF6 has been widely used in gas-insulated switchgear (GIS) because of its excellent properties in insulating and arcextinguishing [1]. However, partial discharges (PDs) will occur with the existence of inevitable intrinsic defects in GIS, causing SF6 decomposes into SO2, SOF2, SO2F2, and H2S under the reaction with trace water and oxygen [2,3]. On one hand, decomposed products of SF6 will destroy the insulation of the equipment; on the other hand, these highly toxic gases are threatened to the health of power workers. Besides, some studies [4e6] showed that types and contents of SF6 decomposition components in GIS reflected the running state of power equipment for further diagnosis of

* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (J. Chu), [email protected] (X. Wang), [email protected] (D. Wang), [email protected] (A. Yang), [email protected] (P. Lv), [email protected] (Y. Wu), [email protected]. cn (M. Rong), [email protected] (L. Gao). https://doi.org/10.1016/j.carbon.2018.04.037 0008-6223/© 2018 Elsevier Ltd. All rights reserved.

insulation faults. So, it's necessary to detect SF6 decomposition components in GIS. Nowadays, many efforts have been developed to analyze the decomposition of SF6 via various techniques, such as infrared absorption spectrometry technology [7], photoacoustic spectroscopy technology [8] and gas chromatography mass spectrometry technology [9]. Comparing to above off-line testing means, gas sensors have great advantages in low cost, easy production, compact size [10e12]. There has been some literature in the field of detecting SF6 decomposed products using gas sensors. Zhang et al. investigated TiO2 nanotube [10], Au-modified rGO nanosheet [12] and CNT [13] to detect H2S, SO2, SO2F2, and SOF2 achieving significant responses, but sensors based on above sensing materials didn't exhibit repeatability. Furthermore, Liu et al. also detected SO2, SO2F2, and SOF2 using NiO-decorated ZnO nanoflower [14] which performed high response with short response-recovery time, but the operating temperature was high about 220 C. Therefore, sensors used to detect SF6 decomposed products with excellent sensing-properties at a relatively low working temperature should be developed. Graphene, the two-dimensional nanosheet based on sp2

96

J. Chu et al. / Carbon 135 (2018) 95e103

hybridized carbon atoms, is a promising material which can be applied in many fields like energy storage, biosensing, and catalysis, because it features high specific surface area, excellent electrocatalytic activity, great mobility and low electrical noise [15e17]. Additionally, graphene has ability to sense gases through transferring charges from/to adsorbed gas molecules, eventually causing a change in electrical conductivity [18]. Reduced graphene oxide (rGO) is the functionalized product of graphene, and it has been considered as a good sensing material due to oxygen functional groups on the surface of rGO providing amounts of active sites for gas molecules to adsorb. Generally, sensors based on rGO can work at a relatively low temperature, but have poor response and long recovery time [19e21]. For traditional chemical gas sensors, various semiconducting metal oxides due to their great responses to target gases are usually used as sensing-materials to detect varieties of dangerous and toxic gases. For example, SnO2 [22], ZnO [23], Cu2O-CuO [24], In2O3 [25], WO3 [26], and TiO2 [27], have been used for detecting nitrogen dioxide(NO2), ethanol, hydrogen sulfide(H2S), carbon monoxide(CO), hydrogen(H2), and ammonia(NH3), respectively. Among them, as a typical n-type wide bandgap (3.6eV) metal oxide semiconductor, SnO2 has attracted more considerable attention. However, the operating temperature of sensors based on SnO2 is usually above 200 C, thereby the high-energy consumption limits its application [28,29]. Recently, the pristine rGO decorated with metal oxides to improve its sensing properties has been proposed. For example, the modification of rGO with SnO2 [30,31], Co3O4 [32], and ZnO [33,34] presented better performances than the pristine rGO. Comparing to previous studies in detecting SF6 decomposition components, we have considered to combine the advantages between SnO2 and rGO, to develop SnO2-rGO sensor to improve repeatability and selectivity. In this work, SnO2-rGO nanocomposite was prepared by onestep hydrothermal synthesis route in which SnO2 particles were easily grown on the surface of rGO sheets. The as-prepared samples were fully characterized by using XRD, RS, SEM, and EDS, confirming the successful preparation of sensing materials. The SnO2rGO nanocomposite was used as gas-sensing material for detecting SF6 decomposition products, exhibiting optimal characteristics at the temperature of about 125 C. It was found that SnO2-rGO sensor exhibited better responses than pristine rGO sensor. Additionally, the underlying sensing mechanism of nanocomposite for the detection of H2S and SOF2 was discussed.

Fig. 1. Schematic illustration for the preparation of SnO2-rGO nanocomposite via hydrothermal treatment. (A colour version of this figure can be viewed online.)

were all added into 40 mL of deionized water with stirring for 30min. Then, the above mixture was transferred into a 100 mL Teflon-lined, stainless-steel autoclave to react at 180 C for 15 h. The resulting precipitates were collected by centrifugation at 3000 rpm for 10min after the autoclave cooled down to room temperature, followed by rinsing with deionized water several times to remove excess chloride ion. Finally, the products, in which the nominal mass fraction of SnO2 was 84.89 wt%, were obtained by freeze drying for 24 h. The pristine rGO was fabricated by the same route without adding SnCl4,5H2O and hydrochloric acid. 2.2. Characterizations The crystal phase and crystallinity of the pristine GO, rGO, and SnO2-rGO nanocomposite were analyzed with X-ray diffraction (Bruker D2 PHASER) using Cu Ka radiation (l ¼ 1.5418 Å) with a scanning range of 13 e80 . The morphology and nanostructure of the samples were examined by scanning electron microscopy (Zeiss GeminiSEM 500). More detailed structural information of the samples was obtained by a laser Raman spectrometer (Renishaw inVia) using a laser wavelength of 633 nm as excitation source.

2. Experimental

2.3. Fabrication and test of gas sensors

All reagents used in the experiment were analytically pure without further purification and purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Graphene oxide (GO) was purchased from Hangzhou Gelanfeng Nano Technology Co., Ltd. (Hangzhou, China). The gases purchased from Dalian Special Gases Co., Ltd. were stored in steel cylinder with various concentration, including 200 ppm calibration gas (in N2) for H2S and SO2, and 20 ppm calibration gas (in He) for SOF2 and SO2F2.

The sensors were fabricated by drop-casting the dispersion of nanomaterials onto a ceramic plate (0.8 cm 1.0 cm), which was coated with interdigital electrodes. The as-prepared materials were dispersed in a mixture containing ionized water and isopropyl alcohol with volume ratio of 1:1 at a concentration of 1 mg/mL, followed by sonicating for 30min. Then, the dispersion was dropcoated on the ceramic substrate, drying in an oven at 60 C for 24 h to facilitate the stability before testing. For investigating the uniformity of sensors, as shown in Fig. S1, three sensors were prepared in a same batch with the same amount of SnO2-rGO nanocomposite, designed as sensor-1, sensor-2, and sensor-3, and the sensor based on pristine rGO was designed as sensor-0 for comparison. The schematic of the experimental instrument for gas sensing was presented in Fig. 2. The sensing performances of the fabricated sensors were tested by a CGS-4TPs intelligent test meter (Beijing Elite Tech. Co., Ltd. China) under the operating temperature ranging from 25 C to 200 C with the relative humidity (RH) being 20%e

2.1. SnO2-rGO nanocomposite synthesis A facile one-step hydrothermal route was used for preparing SnO2-rGO nanocomposite [35,36], and entire procedure for the material synthesis was shown in Fig. 1. In the synthesis, typically, 10 mL homogeneous suspension solution with 5 mg of GO was obtained after ultrasonicating for 30min in deionized (DI) water. After that, 50 mg of SnCl4,5H2O, 10 mL of GO dispersion (0.5 mg/mL), and 0.5 mL of hydrochloric acid (37%)


Fig. 2. Schematic illustration for gas sensing experimental device. (A colour version of this figure can be viewed online.)

25%. The concentrations of each target gas were adjusted by controlling the flow ratio between balance gas (air) and calibration gas, which was achieved by using mass-flow controllers (MFC). The response of the sensor was defined as follows:

DR=R ¼ Ra Rg Ra 100%

(1)

where Ra and Rg represents the resistances of the sensor in dry air and target gas, respectively. The response time (tres) and recovery time (trec) of the sensor are defined as the time taken to reach 90% of the total resistance change. 3. Results and discussion 3.1. Characterization results The XRD patterns for GO, rGO and SnO2-rGO nanocomposite are shown in Fig. 3. In the case of rGO, the characteristic peak is shifted to 24.29 in comparison with GO after hydrothermal reduction, corresponding to the (002) interlayer d spacing of 3.6043 Å [37]. It's easy to find that several strong peaks are emerged at Bragg angles (2q) of 26.61, 33.73 , 51.85 , 65.38 in the XRD spectrum of SnO2rGO sample, which can be attributed to the (110), (101), (211), (112) planes of tetragonal rutile SnO2 (JCPDS Card No.41e1049) [29,38], and it demonstrates the formation of SnO2 crystals. Nevertheless, there is no obvious diffraction peak can be ascribed to rGO in the spectrum of SnO2-rGO nanocomposite, most likely because of the low content of rGO in the composite, and the Bragg peak of rGO at

Fig. 3. XRD spectra of GO, rGO, SnO2-rGO samples. (A colour version of this figure can be viewed online.)

97

(002) plane is covered by the high intensity peak of SnO2 nanoparticles at (110) plane. The presence of rGO in the SnO2-rGO nanocomposite is further investigated by Raman spectra in Fig. 4. The D band at 1342 cm1 accords with the structure defects and partially-disordered structures, and the G band at 1598 cm1 corresponds to the first-order scattering of the E2g mode. These two peaks can be seen in both the pristine rGO and SnO2-rGO nanocomposite, indicating the presence of rGO in the composite. The intensity ratio of D band and G band (ID/IG) increases from 1.07 (GO) to 1.19 (rGO) following the reduction step, indicating partial presence of the graphene structure with decreased average size of sp2 domains. Some studies [39,40] revealed that ID/IG may increase or remain the same after the reduction process of GO. For the SnO2-rGO nanocomposite, ID/IG is estimated to the value of 1.22 which is similar to the spectra of rGO, suggesting the restoration of C¼C bonds. After freeze-drying treatment, the formation of the porous structure on the surface of rGO made gases diffuse and adsorb easily [41,42]. The SEM images of GO, rGO and SnO2-rGO samples are shown in Fig. 5, and we can see the micrometers-long wrinkles of overlapped GO (Fig. 5a) and rGO (Fig. 5b) at the edges, which indicates that graphene sheets have randomly aggregated. During the hydrothermal treatment, GO was reduced to rGO while Sn4þ ions came into being Sn(OH)4, followed by dehydration and crystal formation of SnO2 nanoparticles. Additionally, adding hydrochloric acid into the dispersion was a useful way to assist the attachment of Sn4þ ions to the surface of rGO through inhibiting the hydrolysis process. Thereby, numerous SnO2 nanoparticles with an around size of 10 nm are uniformly attached on the surface of rGO as shown in Fig. 5c and d, and there is no individual SnO2 nanoparticles aggregated. The energy dispersive spectrometer(EDS) of products is shown in Fig. 6, in which only C, O, Sn elements can be observed. It further demonstrates that SnO2 nanoparticles are successfully doping on rGO sheets, and the products are quite pure. According to the results of EDS, the weight fraction of SnO2 in composites is about 83.6%, which is closed to the expected 84.89 wt%. In conclusion, hydrothermal route is an effective way for preparing SnO2-rGO nanocomposite. 3.2. H2S and SOF2 sensing properties From the studies of W. Nakla et al. [43], F. Shao et al. [44] and our

Fig. 4. Raman spectra of GO, rGO, SnO2-rGO samples. (A colour version of this figure can be viewed online.)

98


Fig. 5. SEM images of (a) GO, (b) rGO, (c) SnO2-rGO, (d) high magnification of (c). (A colour version of this figure can be viewed online.)

Fig. 6. EDS spectrum of SnO2-rGO nanocomposite. (A colour version of this figure can be viewed online.)

work (as shown in Fig. S2), SnO2 exhibits poor recovery characteristics at a relatively low temperature, so we have only discussed the characteristics of sensors based on pristine rGO and SnO2-rGO in this paper. Fig. 7 shows the responses of sensor-1 toward 100 ppm H2S and 10 ppm SOF2 at different working temperatures. At 50 C, the responses of sensor toward H2S and SOF2 are only 20.87% and 1.14% respectively, and slowly return to the initial resistance of the sensor near 45 min and 7 min. The responses of the sensor toward target gases increase with the augment of operating temperatures, and attain the maximum values at about 125 C, followed by gradually decreasing. Furthermore, the recovery time decreases with the increase of the operating temperature. Considering high response, short recovery time and low energy assumption, we finally choose 125 C as the optimal working temperature of SnO2-rGO sensor. Fig. 8 illustrates the response-recovery curves of sensor-0 and sensor-1 toward 100 ppm H2S and 10 ppm SOF2 at 125 C. As shown in Fig. 8a, the resistance of SnO2-rGO sensor decreases greatly with the addition of 100 ppm H2S, and the response, response time and

recovery time are 34.31%, 209s, 900s, respectively. Contrarily, the pristine rGO sensor has a very low response to 100 ppm H2S, and the resistance of the sensor can't return to the initial values. It is interesting that SOF2 leads to an increase in the resistance of pristine rGO and SnO2-rGO sensors (Fig. 8b), as is opposite to the H2S-sensing behaviors and can be used to distinguish the two gases. Toward 10 ppm SOF2, the SnO2-rGO sensor has the response, response time and recovery time of 3.24%, 255s, 330s, respectively, while the pristine rGO sensor causes a lower response and longer response time and recovery time. The sensing performances shown in Fig. 8 reveal that SnO2-rGO sensor is more sensitive to H2S and SOF2 than pristine rGO sensor. For practicality, the uniformity of the sensor is an important property. In Fig. 9a and b, three sensors made in a same batch show the continuous response curves upon H2S and SOF2 exposurerelease cycles of varying concentrations at 125 C. As we can see, sensor-1, sensor-2 and sensor-3 have similar dynamic properties, and have ability to restore the initial resistance, exhibiting excellent uniformity. Fig. 9c and d show the relationship between the responses of SnO2-rGO sensors and gas concentrations. It's obvious that the responses increase with the increasing concentrations of target gases. The Freundlich equation based on adsorption on a heterogeneous surface was given by Refs. [45,46]:

jDRj=R ¼ KF C 1=n

(2)

lnðjDRj=RÞ ¼ ln KF þ ð1=nÞln C

(3)

where KF is a constant for the response capacity, and 1/n is an empirical parameter which is related to the response and affected by the heterogeneity of material [47]. The dependences of logarithm responses (Y) on logarithm concentrations (X) of H2S (Fig. 9e) and SOF2 (Fig. 9f) are approximately linear for SnO2-rGO sensors, and the determination coefficients (R2) are 0.9714 and 0.9776, respectively, suggesting sensors following Freundlich isotherm. For evaluating the detection limits (DL) of SnO2-rGO sensor to H2S and SOF2, the signal noise ratio (SNR) formula was given as


99

Fig. 7. The effect of operating temperature on performance of SnO2-rGO sensor upon exposure to (a)100 ppm H2S, (b)10 ppm SOF2. (A colour version of this figure can be viewed online.)

Fig. 8. The response-recovery curves of the pristine rGO and SnO2-rGO sensors toward (a)100 ppm H2S, (b)10 ppm SOF2 at the temperature of 125 C. (A colour version of this figure can be viewed online.)

follows [31]:

Smin =Nnoise ¼ 3

(4)

Where Nnoise is defined as the ratio between standard deviation and mean that is calculated to be equal to 0.09421%, based on 729 data points obtained from the baseline in Fig. S3, and small fluctuation in the baseline indicates that the sensor performs well in stability. Smin is defined as the minimum response of sensor which has potential to reach to 0.2826% according to SNR formula. Therefore, the detection limits in theory can be obtained according to the relationship between response and concentration in Fig. 9, and estimated to 42 ppb and 510 ppb toward H2S and SOF2, respectively. It was considered that the calibration gases that we used adopted N2 and He as balancing gases, which might have influence on the response of sensor. So, in Fig. S4, we further study the responses through respectively exposing SnO2-rGO sensor to the ambient of pure N2 and He, and then, the sensor produces the responses of 1.474% and 0.5623%. Because the concentrations of target gases were adjusted by controlling gas flow rates, we used pure N2 and He environment to approach the condition that the concentrations of target gases were up to the highest, and the responses of the sensor were also the maximum at the same time (44.96% and 4.735%, respectively). Therefore, even if the balancing gases had influences on the responses, it wouldn't greatly impact the use of SnO2-rGO sensor. Fig. 10a and Fig. 10b present four dynamic reversible cycles of sensor-1 toward H2S and SOF2, respectively. It can be easily found that the response-recovery curves were reproduced well. The

coefficients of variation (CV) for four responses to 100 ppm H2S and 10 ppm SOF2 are 1.212% and 3.68%, respectively, demonstrating that SnO2-rGO sensors have a good repeatability toward target gases during cycle test. Fig. 10c shows the sensors continuously sense to target gases for 10 days, during which the responses float slightly, exhibiting quite good long-term stability. We further evaluated the selectivity of sensor-1.20 ppm of typical SF6 decomposition components including H2S, SOF2, SO2, and SO2F2 were used to measure the responses of SnO2-rGO sensor at 125 C. As shown in Fig. 10d, the sensor has much higher responses to H2S and SOF2 than those of other two gases. Moreover, the SnO2-rGO sensor can distinguish the species of H2S and SOF2 in terms of resistance changes, performing good selectivity. Under the action of partial discharges, the SF6 decomposition components were produced in GIS, where SF6 has been utilized as the insulating gas. Whether SF6 as background gas would affect the performance of the sensor was very critical for practical applications. According to our previous research studies [48,49], the adsorption energy and the amount of charge transfer for SF6 were generally far smaller than those of SF6 decomposition components, and charge transfer causing a change in the resistance of materials played an important role in determining the adsorption energy. Therefore, the adsorption of SF6 induced the least charge transfer and adsorption energy among the SF6 decomposition components, which indicated that SF6 was hard to cause a change in the resistance of gas-sensing materials. Additionally, we have also investigated the response of SnO2-rGO sensor to 10000 ppm SF6. The result in Fig. S5 presents that the sensor hardly responses to SF6, according with our previous simulation studies. Moreover, owing to only trace amounts of water existed in GIS, it was reasonable to

100


Fig. 9. The response-recovery curves of three SnO2-rGO sensors to various concentrations of (a) H2S, (b) SOF2 at 125 C. Experimental responses and fitting models toward (c) H2S, and (d) SOF2 based on SnO2-rGO sensors at 125 C. The logarithm of the responses toward (e) H2S, and (f) SOF2 versus the logarithm of gas concentrations. (A colour version of this figure can be viewed online.)

keep a low relatively humidity during the sensing experiment. And the test for investigating the response of SnO2-rGO sensor to high humidity has also been presented in Fig. S6, where only a small response is produced. Therefore, SnO2-rGO sensor had potential to work at an ambient of high humidity, and it could be ascribed to the relatively high temperature of 125 C which weakened the influence brought by water [50]. Table 1 summarizes the H2S and SOF2 sensing performances of varying sensors in previous works. The In2O3 [51] sensor and the PPy-WO3 sensor [52] could work at room temperature, but the recovery time of these sensors were 7200s and 12600s, respectively. Seon-Jin Choi et al. [53], Shudi Peng et al. [54] and Hongcheng Liu et al. [14] respectively demonstrated that SnO2 nanofibers-rGO, ZnO nanorods and NiO-ZnO based sensors had fast recovery time, but their working temperatures were above 200 C which was higher than our sensors (125 C). In comparison with Xiaoxing Zhang et al. [12], our sensor has ability to return to the

initial resistance in a relatively short period. Besides, it can be observed that SnO2-rGO sensor prepared in our work performs excellent uniformity, which was seldom reported for other H2S or SOF2 sensors. What's more, the selectivity of our SnO2-rGO sensor makes it possible to distinguish H2S and SOF2 upon exposure to typical SF6 decomposition components in GIS. 3.3. Gas-sensing mechanism The probable sensing mechanism for the high-performance of SnO2-rGO nanocomposite is proposed as follows. Firstly, the attachment of SnO2 nanoparticles onto rGO sheets brings abundant adsorption sites (including vacancies, defects, functional groups and sp2-bonded carbon) [55e57] and significantly enhances the sensing properties of SnO2-rGO nanocomposite. Next, the assist of SnO2 nanoparticles prevents the aggregation of rGO sheets forming a high surface area, which facilitates the diffusion and adsorption of


101

Fig. 10. The repeatability curve of SnO2-rGO sensor exposed to (a) 100 ppm H2S, (b) 10 ppm SOF2 at 125 C. (c) The long-term stability of SnO2-rGO sensor upon exposure to 100 ppm H2S and 10 ppm SOF2 at 125 C. (d) The selectivity of SnO2-rGO sensors toward 20 ppm gases, including H2S, SOF2, SO2, SO2F2 at 125 C. (A colour version of this figure can be viewed online.)

Table 1 Comparison of H2S and SOF2 sensing performances of varying gas sensors (“d” represents the data was not given in the literature; “√” represents the data was given in the literature). Material

Gas species Concentration (ppm) Temperature ( C) Response (DR/R, Ra/Rg) Response/Recovery time (s) Selectivity Uniformity Reference

In2O3 PPy-WO3 SnO2 nanofibers-rGO ZnO nanorods NiO-ZnO Au-rGO

H2S H2S H2S SOF2

10 1 5 10

RT RT 200 300

35% 81% 34 7%

240/7200 360/12600 ~120/~550 10/17

✓ ✓ ✓ e

e e e e

[51] [52] [53] [54]

SOF2 H2S SOF2 H2S SOF2

100 100 100 100 10

260 RT RT 125 125

22.25 28.15% 23.83% 33.02% 3.24%

18/22 e

e ✓

e e

[14] [12]

209/900 255/330

✓

✓

This work

SnO2-rGO

gas molecules [37]. In addition, favorable electrical transport properties of rGO enhances electrons transfer, and that Sn doping leads to conductivity reduction in comparison with pristine rGO indicates electrons transfer from SnO2 to rGO. And then, a barrier has formed at the surface between p-type rGO nanosheet and ntype SnO2 nanoparticles, causing the electron depletion region [53]. In ambient air, chemisorbed ionized oxygen species (O-2) capturing electrons from sensing material further alter the electron depletion region and the sensor resistance, which can be expressed as follows [35]:

O2ðgasÞ /O2ðadsÞ

(5)

O2ðadsÞ þe /O 2ðadsÞ

(6)

When the sensor is exposed to target gas (e.g. H2S, SOF2), the gas molecules can easily adsorb on active sites of SnO2-rGO nanocomposite and react with preadsorbed oxygen adions. As a result, electrons trapped in chemisorbed ionized oxygen species transfer

between gas molecules and sensing material, modifying electron depletion region at the interface of SnO2-rGO nanocomposite. The surface reactions of H2S and SOF2 with chemisorbed ionized oxygen can be respectively proposed as follows: 2 H2 S þ 3 O 2ðadsÞ /2 H2 O þ 2 SO2 þ 3 e

(7)

SOF2 þ e /2 SOF 2

(8)

SOF2 þ O 2ðadsÞ /2 SOF2 þ O2

(9)

In Fig. 11, a possible working principle is used to explain the selectivity of SnO2-rGO sensor for distinguishing H2S and SOF2. As reported in the literature [58,59], H2S is a typical reducing gas. It should be noted that the decreased resistance is observed in SnO2rGO sensor upon exposure to H2S, which can be ascribed to the increase of electronic carrier concentration. The reason of this phenomenon is that the electrons trapped by chemisorbed ionized oxygen feed back to the conduction band of gas-sensing material,

102


References

Fig. 11. Schematic of a proposed sensing mechanism: the adsorption of H2S and SOF2 molecules onto the SnO2-rGO nanocomposite. (A colour version of this figure can be viewed online.)

thereby leading to the narrowing width of electron depletion region. On the contrary, SOF2 attracts electrons from SnO2-rGO, and decreases the electronic carrier concentration, leading to the increasing resistance of the gas sensor. In our work, the incorporation of SnO2 and rGO facilitates the detection of H2S and SOF2 in comparison to their constituents alone. In addition, as a p-type semiconductor, the resistance of pristine rGO should present an increase and a decrease upon exposure to H2S and SOF2, respectively. However, it reflects a property of n-type semiconductor in Fig. 8. Some studies suggested that the pristine rGO might have an n-p transition under thermal reduction [60], and trace impurities mixed in pristine rGO was another possible reason [12]. Overall, this phenomenon needs to be further investigated. 4. Conclusion This paper introduced a high-performance sensor employing SnO2-rGO nanocomposite as sensing material for the detection of H2S and SOF2. The sensor was prepared by a facile one-step hydrothermal synthesis route, and a series of approaches like XRD, RS, SEM, EDS were used to characterized its morphology and nanostructure. The gas-sensing experiments indicated that doping of rGO with SnO2 nanoparticles significantly enhanced gas sensing properties at the operating temperature of 125 C. Moreover, the SnO2-rGO sensor with excellent uniformity had ability to select four decomposition products of SF6, and revealed the potential to distinguish H2S and SOF2 by the direction of the resistance change. In addition, the detect limits of SnO2-rGO sensor in theory could reach to 42 ppb and 510 ppb toward H2S and SOF2, respectively. Thus, the high-performance gas sensors described here had potential to be deployed for online monitoring of GIS. Acknowledgements This work was supported by State Grid Corporation of China (No. SGSXC200XTJS1700084), National Science Foundation of China (No. 51521065), National Key Basic Research Program (“973” Program) of China (No. 2015CB251001), Young Elite Scientists Sponsorship Program by CAST and the Fundamental Research Funds for the Central Universities. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.carbon.2018.04.037.

[1] G. Ma, Z. Wu, H. Zhou, J. Jiang, W. Chen, S. Zheng, et al., A wireless and passive online temperature monitoring system for GIS based on Surface-AcousticWave sensor, IEEE Trans. Power Deliv. 31 (3) (2016) 1270e1280. [2] R.J. Vanbrunt, J.T. Herron, Fundamental processes of SF6 decomposition and oxidation in glow and corona discharges, IEEE Electr. Insul. Mag. 25 (1) (1990) 75e94. [3] C. Beyer, H. Jenett, D. Klockow, Influence of reactive SFx gases on electrode surfaces after electrical discharges under SF6 atmosphere, IEEE T Dielect El In. 7 (2) (2000) 234e240. [4] J. Tang, F.P. Zeng, J.Y. Pan, X.X. Zhang, Q. Yao, J.J. He, et al., Correlation analysis between formation process of SF6 decomposed components and partial discharge qualities, IEEE T Dielect El 20 (3) (2013) 166e177. [5] J. Tang, F. Liu, X. Zhang, Q. Meng, J. Zhou, Partial discharge recognition through an analysis of SF6 decomposition products part 1: decomposition characteristics of SF6 under four different partial discharges, IEEE Trans. Dielectr. Electr. Insul. 19 (1) (2012) 29e36. [6] Q. Li, H. Cong, J. Xing, B. Qi, C. Li, On-line temperature monitoring of the GIS contacts based on infrared sensing technology, J Electr Eng Technol 9 (4) (2014) 1385e1393. [7] R. Kurte, H.M. Heise, D. Klockow, Quantitative infrared spectroscopic analysis of SF6 decomposition products obtained by electrical partial discharges and sparks using PLS-calibrations. Special issue, J. Mol. Struct. 565 (2001) 505e513. [8] J. Luo, Y.H. Fang, Z.X. Su, D.C. Li, Y.D. Zhao, A.J. Wang, et al., The research of temperature properties of photoacoustic spectroscopy detection for SF6 decomposition products in gas insulated switchgear, Anal Methods-Uk 7 (9) (2015) 3806e3813. [9] O. Koreh, T. Rikker, G. Molnar, B.M. Mahara, K. Torkos, J. Borossay, Study of decomposition of sulphur hexafluoride by gas chromatography mass spectrometry, Rapid Commun. Mass Spectrom. 11 (15) (1997) 1643e1648. [10] X. Zhang, J. Zhang, Y. Jia, P. Xiao, J. Tang, TiO2 nanotube array sensor for detecting the SF6 decomposition product SO2, Sensors-Basel 12 (12) (2012) 3302e3313. [11] X. Zhang, L. Yu, J. Tie, X. Dong, Gas sensitivity and sensing mechanism studies on Au-Doped TiO2 nanotube arrays for detecting SF6 decomposed components, Sensors-Basel 14 (10) (2014) 19517e19532. [12] X. Zhang, L. Yu, X. Wu, W. Hu, Experimental sensing and density functional theory study of H2S and SOF2 adsorption on Au-Modified graphene, Adv. Sci. 2 (2015) 150010111. [13] X. Zhang, F. Meng, B. Yang, Use of hydroxyl-modified carbon nanotubes for detecting SF6 decomposition products under partial discharge in gas insulated switchgear, IEEE T Dielect El In 20 (6) (2013) 2246e2253. [14] H. Liu, Q. Zhou, Q. Zhang, C. Hong, L. Xu, L. Jin, et al., Synthesis, characterization and enhanced sensing properties of a NiO/ZnO p-n junctions sensor for the SF6 decomposition byproducts SO2, SO2F2, and SOF2, Sensors-Basel 17 (2017) 9134. [15] Y. Zhu, S. Murali, W. Cai, X. Li, J.W. Suk, J.R. Potts, et al., Graphene and graphene oxide: synthesis, properties, and applications, Adv. Mater. 22 (46) (2010) 5226. [16] C. Lee, X. Wei, J.W. Kysar, J. Hone, Measurement of the elastic properties and intrinsic strength of monolayer graphene, Science 321 (5887) (2008) 385e388. [17] Y. Su, G. Xie, S. Wang, H. Tai, Q. Zhang, H. Du, et al., Novel high-performance self-powered humidity detection enabled by triboelectric effect, Sensor. Actuator. B Chem. 251 (2017) 144e152. [18] F. Schedin, A.K. Geim, S.V. Morozov, E.W. Hill, P. Blake, M.I. Katsnelson, et al., Detection of individual gas molecules adsorbed on graphene, Nat. Mater. 6 (9) (2007) 652e655. [19] S. Prezioso, F. Perrozzi, L. Giancaterini, C. Cantalini, E. Treossi, V. Palermo, et al., Graphene oxide as a practical solution to high sensitivity gas sensing, J. Phys. Chem. C 117 (20) (2013) 10683e10690. [20] G. Lu, L.E. Ocola, J. Chen, Reduced graphene oxide for room-temperature gas sensors, Nanotechnology 20 (2009) 44550244. [21] Y. Su, G. Xie, J. Chen, H. Du, H. Zhang, Z. Yuan, et al., Reduced graphene oxidepolyethylene oxide hybrid films for toluene sensing at room temperature, RSC Adv. 6 (100) (2016) 97840e97847. [22] D. Yang, I. Kamienchick, D.Y. Youn, A. Rothschild, I. Kim, Ultrasensitive and highly selective gas sensors based on electrospun SnO2 nanofibers modified by Pd loading, Adv. Funct. Mater. 20 (24) (2010) 4258e4264. [23] J. Xu, Z. Xue, N. Qin, Z. Cheng, Q. Xiang, The crystal facet-dependent gas sensing properties of ZnO nanosheets: experimental and computational study, Sensor. Actuator. B Chem. 242 (2017) 148e157. [24] F. Meng, X. Di, H. Dong, Y. Zhang, C. Zhu, C. Li, et al., PPb H2S gas sensing characteristics of Cu2O/CuO sub-microspheres at low-temperature, Sensor. Actuator. B Chem. 182 (2013) 197e204. [25] N. Singh, R.K. Gupta, P.S. Lee, Gold-Nanoparticle-Functionalized In2O3 nanowires as CO gas sensors with a significant enhancement in response, Acs Appl Mater Inter 3 (7) (2011) 2246e2252. [26] M. Zhao, J.X. Huang, C.W. Ong, Diffusion-controlled H2 sensors composed of Pd-coated highly porous WO3 nanocluster films, Sensor. Actuator. B Chem. 191 (2014) 711e718. [27] S.G. Pawar, S.L. Patil, M.A. Chougule, B.T. Raut, P.R. Godase, R.N. Mulik, et al.,


[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

New method for fabrication of CSA doped PANI-TiO2 Thin-Film ammonia sensor, IEEE Sensor. J. 11 (11) (2011) 2980e2985. L. Mei, Y. Chen, J. Ma, Gas sensing of SnO2 nanocrystals revisited: developing Ultra-Sensitive sensors for detecting the H2S leakage of biogas, Sci Rep-Uk 4 (2014) 6028. B. Wang, L.F. Zhu, Y.H. Yang, N.S. Xu, G.W. Yang, Fabrication of a SnO2 nanowire gas sensor and sensor performance for hydrogen, J. Phys. Chem. C 112 (17) (2008) 6643e6647. S. Cui, Z. Wen, E.C. Mattson, S. Mao, J. Chang, M. Weinert, et al., Indium-doped SnO2 nanoparticle-graphene nanohybrids: simple one-pot synthesis and their selective detection of NO2, J. Mater. Chem. 1 (14) (2013) 4462e4467. Z. Song, Z. Wei, B. Wang, Z. Luo, S. Xu, W. Zhang, et al., Sensitive roomtemperature H2S gas sensors employing SnO2 quantum Wire/Reduced graphene oxide nanocomposites, Chem. Mater. 28 (4) (2016) 1205e1212. C. Ning, L. Xiaogan, W. Xueyan, Y. Jun, W. Jing, T. Zhenan, et al., Enhanced room temperature sensing of Co3O4-intercalated reduced graphene oxide based gas sensors, Sensor. Actuator. B Chem. 188 (2013) 902e908. S. Liu, B. Yu, H. Zhang, T. Fei, T. Zhang, Enhancing NO2 gas sensing performances at room temperature based on reduced graphene oxide-ZnO nanoparticles hybrids, Sensor. Actuator. B Chem. 202 (2014) 272e278. Y. Su, G. Xie, H. Tai, S. Li, B. Yang, S. Wang, et al., Self-powered room temperature NO2 detection driven by triboelectric nanogenerator under UV illumination, Nanomater. Energy 47 (2018) 316e324. H. Zhang, J. Feng, T. Fei, S. Liu, T. Zhang, SnO2 nanoparticles-reduced graphene oxide nanocomposites for NO2 sensing at low operating temperature, Sensor. Actuator. B Chem. 190 (2014) 472e478. D. Zhang, H. Chang, P. Li, R. Liu, Q. Xue, Fabrication and characterization of an ultrasensitive humidity sensor based on metal oxide/graphene hybrid nanocomposite, Sensor. Actuator. B Chem. 225 (2016) 233e240. D. Zhang, J. Liu, H. Chang, A. Liu, B. Xia, Characterization of a hybrid composite of SnO2 nanocrystal-decorated reduced graphene oxide for ppm-level ethanol gas sensing application, RSC Adv. 5 (24) (2015) 18666e18672. L. Wang, Z. Lou, T. Zhang, H. Fan, X. Xu, Facile synthesis of hierarchical SnO2 semiconductor microspheres for gas sensor application, Sensor. Actuator. B Chem. 155 (1) (2011) 285e289. C. Su, Y. Xu, W. Zhang, J. Zhao, X. Tang, C. Tsai, et al., Electrical and spectroscopic characterizations of Ultra-Large reduced graphene oxide monolayers, Chem. Mater. 21 (23) (2009) 5674e5680. S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, et al., Synthesis of graphene-based nanosheets via chemical reduction of exfoliated graphite oxide, Carbon 45 (7) (2007) 1558e1565. S. Yan, X. Liang, H. Song, S. Ma, Y. Lu, Synthesis of porous CeO2-SnO2 nanosheets gas sensors with enhanced sensitivity, Ceram. Int. 44 (1) (2018) 358e363. N. Cao, Q. Lyu, J. Li, Y. Wang, B. Yang, S. Szunerits, et al., Facile synthesis of fluorinated polydopamine/chitosan/reduced graphene oxide composite aerogel for efficient oil/water separation, Chem. Eng. J. 326 (2017) 17e28. W. Nakla, A. Wisitsora-at, A. Tuantranont, P. Singjai, S. Phanichphant, C. Liewhiran, H2S sensor based on SnO2 nanostructured film prepared by high current heating, Sensor. Actuator. B Chem. 203 (2014) 565e578. F. Shao, M.W.G. Hoffmann, J.D. Prades, R. Zamani, J. Arbiol, J.R. Morante, et al., Heterostructured p-CuO (nanoparticle)/n-SnO2 (nanowire) devices for

103

selective H2S detection, Sensor. Actuator. B Chem. 181 (2013) 130e135. [45] H. Liu, S. Xu, M. Li, G. Shao, H. Song, W. Zhang, et al., Chemiresistive gas sensors employing solution-processed metal oxide quantum dot films, Appl. Phys. Lett. 105 (16) (2014) 163104. [46] J. Pan, Y. Ma, J. Zeng, X. Niu, T. Zhang, F. Qiu, et al., Facile assembly of hollow polydopamine capsules onto macroporous poly(glycidyl methacrylate) foams for simultaneous removal of l-cyhalothrin and copper ions, Chem. Eng. J. 302 (2016) 670e681. [47] Z. Song, J. Liu, Q. Liu, H. Yu, W. Zhang, Y. Wang, et al., Enhanced H2S gas sensing properties based on SnO2 quantum wire/reduced graphene oxide nanocomposites: equilibrium and kinetics modeling, Sensor. Actuator. B Chem. 249 (2017) 632e638. [48] A. Yang, D. Wang, X. Wang, J. Chu, P. Lv, Y. Liu, et al., Phosphorene: a promising candidate for highly sensitive and selective SF6 decomposition gas sensors, IEEE Electron. Device Lett. 38 (7) (2017) 963e966. [49] D. Wang, X. Wang, A. Yang, J. Chu, P. Lv, Y. Liu, et al., MoTe2: a promising candidate for SF6 decomposition gas sensors with high sensitivity and selectivity, IEEE Electron. Device Lett. 39 (2) (2018) 292e295. [50] M. Righettoni, A. Tricoli, S.E. Pratsinis, Si:WO3 sensors for highly selective detection of acetone for easy diagnosis of diabetes by breath analysis, Anal. Chem. 82 (9) (2010) 3581e3587. [51] M. Kaur, N. Jain, K. Sharma, S. Bhattacharya, M. Roy, A.K. Tyagi, et al., Roomtemperature H2S gas sensing at ppb level by single crystal In2O3 whiskers, Sensor. Actuator. B Chem. 133 (2) (2008) 456e461. [52] P. Su, Y. Peng, Fabrication of a room-temperature H2S gas sensor based on ppy/WO3 nanocomposite films by in-situ photopolymerization, Sensor. Actuator. B Chem. 193 (2014) 637e643. [53] S. Choi, B. Jang, S. Lee, B.K. Min, A. Rothschild, I. Kim, Selective detection of acetone and hydrogen sulfide for the diagnosis of diabetes and halitosis using SnO2 nanofibers functionalized with reduced graphene oxide nanosheets, Acs Appl Mater Inter 6 (4) (2014) 2588e2597. [54] S. Peng, G. Wu, W. Song, Q. Wang, Application of Flower-Like ZnO nanorods gas sensor detecting SF6 decomposition products, J. Nanomater. (2013) 135147. [55] Q. Lin, Y. Li, M. Yang, Tin oxide/graphene composite fabricated via a hydrothermal method for gas sensors working at room temperature, Sensor. Actuator. B Chem. 173 (2012) 139e147. [56] J.D. Fowler, M.J. Allen, V.C. Tung, Y. Yang, R.B. Kaner, B.H. Weiller, Practical chemical sensors from chemically derived graphene, ACS Nano 3 (2) (2009) 301e306. [57] D. Zhang, A. Liu, H. Chang, B. Xia, Room-temperature high-performance acetone gas sensor based on hydrothermal synthesized SnO2-reduced graphene oxide hybrid composite, RSC Adv. 5 (4) (2015) 3016e3022. [58] X. Zhang, Y. Gui, X. Dong, Preparation and application of TiO2 nanotube array gas sensor for SF6-Insulated equipment detection: a review, Nanoscale Res Lett. 11 (2016) 302. [59] X. Zhang, H. Cui, Y. Gui, J. Tang, Mechanism and application of carbon nanotube sensors in SF6 decomposed production detection: a review, Nanoscale Res Lett. 12 (2017) 177. [60] Y. Zhou, X. Lin, Y. Huang, Y. Guo, C. Gao, G. Xie, et al., Impact of further thermal reduction on few-layer reduced graphene oxide film and its n-p transition for gas sensing, Sensor. Actuator. B Chem. 235 (2016) 241e250.