- Email: [email protected]
Realization of Au-decorated WS2 nanosheets as low power-consumption and selective gas sensors
Sensors & Actuators: B. Chemical 296 (2019) 126659
Contents lists available at ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Realization of Au-decorated WS2 nanosheets as low power-consumption and selective gas sensors ⁎
Jae-Hun Kima, Ali Mirzaeib,c, Hyoun Woo Kimb,d, , Sang Sub Kima,
T
⁎⁎
a
Department of Materials Science and Engineering, Inha University, Incheon, 22212, Republic of Korea The Research Institute of Industrial Science, Hanyang University, Seoul, 04763, Republic of Korea c Department of Materials Science and Engineering, Shiraz University of Technology, Shiraz, Iran d Division of Materials Science and Engineering, Hanyang University, Seoul, 04763, Republic of Korea b
A R T I C LE I N FO
A B S T R A C T
Keywords: 2D WS2 Au-decoration Self-heating Gas sensor Sensing mechanism
We introduced a selective and self-heated CO gas sensor based on Au-decorated WS2 nanosheets. Au nanoparticles (NPs) were decorated on WS2 NSs using UV irradiation technique at different irradiation times (1, 15 and 30 s). SiO2 grown on Si was used as substrates and the gas sensors with top electrode configuration were fabricated. Gas sensors were able to work under self-heating mode with an applied voltage of 2 V for selective CO sensing. The results of CO gas sensing demonstrated the promising effects of Au for the enhanced response and selectivity towards CO gas. Also, the low-voltage operation at room temperature resulted in low-power consumption. This study demonstrates realization of selective CO gas sensor using a facile approach along with low power consumption.
1. Introduction Although graphene, the best-known two-dimensional (2D) material, has very exciting properties from both theoretical and practical standpoints, accessing these properties in practical applications can be a challenge. It is chemically inert and can only be activated through functionalization with other materials, a process which will sometimes cause degradation of some of its outstanding properties. In contrast, 2D layered transition metal dichalcogenides (TMDs), with the general formula MX2 where M is a transition metal (usually, Mo, W, Nb, Ti, Ta) and X is a chalcogen (S, Se, Te), possess very versatile chemistries, which provides opportunities for researchers in different fields to develop TMD-based devices with ease [1]. TMDCs can be prepared in 2D sheets with high surface-area, are semiconducting with tunable band gaps and good mobility of charge carriers (∼500 cm2/(V.s)), are made of earth-abundant materials, and exhibit low toxicity [2–5]. Among the different TMDs, tungsten disulfide (WS2) is of particular interest as it possesses desirable properties such as excellent thermal stability, oxidative resistance, favorable band structure and high phonon-limited electron mobility [6]. Semiconducting WS2 in bulk form has an indirect band gap of 1.4 eV, while a monolayer of WS2 has a wide, direct band gap of 2.1 eV [5]. The structure of WS2 is very useful for gas sensing applications. It is composed of vertically-stacked layers
⁎
of covalently bonded WeS atoms. Each layer is connected to its neighboring layers by relatively weak van der Waals forces, which allow target gases to freely diffuse between the layers [7]. Due to these advantageous structural properties, WS2 is a promising candidate for the realization of high-performance gas sensors at and near room temperature [8–11]. In addition, WS2 has also been used as humidity sensor [12,13], strain sensor [14], DNA hybridization sensor [15] and evanescent wave absorption sensor [16]. In the context of gas sensing applications, pristine WS2 gas sensors show poor response [17]. Moreover, because of strong gas adsorption, the recovery of pristine WS2 gas sensors is generally long and sometimes incomplete at room temperature, limiting applications in real environments [18]. A promising strategy to enhance gas response of WS2 nanosheets (NSs) is fabrication of composites [19]. For example, WS2/WO3 composites were reported for improved sensing performance [20]. Another promising strategy to enhance overal sensing properties of resistance-based gas sensors is decoration of noble metals on the surfaces of the sensing layer. This is due to the catalytic effect of metal and lower energy required for adsorption and interaction of gas molecules with the sensing layer [21,22]. For example, polyakov et al. reported enhanced sensing reponse of WS2 NSs decorated with Au nanoparticles [23]. In this study, we fabricated pristine and Au-decorated WS2 gas
Corresponding author at: The Research Institute of Industrial Science, Hanyang University, Seoul, 04763, Republic of Korea. Corresponding author at: Department of Materials Science and Engineering, Inha University, Incheon 22212, Republic of Korea. E-mail addresses: [email protected] (H.W. Kim), [email protected] (S.S. Kim).
⁎⁎
https://doi.org/10.1016/j.snb.2019.126659 Received 27 February 2019; Received in revised form 4 June 2019; Accepted 4 June 2019 Available online 08 June 2019 0925-4005/ © 2019 Elsevier B.V. All rights reserved.
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
(UHV) using irradiation of HeI 21.2 eV ultraviolet light. The temperature of the gas sensors in self-heating mode was measured by an infrared imaging camera (FLIR E75). The thermal resolution of the IR camera is less than 20 mK at 30 °C. Therefore, the camera can detect even the 0.1 °C-difference in temperature.
sensors for CO gas sensing. We investigated the effects of irradiation time for Au deposition on WS2 and studied the effects of applied voltage on the sensing performance of gas sensor. It was shown that selective CO gas sensors with very low power consumption (28.6 μW) can be fabricated using Au-decorated WS2 NSs. The sensing mechanism is discussed in detail. This study has opened new doors for the fabrication of low-power and selective CO gas sensors using a combination of noble metals and TMDs.
2.3. Gas sensing tests The gas sensing tests were similar to those elsewhere [24]. The fabricated gas sensors were put into the gas chamber with possibility of temperature control. By varying the mixing ratio of dry air to the target gas using mass flow controllers with a total stream of 500 sccm, gas flow was controlled during the gas sensing tests. The sensors’ resistances were recorded using a source meter as a function of time in different atmospheres. Both pristine and Au-decorated WS2 gas sensors were studied under the same sensing setup and gas flow conditions. The sensor response (R) was defined as R = Rg/Ra (for reducing gases) and R = Ra/Rg, (for NO2 as an oxidizing gas) where Ra and Rg are the resistances of the gas sensor in air and in the presence of the target gas, respectively. For self-heating experiments, different voltages were applied to the gas sensor via the source meter during measurements at room temperature. The response time was calculated as the time it took the resistance of the sensor to change to 90% of the original baseline, and the recovery time was calculated as the time needed for 90% of the signal to be recovered [25].
2. Experimental 2.1. Fabrication of gas sensors For fabrication of gas sensor, SiO2-coated Si substrate was equipped with interdigitated bi-layered electrodes comprising of Ti/Pt (50 nm/ 200 nm thick) layers that were sequentially deposited on the substrates by a sputtering process. First, a 5 mg monolayer WS2 powder (ACS Materials) with a thickness of ˜ 1 nm, was dispersed in 0.01 ml of 2propanol for 15 min. Then, 0.075 μl of the solution was drop coated (in three drops) onto the prepared substrates. Afterwards, the substrates with deposited WS2 were dried at 60 °C for 10 min. For fabrication of Au-decorated WS2, the surface of the WS2 NSs were decorated with Au NPs using an ultraviolet (UV) irradiation. First, 0.4 g of HAuCl4·H2O was dissolved in 2-propanol for 15 min. Then, the synthesized WS2 NSs were dipped into the gold solution and subsequently they were UV irradiated using a halogen lamp (intensity =0.11 mW/cm2, λ = 360 nm) at different times (1–30 s). Finally, the Au-decorated WS2 NSs were annealed at 500 °C for 30 min in N2 atmosphere. The dried Au-decorated WS2 NSs powder was dispersed in 2propanol for 15 min. Subsequent sensor`s fabrication was similar to pristine WS2 gas sensor. Fig. 1a and b schematically shows the fabrication process of pristine and Au-decorated WS2 NSs gas sensors, respectively.
3. Results and discussion 3.1. Structural and morphological analyses Fig. 2a and its inset, shows FE-SEM images of pristine WS2 NSs deposited on the substrate equipped with interdigitated electrodes. It can be seen that the WS2 almost fully covered the surface of substrate with sufficient contacts among them. Fig. 2b provides the TEM micrograph of pristine WS2 that clearly show the ultrathin nature of WS2 NSs and Fig. 2c confirms the presence of Au along with WS2 as the latticeresolved TEM image shows lattice fringes with spacings of 0.236 nm and 0.270 nm that can be indexed as the (111) and (100) planes of Au [26] and WS2, [27], respectively. Further, TEM images of pristine and Au-decorated WS2 NSs at different irradiation times are shown in Fig. 3(a)–(d) for pristine, 1, 15, and 30 s irradiated samples, respectively. These images clearly demonstrate the presence of Au NPs on the surfaces of WS2 NSs. Particle density (amount/50 nm2 WS2 NS) and size of Au NPs were calculated from TEM images and are presented in Fig. 3(e). Particle sizes increase with increasing irradiation time,
2.2. Characterization Microstructural and morphological studies were performed using field-emission scanning electron microscopy (HR-SEM SU 8010) and transmission electron microscopy (FE-TEM JEM-2100 F). Raman spectra were obtained using a LabRAM HR spectrometer (HORIBA Jobin Yvon, France) with a laser excitation wavelength of 514.5 nm. XPS spectrum was recorded using a K-Alpha system of Thermo Scientific, equipped with a monochromatic Al-Kα source (1486.6 eV). Ultraviolet photoelectron spectroscopy (UPS) measurements (Thermo Fisher Scientific Co. Theta probe) were performed in ultrahigh vacuum
Fig. 1. Schematic representation of preparation steps of pristine and Au-decorated WS2 NSs. 2
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 2. (a) FE-SEM images of pristine WS2 NSs on the substrate. Inset in (a) shows a higher magnification image. (b) TEM image of pristine WS2 NSs (c) latticeresolved TEM image of Au-decorated WS2 NSs.
(530.7 eV) and the second peak (532.6 eV) have been slightly shifted, which may be due to electronic interaction between Au and WS2. A comparison between O 1s regions of pristine and Au-decorated WS2 NSs exhibits the higher amount of adsorbed oxygen species on the surface of Au-decorated WS2 NSs, presumably due to catalytic effect of Au on the surface of WS2. Accordingly, a higher gas response can be expected for this gas sensor. Furthermore, we obtained the XPS regions of W 4f, W 5p3/2 and S 2p at different UV irradiation times (0, 1, 15, 30 s) and the results are presented in Fig. S1. As shown in Fig. S1(a), the peaks related to the W 4f7/2, W 4f 5/2 and W 5p 3/2 can be seen. Also, as shown in Fig. S1(b), the peaks related to the S 2p3/2 and S 2p1/2 can be seen. Overall, it can be seen that the peaks shift to lower binding energies with increasing of UV irradiation time. Also, for the 30 s irradiated sample, the 5p3/2 peak has been disappeared in the W4f region. Therefore, it seems that with increasing UV irradiation time, more Au particles can be reduced which resulted in more intense electronic interaction between Au and WS2 and thus a shift in the XPS peaks.
whereas particle density shows a maximum for 15 s irradiation time. It seems that 1 s irradiation is not enough to produce enough Au NPs on the surface of WS2, whereas 30 s irradiation will result in producing large amount of Au NPs, being agglomerated and larger, which eventually leads to a decrease of particle density. The Raman spectra of pristine and Au-decorated WS2 NSs, respectively, are shown in Fig. 4(a) and (b). Two broad peaks around 421 cm−1 and 352 cm−1 belong to the out-of-plane A 1 g vibration mode of S atoms and in-plane vibration mode, E21g , respectively [21]. The ratio of the A1 g to E21g peaks is 0.693 for the pristine WS2 NSs, which will increase to 0.872 for Au-decoratedWS2 NSs. Fig. 5 provides the XPS survey of Au-functionalized WS2 NSs. The XPS peaks related to O 1s, W 4p, W 4d, W 4f, S 2p, C 1s, Si 2p, Au 4d and Au 4f can be seen, which demonstrates the existence of Au on WS2 NSs. Fig. 5c and d show the deconvoluted O1 s region for pristine and Au-decorated WS2 NSs, respectively. For pristine WS2 NSs, O 1s peak can be deconvoluted into two peaks located at 530.7 eV and 532.6 eV. The first peak can be assigned to adsorbed oxygen species such as CO2, O2 and adsorbed H2O [28], on the surface of WS2. Since adsorbed oxygen directly affect the gas sensing properties of the gas sensor, this peak can predict the gas sensing properties. The second peak can be attributed to WeO bonds, which is due to the surface oxidation of WS2 during the preparation process [29]. For Au-decorated WS2 NWs, locations of the first peak
3.2. Gas sensing studies Before self-heating tests, we tested the behavior of gas sensor under external heating at different temperatures. Fig. 6a shows dynamic 3
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 3. TEM images of Au-decorated WS2 NSs under different UV irradiation times, (a) 0 s, (b) 1 s, (c) 15 s and (d) 30 s. (e) Particle density and average of particle size of Au NPs as a function of UV irradiation time.
10 V, respectively. Accordingly, it can be deduced that self-heating effect works on the gas sensors and can be used for sensing studies. We investigated the effect of applied voltage on the response of WS2 NSs, with H2S gas. Fig. 8a, shows the dynamic resistance changes of pristine WS2 gas sensor to 1, 10 and 50 ppm H2S gas under 1–10 V applied voltages without external heating. The sensor shows a response to all concentrations of H2S gas under all applied voltages. The resistance decreases with increasing the applied voltage due to the fact that higher applied voltage will cause more electrons jump to conduction band of WS2 NS (Fig. 8b). As presented in Fig. 9a, the magnitude of the response greatly depends on the applied voltage and the highest response is observed under 2 V applied voltage and with further increase of the applied voltage, the response of gas sensor decreases. Since the WS2 sensor
resistance curves of pristine WS2 NSs gas sensor to 1, 10 and 50 ppm H2S gas at various temperatures. It can be seen that the gas sensor shows n-type behavior, where the resistance decreases with introduction of H2S gas. In addition, the sensor also exhibits good reversibility, where upon stoppage of H2S gas, the resistance returns to its initial value in air. As plotted in Fig. 6b, the response of sensor increases with increasing the sensing temperature, in which higher temperatures provide more energy to promote the adsorption and reaction processes on the surface of the gas sensor. The highest gas response at 120 °C was 1.35 for 50 ppm H2S gas. Fig. 7 shows the temperature of pristine WS2 gas sensor as a function of voltage along with corresponding thermographs. The temperature of gas sensor at 0 V was 23.5 °C, which can be increased to 31.6, 31.7, 31.8, 31.9 and 32.2 °C, under an applied voltages of 1, 2, 3, 5 and
4
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
catalytic performance of Au towards CO gas. Fig. 10a shows corresponding response versus applied voltage for 50 ppm CO gas. The sensor shows a good response to CO gas under different applied voltages, with the maximum response being located at 2 V. For comparison, the response of the pristine WS2 NSs gas sensor is also shown, which clearly demonstrate the promising effect of Au for CO sensing. Fig. 10b shows the initial resistance of Au-decorated WS2 (irradiated for 1 s) as a function of applied voltage. The decrease of resistance with increasing the applied voltage will be related to the increased conduction electrons. The initial resistance of the Au-decorated WS2 gas sensor is higher than that of the pristine WS2 NSs gas sensor owing to transfer of electrons from WS2 to Au NPs, which will be further discussed in sensing mechanism section. It is anticipated that increasing the applied voltage will slightly increase the sensing temperature, promoting the sensing process. However, the trend in Fig. 10a indicates that another mechanism operates. Fig. 10b indicates that higher applied voltage resulted in lower initial resistance. Low initial resistance corresponds to a high initial conduction volume/regions. In this case, the same amount of gases adsorbed/removed with the higher conduction volume will bring about a lower sensor response [30]. After finding of the optimal applied voltage (2 V), the Au-decorated WS2 NSs gas sensors irradiated at different times were exposed to 1, 10 and 50 ppm CO gas at 2 V to find optimal gas sensor in term of response to CO gas. Fig. S4 (Supplementary Information) shows corresponding dynamic resistance curves and Fig. 11a summarizes the response values versus the Au irradiation times and Au particle sizes. For all concentrations of CO gas, the Au-decorated WS2 NSs gas sensor irradiated for 15 s showed the maximum response to this gas. This can be related to the amount and size of Au NPs on the surface of WS2 which will be discussed in the sensing mechanism section. After finding of the optimal gas sensor, in order to evaluate selectivity of Au-decorated WS2 (irradiated for 15 s) gas sensor, it was exposed to 1, 10 and 50 ppm of interfering gases and dynamic resistance curves under 2 V applied voltage are shown in Fig. S5 (Supplementary Information). Fig. 11b compares the selectivity patterns of pristine and Au-decorated WS2 NSs gas sensor
Fig. 4. Raman spectra of (a) pristine WS2 and (b) Au-decorated WS2 NSs.
showed its best performance under applied voltage of 2 V, selectivity tests in the presence of interfering gases were performed under this condition. Fig. S2 (Supplementary Information) shows the dynamic resistance plots to 1, 10 and 50 ppm of different gases. As it can be seen, in the selectivity pattern provided in Fig. 9b. The response of gas sensor to 50 ppm NO2, H2, CO, C7H8, C6H6, C3H6O, C2H6O and H2S is 1.044, 1.023, 1.051, 1.029, 1.055, 1.033, 1.037 and 1.061 respectively. Even though the response to H2S gas (1.061) is higher than the response to others, the sensor suffers from lack of selectivity which can limit its application. To investigate the effect of applied voltage on the sensing behavior of Au-decorated WS2 NSs, the Au-decorated WS2 (irradiated for 1 s) gas sensor was exposed to 1, 10 and 50 ppm CO gas under different applied voltages and the dynamic resistance plots are presented in Fig. S3 (Supplementary Information). Choice of CO gas was due to well-known
Fig. 5. XPS survey spectra of (a) pristine WS2 and (b) Au-decorated WS2 NSs. (c) XPS O 1s region of pristine WS2 and (d) Au-decorated WS2 NSs. 5
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 8. (a) Dynamic resistance curves of pristine WS2 NSs to 1, 10 and 50 ppm H2S gas under different applied voltages (1–10 V). (b) Change of initial resistance with varying the applied voltage.
(irradiated for 15 s) for 50 ppm tested gases under an applied voltage of 2 V. The Au-decorated WS2 NSs gas sensor not only shows higher response to all tested gases relative to pristine WS2 NS gas sensor, but it also shows the highest response to CO gas (Ra/Rg = 1.48), which is much higher than that obtained for the pristine WS2 gas sensor (Ra/ Rg = 1.05). In the next step, the sensing behavior of Au-decorated WS2 gas sensor (irradiated for 15 s) was compared to pristine WS2 gas sensor under an applied voltage of 2 V. Fig. S6a (Supplementary Information) compares the dynamic resistance curves of pristine and Au-decorated WS2 NSs gas sensors (irradiated for 15 s) to different concentrations of CO gas and corresponding calibration curves for both gas sensors are reveled in Fig. S6b (Supplementary Information). Obviously, the response of Au-decorated gas sensor is higher than that of pristine gas sensor at all CO gas concentrations. Even though some papers reported enhanced response to NO2 gas after Au-decoration [31,32], in this study, the response to NO2 was not significantly enhanced after Audecoration. The possible reasons are explained in section 3.3 (sensing mechanism). Fig. 12a shows dynamic resistance curves of Au-decorated WS2 NSs gas sensor (irradiated for 15 s) to very low concentrations of CO gas under an applied voltage of 2 V. As shown, this sensor even shows a noticeable response to 0.2 ppm of CO gas. Fig. 12b shows the corresponding calibration histogram where the response increases with increasing of the CO concentration. In Fig. 12c, response times and recovery times for different CO concentrations are plotted. At low concentrations of CO gas, the response time is relatively long, presumably because the diffusion of CO gas to deeper parts of the sensor takes more time. Regardless of CO concentration, recovery times are
Fig. 6. (a) Dynamic resistance curves of pristine WS2 NSs gas sensor to 1, 10 and 50 ppm H2S gas at different temperatures (b) Response to 1, 10 and 50 ppm H2S gas versus operating temperature.
Fig. 7. Change of temperature of Au-decorated WS2 NSs gas sensor (irradiated for 15 s) by varying the applied voltage along with corresponding thermographs.
6
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 9. (a) Response of pristine WS2 gas sensor to different concentrations of H2S gas under different applied voltages. (b) Selectivity pattern of pristine WS2 gas sensor to 50 ppm of different gases at 2 V. Fig. 10. (a) Response of Au-WS2 gas sensor (irradiated for 1 s) to 50 ppm CO gas as a function of external voltage. For comparison, the response of pristine WS2 gas at 2 V is also shown. (b) Initial resistance of Au-WS2 gas sensor (irradiated for 1 s) as a function of external voltage. For comparison, the initial resistance of pristine WS2 gas at 2 V is also shown.
3.3. Gas sensing mechanism
shorter than response times. We surmise that injected air quickly replaces CO molecules on the surface of gas sensor during the recovery period. Fig. 13 shows performance of Au-decorated WS2 NSs gas sensor (irradiated for 15 s) in the presence of different relative humidity (RH %). Fig. 13a shows dynamic resistance plots of Au-WS2 NSs gas sensor (irradiated for 15 s) to 50 ppm CO gas under 2 V external voltage at different RH%. As shown in Fig. 13b, with increases of RH% from zero to 60%, the sensor response decreases. This is due to fact that water molecules are adsorbed on the surface of gas sensor, decreasing the available sites for adsorption of CO gas. Also, in Fig. 13c the response and recovery times in the presence of RH% are plotted. With increase of RH% both response and recovery time will be longer. When RH% increases CO molecules cannot easily reach to available sites on the surface of gas sensor. Also, with increasing RH%, desorption of CO becomes more difficult.
As it was shown in gas sensing tests, WS2 NSs gas sensor revealed ntype semiconducting behavior, in which the resistance decreased in the presence of H2S gas, which is a reducing gas. Accordingly, the main charge carriers are electrons in WS2 NSs and we will apply the mechanism of n-type gas sensors in this study using an electron depletion layer (EDL) concept. Initially, when the pristine WS2 sensor is in air at near room temperature, oxygen molecules are adsorbed on the surface of WS2 and extract electrons from conduction band of WS2 owing to their high electron affinity [8]:
O2 (air ) → O2 (ads )
(1)
O2 (ads ) + e → O2−
(2)
Therefore, adsorbed oxygen species will cause formation of an EDL on the surface of WS2 NSs and thus decreases the resistance of gas sensor. When WS2 gas sensor is exposed to H2S and CO gases gas, Fig. 11. (a) Responses of Au-decorated WS2 NSs (irradiated for 15 s) gas sensors to CO gas with varying average particle size of Au NPs and irradiation time under an applied voltage of 2 V. (b) Comparison between the selectivity of pristine and Au-WS2 (irradiated for 15 s) gas sensors to a variety of gases at the concentration of 50 ppm and under an applied voltage of 2 V.
7
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 12. (a) Dynamic resistance curves of the Au-decorated WS2 NSs gas sensor (irradiated for 15 s) to low concentrations of CO gas at 2 V and (b) Corresponding histogram of response at different concentrations. (c) Response and recovery times versus CO concentration of the Au-decorated WS2 NSs.
Fig. 13. (a) Dynamic resistance curves of Au-WS2 NSs gas sensor (irradiated for 15 s) to 50 ppm CO gas at 2 V at different RH%. (b) Histogram of response of Au-decorated WS2 NSs gas sensor (c) corresponding response and recovery times as a function of RH%.
following reaction can take place on the surface of WS2:
H2 S + 3O2− → 2H2 O + 2SO2 + 3e−
(3)
2CO + O2− → 2CO2 + e−
(4)
sensing temperature (due to self-heating), and kinetic diameter of target gases. Au NPs are known to be highly efficient CO oxidation catalysts even at low temperatures [34,35]. The good selectivity for CO gas in the presence of Au NPs is firstly related to the low oxidation barrier for complete oxidation of CO (1.20 eV). The interaction between CO and Au NPs is very complex, and depends strongly on the overlap of the 5σ and 2 π* orbitals of CO with the d orbitals of the Au NP. CO donates electrons from its σ orbitals to the dσ orbitals of Au, and the filled dπ orbitals of Au donate electrons back into the empty π orbitals of CO. Furthermore, the down-shift of the Au d-band center increases the stability of the CO/Au interface, resulting in easy conversion of CO to CO2 [25,30]. Furthermore, since target gases need an energy to overcome activation energy of adsorption, it can be surmised that excerpt for CO gas, operating temperature cannot provide enough energy for adsorption of target gases. Finally, the smaller kinetic diameter of CO (3.76 Å) relative to C6H6 (5.85 Å), C7H8 (5.9 Å) [36], H2S (3.6 Å) [37], C2H6O (4.53 Å) [38], C3H6O (4.6 Å) [39], NO2 (04.01 to 05.02 Å) [40], will result in deeper diffusion of CO gas into the sensor, bring about a higher response. To obtain work functions of Au and WS2, UPS spectra (Fig. S7 (Supplementary Information)) were used and using the same procedure presented in Ref. 30, the work functions of Au and WS2 were calculated to be 5.35 and 4.85 eV, respectively (Fig. 14a)). As shown in Fig. 14b, because of work function difference between Au and WS2, Au will act as electron acceptor and electrons are transferred from WS2 to Au.
The above reactions release captured electrons back to the WS2, thereby decreasing the width of the EDL and decreasing the resistance of WS2 gas sensor. Furthermore, in the contact area between WS2-WS2 NSs, homojunctions are formed, creating a potential barrier. In the presence of reducing gases, the height of potential barrier decreases and modulation of the resistance occurs. However, the response of the pristine WS2 gas sensor was not as high as that of the Au-decorated gas sensor. In the presence of Au NPs, owing to spill over effect of Au NPs [33], oxygen species firstly will be adsorbed on the surfaces of Au and then are dissociated to oxygen atoms and then will be transferred to the surfaces of WS2. Therefore, the number of chemisorbed oxygen ions on the surfaces of WS2 NSs increases, and the EDL will be thicker relative EDL width on pristine WS2 gas sensor. Accordingly, the initial resistance of the gas sensor increases. The existence of a high number of oxygen ion species will facilitate fast surface sensing reactions between gas molecules and surface oxygen ion species, which lead to an increase in the response of the gas sensor. Accordingly, after exposure of gas sensor to target gas molecules, a greater decrease in the resistance of gas sensor causes a higher response of the gas sensor. This mechanism works for almost all tested gases. The good selectivity to CO gas can be due to catalytic effect of Au, 8
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
than C–S NWs, due to structural reasons. Second, it is possible that more noticeable self-heating will appear at > 10 V. Generally, there is an optimal value of applied voltage where the maximum response is observed under self-heating mode of operation of gas sensor. In lower applied voltages there is no enough heat inside of the sensing layer and for higher applied voltages most of the generated heat may be dissipated to the surrounding atmosphere or substrate [42]. In this study, the maximum response was observed at 2 V. The power consumption was calculated using W = V2/R relation, where V is applied voltage and R is the measured resistance. In this study, a power consumption of 28.6 μW was obtained based on (V = 2 V) and (R = 139,865 Ω) values. Table S1 (Supporting Information) compares the power consumption of some self-heated gas sensors with our present sensor, demonstrating a relatively low-power consumption of the present sensor. Accordingly, in this particular sensors based on Au-decorated WS2 NSs, their materials, structures, and self-heating allow the sufficient sensing at a low voltage of 2 V, contributing to the low powerconsumption. The 2D materials gas sensors will have a lot of advantages, in comparison to traditional semiconductor metal oxides. 2D materials can be easily fabricated onto flat substrates and the well-developed semiconductor fabrication technologies can be synergetically employed. Also, they can be efficiently combined into flexible and stretchable substrates, enlightening their use in future devices and systems. 2D materials generally have large surface areas maximizing the interaction between the sensing layer and the target gas which is beneficial for sensing studies. [43]. With its layers being able to be exfoliated, the number of layer can be controlled, tuning the semiconducting properties [44,45]. Furthermore, 2D graphene, with its fascinating properties of zero bandgap, high carrier mobility at room temperature, and lowest noise, was revealed to be able to detect a single gas molecule [46]. Similarly, other 2D materials, such as transition metal dichalcogenides (TMD), have outstanding materials and electronic properties, being expected to overcome the limitation of current gas sensors. In particular, the WS2 has a wider working temperature range than other TMDs and it is the most thermodynamically stable. Furthermore among TMDs it has the largest mobility of the carriers, making it very useful for sensing studies [47]. However, generally the sensing layer based on the NSs will lose its most fascinating properties if the NSs aggregate during fabrication of gas sensor. In fact, the very high surface area of a single NS will dramatically decreases due to aggregation, reducing the active sites and adsorption capability for gas molecules [48]. Even though the response of 2D materials may be lower that corresponding nanowires, in some aspects they are superior. For example, Tonezzer et al. reported gas sensing properties of 1D and 2D ZnO [49]. At the same operative conditions, ZnO NWs showed a higher response compared to similar thickness NSs but 2D NSs demonstrate an improved limit of detection, due to having a larger cross-section, which increased the base current, and lowered the limit of detection. In spite of several drawbacks mentioned above, 2D materials will have enough advantages to offset the above shortcomings.
Fig. 14. (a) Band structures of WS2 and Au before contact and after contact in air. (b) Change of height of Schottky barrier formed between Au and WS2 after exposure to CO gas.
Accordingly, resulted higher initial resistance in WS2 will cause a higher sensor response because the relative change of resistance will become larger upon the adsorption/desorption of the same amount of gas species. In addition, a potential energy barrier will be developed at the interface between Au and WS2 due to the band bending which is caused by the difference in the Fermi levels of the Au and WS2. Electrons must overcome this potential energy barrier in order to cross this interface [33]. Owing to presence of potential barrier, the number of electrons in EDL of WS2 decreases, resulting in increase of the initial resistance of the gas sensor. Thus, In CO ambient, the released electrons will transfer to the WS2, leading to a decrease the height of Schottky barriers (Fig. 14b). This can contribute to the decrease of the sensor`s resistance, ultimately enhancing the sensor response. An optimal Au density (irradiation for 15 s) led to the best sensing performance. When Au density on the surfaces of WS2 was low (irradiation for 1 s and 30 s), the number of Au/WS2 heterojunctions was not enough to significantly modulate the resistance of gas sensor. Also, when the size of Au NPs is high (irradiation for 30 s), their catalytic activity decreases due to lower surface area relative to finer Au NPs. Au NPs irradiated for 1 s have the finest particle size, however, their amounts is low, being relative to other samples. Au NPs irradiated at 15 s not only have fine size but their density on the surfaces of WS2 in the highest, which causes the best sensing performance of this gas sensor. Self-heating or Joule heating effect in gas sensors means heat generation inside of the sensing layer under influence of applied voltage. WS2 NSs have an intrinsic resistivity, contributing to heat generation within the gas sensor. Indeed, since the charge carriers flows through the WS2, heat can be generated inside WS2 NSs grains. Furthermore, a considerable number of the WS2 NSs are in direct contact with each other and therefore, Joule heating in WS2-WS2 homojunctions also occurs, where the resistance for flow of electrons is high. However, in comparison to Pt-functionalized [41] and Au-functionalized [30] SnO2–ZnO C-S NW sensors the self-heating in the rage of 1–10 V is not significant. Instead, it is noteworthy that the present sensor exhibited a significant self-heating (i.e. 8.1 °C) by elevating the voltage from 0 to 1 V, which is distinguishable larger than those of Pt-functionalized (i.e. -0.2 °C) and Au-functionalized (i.e. 1.6 °C) SnO2-ZnO C-S NW sensors. It is not clear at this moment why Au-decorated WS2 NSs gas sensor exhibits the different dependence on applied voltage, in comparison to Ptfunctionalized and Au-functionalized SnO2–ZnO C-S NW sensors. First, it is possible that Au-decorated WS2 NSs generate less Joule heating
4. Conclusions In this study, Au-decorated WS2 NSs were prepared for gas sensing application. Au-decoration was performed using UV irradiation at three different times of 1, 15 and 30 s. They were fully characterized by different characterization techniques. Au-decoration at 15 s under UV irradiation resulted in the realization of the best performance gas sensor, where CO gas sensing results showed the possibility of detection of CO gas at low applied voltage with low-power consumption and sufficient gas selectivity. The results obtained in this study demonstrated the possibility of fabricating cheap, low power consumption, selective and high-performance CO gas sensors using Au-decoration on WS2 NSs. 9
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Acknowledgments
18952–18962. [24] J.-H. Kim, J.-H. Lee, A. Mirzaei, H.W. Kim, S.S. Kim, SnO2 (n)-NiO (p) composite nanowebs: gas sensing properties and sensing mechanisms, Sens. Actuators B Chem. 258 (2018) 204–214. [25] Z.U. Abideen, J.-H. Kim, A. Mirzaei, H.W. Kim, S.S. Kim, Sensing behavior to ppmlevel gases and synergistic sensing mechanism in metal-functionalized rGO-loaded ZnO nanofibers, Sens. Actuators B Chem. 255 (2018) 1884–1896. [26] S.-I. Naya, T. Kume, R. Akashi, M. Fujishima, H. Tada, Red-light-driven water splitting by Au (core)–CdS (shell) half-cut nanoegg with heteroepitaxial junction, J. Am. Chem. Soc. 140 (2018) 1251–1254. [27] F. Huang, J. Jian, R. Wu, Few-layer thick WS2 nanosheets produced by intercalation/exfoliation route, J. Mater. Sci. 51 (2016) 10160–10165. [28] V. Kumar, V. Kumar, S. Som, M.M. Duvenhage, O.M. Ntwaeaborwa, H.C. Swart, Effect of Eu doping on the photoluminescence properties of ZnO nanophosphors for red emission applications, Appl. Surf. Sci. 308 (2014) 419–430. [29] J. Ren, Z. Wang, F. Yang, R.P. Ren, Y.K. Lv, Freestanding 3D single-wall carbon nanotubes/WS2 nanosheets foams as ultra-long-life anodes for rechargeable lithium ion batteries, Electroch. Acta 267 (2018) 133–140. [30] J.-H. Kim, A. Mirzaei, H.W. Kim, S.S. Kim, Low power-consumption CO gas sensors based on Au-functionalized SnO2-ZnO core-shell nanowires, Sens. Actuators B Chem. 267 (2018) 597–607. [31] J. Liang, K. Zhu, R. Yang, M. Hu, Room temperature NO2 sensing properties of Audecorated vanadium oxide nanowires sensor, Interceram - Int. Ceram. Rev. 44 (2018) 2261–2268. [32] L. Yuan, M. Hu, Y. Wei, W. Ma, Enhanced NO2 sensing characteristics of Au modified porous silicon/thorn-sphere-like tungsten oxide composites, Appl. Surf. Sci. 389 (2016) 824–834. [33] D.R. Miller, S.A. Akbar, P.A. Morris, Nanoscale metal oxide-based heterojunctions for gas sensing: a review, Sens. Actuators B Chem. 204 (2014) 250–272. [34] I.X. Green, W. Tang, M. Neurock, J.T.Jr. Yates, Spectroscopic observation of dual catalytic sites during oxidation of CO on a Au/TiO2 catalyst, Science 333 (2011) 736–739. [35] J. Saavedra, H.A. Doan, C.J. Pursell, L.C. Grabow, B.D. Chandler, The critical role of water at the gold–titania interface in catalytic CO oxidation, Science 345 (2014) 1599–1602. [36] N.O. Chisholm, H.H. Funke, R.D. Noble, J.L. Falconer, Effect of toluene adsorption on permeation through SAPO-34 membranes, J. Mem. Sci. 560 (2018) 108–114. [37] L. Sigot, G. Ducom, P. Germain, Adsorption of hydrogen sulfide (H2S) on zeolite (Z): retention mechanism, Chem. Eng. J. 287 (2016) 47–53. [38] H. Qiu, M. Xue, C. Shen, Z. Zhang, W. Guo, Graphynes for water desalination and gas separation, Adv. Mater. (2019) 1803772. [39] T. Zhou, Y. Sang, X. Wang, C. Wu, D. Zeng, C. Xie, Pore size dependent gas-sensing selectivity based on ZnO@ ZIF nanorod arrays, Sens. Actuators B Chem. 258 (2018) 1099–1106. [40] L. Xie, F. Liu, K. Liu, X. Shi, H. He, Inhibitory effect of NO2 on the selective catalytic reduction of NOx with NH3 over one-pot-synthesized Cu–SSZ-13 catalyst, Catal. Sci. Tech. 4 (2014) 1104–1110. [41] J.-H. Kim, H.W. Kim, S.S. Kim, Self-heating effects on the toluene sensing of Ptfunctionalized SnO2-ZnO core-shell nanowires, Sens. Actuators B Chem. 251 (2017) 781–794. [42] K. Chikkadi, M. Muoth, V. Maiwald, C. Roman, C. Hierold, Ultra-low power operation of self-heated, suspended carbon nanotube gas sensors, Appl. Phys. Lett. 103 (2013) 223109. [43] S. Yang, C. Jiang, S.H. Wei, Gas sensing in 2D materials, Appl. Phys. Rev. 4 (2017) 021304. [44] K.F. Mak, C. Lee, J. Hone, J. Shan, T.F. Heinz, Atomically thin MoS2: a new directgap semiconductor, Phys. Rev. Lett. 105 (2010) 136805. [45] H. Liu, A.T. Neal, Z. Zhu, Z. Luo, X. Xu, D. Tomanek, P.D. Ye, Phosphorene: an explored 2D semiconductor with a high hole mobility, ACS Nano 8 (2014) 4033–4041. [46] F. Schedin, A.K. Geim, S.V. Morozov, E.W. Hill, P. Blake, M.I. Katsnelson, K.S. Novoselov, Detection of individual gas molecules adsorbed on graphene, Nat. Mater. 6 (2007) 652–655. [47] K. Sarvazad, M. Elahi, F. Ahmadian, A. Boochani, Hydrogen adsorption on WS2 decorated by Na and Li, Superlat. Micros. 128 (2018) 392–401. [48] X. Liu, T. Ma, N. Pinna, J. Zhang, Two‐dimensional nanostructured materials for gas sensing, Adv. Fun. Mater. 27 (2017) 1702168. [49] M. Tonezzer, T.T. Le Dang, N. Bazzanella, V.H. Nguyen, S. Iannotta, Comparative gas-sensing performance of 1D and 2D ZnO nanostructures, Sens. Actuators B Chem. 220 (2015) 1152–1160.
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2017R1A6A3A11030900 and 2016R1A6A1A03013422). Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.snb.2019.126659. References [1] M. Chhowalla, H.S. Shin, G. Eda, L.-J. Li, K.P. Loh, H. Zhang, The chemistry of twodimensional layered transition metal dichalcogenide nanosheets, Nat. Chem. 5 (2013) 263. [2] W. Yan, A. Harley-Trochimczyk, H. Long, L. Chan, T. Pham, M. Hu, Y. Qin, A. Zettl, C. Carraro, M.A. Worsley, Conductometric gas sensing behavior of WS2 aerogel, FlatChem 5 (2017) 1–8. [3] N. Huo, S. Yang, Z. Wei, S.-S. Li, J.-B. Xia, J. Li, Photoresponsive and gas sensing field-effect transistors based on multilayer WS2 nanoflakes, Sci. Rep. 4 (2014) 05209. [4] K. Kalantar‐zadeh, J.Z. Ou, T. Daeneke, M.S. Strano, M. Pumera, S.L. Gras, Two‐dimensional transition metal dichalcogenides in Biosystems, Adv. Funct. Mater. 25 (2015) 5086–5099. [5] A.S. Pawbake, R.G. Waykar, D.J. Late, S.R. Jadkar, Highly transparent wafer-scale synthesis of crystalline WS2 nanoparticle thin film for photodetector and humiditysensing applications, ACS Appl. Mater. Interfaces 8 (2016) 3359–3365. [6] Z. Jin, M. Chao, K. Xiao, L. Zhang, X.-L. Liu, Effect of WS2 particle size on mechanical properties and tribological behaviors of Cu-WS2 composites sintered by SPS, Trans. Nonferrous Met. Soc. China 28 (2018) 1176–1185. [7] M. Pumera, A.H. Loo, Layered transition-metal dichalcogenides (MoS2 and WS2) for sensing and biosensing, TrAC, Trends Anal. Chem. 61 (2014) 49–53. [8] Z. Qin, C. Ouyang, J. Zhang, L. Wan, S. Wang, C. Xie, D. Zeng, 2D WS2 nanosheets with TiO2 quantum dots decoration for high-performance ammonia gas sensing at room temperature, Sens. Actuators B Chem. 253 (2017) 1034–1042. [9] Z. Qin, D. Zeng, J. Zhang, C. Wu, Y. Wen, B. Shan, C. Xie, Effect of layer number on recovery rate of WS2 nanosheets for ammonia detection at room temperature, Appl. Surf. Sci. 414 (2017) 244–250. [10] W. Yan, M.A. Worsley, T. Pham, A. Zettl, C. Carraro, R. Maboudian, Effects of ambient humidity and temperature on the NO2 sensing characteristics of WS2/ graphene aerogel, Appl. Surf. Sci. 450 (2018) 372–379. [11] T. Xu, Y. Liu, Y. Pei, Y. Chen, Z. Jiang, Z. Shi, J. Xu, D. Wu, Y. Tian, X. Li, The ultrahigh NO2 response of ultra-thin WS2 nanosheets synthesized by hydrothermal and calcination processes, Sens. Actuators B Chem. 259 (2018) 789–796. [12] S. Leonardi, W. Wlodarski, Y. Li, N. Donato, Z. Sofer, M. Pumera, G. Neri, A highly sensitive room temperature humidity sensor based on 2D-WS2 nanosheets, FlatChem 9 (2018) 21–26. [13] Y. Chen, Y. Pei, Z. Jiang, Z. Shi, J. Xu, D. Wu, T. Xu, Y. Tian, X. Wang, X. Li, Humidity sensing properties of the hydrothermally synthesized WS2-modified SnO2 hybrid nanocomposite, Appl. Surf. Sci. 447 (2018) 325–330. [14] H.-Y. Qi, W.-T. Mi, H.-M. Zhao, T. Xue, Y. Yang, T.-L. Ren, A large-scale spray casting deposition method of WS2 Films for high-sensitive, flexible and transparent sensor, Mater. Lett. 201 (2017) 161–164. [15] M.S. Rahman, M.R. Hasan, K.A. Rikta, M. Anower, A novel graphene coated surface plasmon resonance biosensor with tungsten disulfide (WS2) for sensing DNA hybridization, Opt. Mater. (Amst) 75 (2018) 567–573. [16] S. Zhang, Y. Zhao, C. Zhang, S. Jiang, C. Yang, X. Xiu, C. Li, Z. Li, X. Zhao, B. Man, In-situ growth of AuNPs on WS2@U-bent optical fiber for evanescent wave absorption sensor, Appl. Surf. Sci. 441 (2018) 1072–1078. [17] X. Li, X. Li, Z. Li, J. Wang, J. Zhang, WS2 nanoflakes based selective ammonia sensors at room temperature, Sens. Actuators B Chem. 240 (2017) 273–277. [18] C.C. Mayorga‐Martinez, A. Ambrosi, A.Y.S. Eng, Z. Sofer, M. Pumera, Metallic 1T‐WS2 for selective impedimetric vapor sensing, Adv. Funct. Mater. 25 (2015) 5611–5616. [19] J.H. Cha, S.J. Choi, S. Yu, I.D. Kim, 2D WS2-edge functionalized multi-channel carbon nanofibers: effect of WS2 edge-abundant structure on room temperature NO2 sensing, J. Mater. Chem. A Mater. Energy Sustain. 5 (2017) 8725–8732. [20] F. Perrozzi, S.M. Emamjomeh, V. Paolucci, G. Taglieri, L. Ottaviano, C. Cantalini, Thermal stability of WS2 flakes and gas sensing properties of WS2/WO3 composite to H2, NH3 and NO2, Sens. Actuators B Chem. 243 (2017) 812–822. [21] K.Y. Ko, J.-G. Song, Y. Kim, T. Choi, S. Shin, C.W. Lee, K. Lee, J. Koo, H. Lee, J. Kim, T. Lee, J. Park, H. Kim, Improvement of gas-sensing performance of large-area tungsten disulfide nanosheets by surface functionalization, ACS Nano 10 (2016) 9287–9296. [22] D.-H. Baek, J. Kim, MoS2 gas sensor functionalized by Pd for the detection of hydrogen, Sens. Actuators B Chem. 250 (2017) 686–691. [23] A.Y. Polyakov, D.A. Kozlov, V.A. Lebedev, R.G. Chumakov, A.S. Frolov, L.V. Yashina, M.N. Rumyantseva, E.A. Goodilin, Gold decoration and photoresistive response to nitrogen dioxide of WS2 nanotubes, Chem. -A Eur. J. 24 (2018)
Jae-Hun Kim received his B.S. degree from Gyeongsang National University, Republic of Korea in 2013. In February 2015, he received M.S. degree from Inha University, Republic of Korea. He is now working as a Ph.D. candidate at Inha University, Republic of Korea. He has been working on oxide nanowire gas sensors. Ali Mirzaei received his Ph.D. degree in Materials Science and Engineering from Shiraz University in 2016. He was visiting student at Messina University, Italy in 2015. Since 2016 he is postdoctoral fellow at Hanyang University in Seoul. He is interested in the synthesis and characterization of nanocomposites for gas sensing applications. Hyoun Woo Kim joined the Division of Materials Science and Engineering atHanyang University as a full professor in 2011. He received his B.S. and M. S. degreesfrom Seoul National University and his Ph.D. degree from Massachusetts Instituteof Technology (MIT) in electronic materials in 1986, 1988, and 1994, respectively.He was a senior
10
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
University and his M.S and Ph.D. degrees from Pohang University of Science and Technology (POSTECH) in Material Science and Engineering in 1987, 1990, and 1994, respectively. He was a visiting researcher at the National Research in Inorganic Materials (currently NIMS), Japan for 2 years each in 1995 and in 2000. In 2006, he was a visiting professor at Department of Chemistry, University of Alberta, Canada. In 2010, he also served as a cooperative professor at Nagaoka University of Technology, Japan. His research interests include the synthesis and applications of nanomaterials such as nanowires and nanofibers, functional thin films, and surface and interfacial characterizations.
researcher in the Samsung Electronics Co., Ltd. from 1994 to 2000.He has been a professor of materials science and engineering at Inha Universityfrom 2000 to 2010. He was a visiting professor at the Department of Chemistryof the Michigan State University, in 2009. His research interests include the one-dimensional nanostructures, nanosheets, and gas sensors. Sang Sub Kim joined the Department of Materials Science and Engineering, Inha University, in 2007 as a full professor. He received his B.S. degree from Seoul National
11
Contents lists available at ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Realization of Au-decorated WS2 nanosheets as low power-consumption and selective gas sensors ⁎
Jae-Hun Kima, Ali Mirzaeib,c, Hyoun Woo Kimb,d, , Sang Sub Kima,
T
⁎⁎
a
Department of Materials Science and Engineering, Inha University, Incheon, 22212, Republic of Korea The Research Institute of Industrial Science, Hanyang University, Seoul, 04763, Republic of Korea c Department of Materials Science and Engineering, Shiraz University of Technology, Shiraz, Iran d Division of Materials Science and Engineering, Hanyang University, Seoul, 04763, Republic of Korea b
A R T I C LE I N FO
A B S T R A C T
Keywords: 2D WS2 Au-decoration Self-heating Gas sensor Sensing mechanism
We introduced a selective and self-heated CO gas sensor based on Au-decorated WS2 nanosheets. Au nanoparticles (NPs) were decorated on WS2 NSs using UV irradiation technique at different irradiation times (1, 15 and 30 s). SiO2 grown on Si was used as substrates and the gas sensors with top electrode configuration were fabricated. Gas sensors were able to work under self-heating mode with an applied voltage of 2 V for selective CO sensing. The results of CO gas sensing demonstrated the promising effects of Au for the enhanced response and selectivity towards CO gas. Also, the low-voltage operation at room temperature resulted in low-power consumption. This study demonstrates realization of selective CO gas sensor using a facile approach along with low power consumption.
1. Introduction Although graphene, the best-known two-dimensional (2D) material, has very exciting properties from both theoretical and practical standpoints, accessing these properties in practical applications can be a challenge. It is chemically inert and can only be activated through functionalization with other materials, a process which will sometimes cause degradation of some of its outstanding properties. In contrast, 2D layered transition metal dichalcogenides (TMDs), with the general formula MX2 where M is a transition metal (usually, Mo, W, Nb, Ti, Ta) and X is a chalcogen (S, Se, Te), possess very versatile chemistries, which provides opportunities for researchers in different fields to develop TMD-based devices with ease [1]. TMDCs can be prepared in 2D sheets with high surface-area, are semiconducting with tunable band gaps and good mobility of charge carriers (∼500 cm2/(V.s)), are made of earth-abundant materials, and exhibit low toxicity [2–5]. Among the different TMDs, tungsten disulfide (WS2) is of particular interest as it possesses desirable properties such as excellent thermal stability, oxidative resistance, favorable band structure and high phonon-limited electron mobility [6]. Semiconducting WS2 in bulk form has an indirect band gap of 1.4 eV, while a monolayer of WS2 has a wide, direct band gap of 2.1 eV [5]. The structure of WS2 is very useful for gas sensing applications. It is composed of vertically-stacked layers
⁎
of covalently bonded WeS atoms. Each layer is connected to its neighboring layers by relatively weak van der Waals forces, which allow target gases to freely diffuse between the layers [7]. Due to these advantageous structural properties, WS2 is a promising candidate for the realization of high-performance gas sensors at and near room temperature [8–11]. In addition, WS2 has also been used as humidity sensor [12,13], strain sensor [14], DNA hybridization sensor [15] and evanescent wave absorption sensor [16]. In the context of gas sensing applications, pristine WS2 gas sensors show poor response [17]. Moreover, because of strong gas adsorption, the recovery of pristine WS2 gas sensors is generally long and sometimes incomplete at room temperature, limiting applications in real environments [18]. A promising strategy to enhance gas response of WS2 nanosheets (NSs) is fabrication of composites [19]. For example, WS2/WO3 composites were reported for improved sensing performance [20]. Another promising strategy to enhance overal sensing properties of resistance-based gas sensors is decoration of noble metals on the surfaces of the sensing layer. This is due to the catalytic effect of metal and lower energy required for adsorption and interaction of gas molecules with the sensing layer [21,22]. For example, polyakov et al. reported enhanced sensing reponse of WS2 NSs decorated with Au nanoparticles [23]. In this study, we fabricated pristine and Au-decorated WS2 gas
Corresponding author at: The Research Institute of Industrial Science, Hanyang University, Seoul, 04763, Republic of Korea. Corresponding author at: Department of Materials Science and Engineering, Inha University, Incheon 22212, Republic of Korea. E-mail addresses: [email protected] (H.W. Kim), [email protected] (S.S. Kim).
⁎⁎
https://doi.org/10.1016/j.snb.2019.126659 Received 27 February 2019; Received in revised form 4 June 2019; Accepted 4 June 2019 Available online 08 June 2019 0925-4005/ © 2019 Elsevier B.V. All rights reserved.
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
(UHV) using irradiation of HeI 21.2 eV ultraviolet light. The temperature of the gas sensors in self-heating mode was measured by an infrared imaging camera (FLIR E75). The thermal resolution of the IR camera is less than 20 mK at 30 °C. Therefore, the camera can detect even the 0.1 °C-difference in temperature.
sensors for CO gas sensing. We investigated the effects of irradiation time for Au deposition on WS2 and studied the effects of applied voltage on the sensing performance of gas sensor. It was shown that selective CO gas sensors with very low power consumption (28.6 μW) can be fabricated using Au-decorated WS2 NSs. The sensing mechanism is discussed in detail. This study has opened new doors for the fabrication of low-power and selective CO gas sensors using a combination of noble metals and TMDs.
2.3. Gas sensing tests The gas sensing tests were similar to those elsewhere [24]. The fabricated gas sensors were put into the gas chamber with possibility of temperature control. By varying the mixing ratio of dry air to the target gas using mass flow controllers with a total stream of 500 sccm, gas flow was controlled during the gas sensing tests. The sensors’ resistances were recorded using a source meter as a function of time in different atmospheres. Both pristine and Au-decorated WS2 gas sensors were studied under the same sensing setup and gas flow conditions. The sensor response (R) was defined as R = Rg/Ra (for reducing gases) and R = Ra/Rg, (for NO2 as an oxidizing gas) where Ra and Rg are the resistances of the gas sensor in air and in the presence of the target gas, respectively. For self-heating experiments, different voltages were applied to the gas sensor via the source meter during measurements at room temperature. The response time was calculated as the time it took the resistance of the sensor to change to 90% of the original baseline, and the recovery time was calculated as the time needed for 90% of the signal to be recovered [25].
2. Experimental 2.1. Fabrication of gas sensors For fabrication of gas sensor, SiO2-coated Si substrate was equipped with interdigitated bi-layered electrodes comprising of Ti/Pt (50 nm/ 200 nm thick) layers that were sequentially deposited on the substrates by a sputtering process. First, a 5 mg monolayer WS2 powder (ACS Materials) with a thickness of ˜ 1 nm, was dispersed in 0.01 ml of 2propanol for 15 min. Then, 0.075 μl of the solution was drop coated (in three drops) onto the prepared substrates. Afterwards, the substrates with deposited WS2 were dried at 60 °C for 10 min. For fabrication of Au-decorated WS2, the surface of the WS2 NSs were decorated with Au NPs using an ultraviolet (UV) irradiation. First, 0.4 g of HAuCl4·H2O was dissolved in 2-propanol for 15 min. Then, the synthesized WS2 NSs were dipped into the gold solution and subsequently they were UV irradiated using a halogen lamp (intensity =0.11 mW/cm2, λ = 360 nm) at different times (1–30 s). Finally, the Au-decorated WS2 NSs were annealed at 500 °C for 30 min in N2 atmosphere. The dried Au-decorated WS2 NSs powder was dispersed in 2propanol for 15 min. Subsequent sensor`s fabrication was similar to pristine WS2 gas sensor. Fig. 1a and b schematically shows the fabrication process of pristine and Au-decorated WS2 NSs gas sensors, respectively.
3. Results and discussion 3.1. Structural and morphological analyses Fig. 2a and its inset, shows FE-SEM images of pristine WS2 NSs deposited on the substrate equipped with interdigitated electrodes. It can be seen that the WS2 almost fully covered the surface of substrate with sufficient contacts among them. Fig. 2b provides the TEM micrograph of pristine WS2 that clearly show the ultrathin nature of WS2 NSs and Fig. 2c confirms the presence of Au along with WS2 as the latticeresolved TEM image shows lattice fringes with spacings of 0.236 nm and 0.270 nm that can be indexed as the (111) and (100) planes of Au [26] and WS2, [27], respectively. Further, TEM images of pristine and Au-decorated WS2 NSs at different irradiation times are shown in Fig. 3(a)–(d) for pristine, 1, 15, and 30 s irradiated samples, respectively. These images clearly demonstrate the presence of Au NPs on the surfaces of WS2 NSs. Particle density (amount/50 nm2 WS2 NS) and size of Au NPs were calculated from TEM images and are presented in Fig. 3(e). Particle sizes increase with increasing irradiation time,
2.2. Characterization Microstructural and morphological studies were performed using field-emission scanning electron microscopy (HR-SEM SU 8010) and transmission electron microscopy (FE-TEM JEM-2100 F). Raman spectra were obtained using a LabRAM HR spectrometer (HORIBA Jobin Yvon, France) with a laser excitation wavelength of 514.5 nm. XPS spectrum was recorded using a K-Alpha system of Thermo Scientific, equipped with a monochromatic Al-Kα source (1486.6 eV). Ultraviolet photoelectron spectroscopy (UPS) measurements (Thermo Fisher Scientific Co. Theta probe) were performed in ultrahigh vacuum
Fig. 1. Schematic representation of preparation steps of pristine and Au-decorated WS2 NSs. 2
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 2. (a) FE-SEM images of pristine WS2 NSs on the substrate. Inset in (a) shows a higher magnification image. (b) TEM image of pristine WS2 NSs (c) latticeresolved TEM image of Au-decorated WS2 NSs.
(530.7 eV) and the second peak (532.6 eV) have been slightly shifted, which may be due to electronic interaction between Au and WS2. A comparison between O 1s regions of pristine and Au-decorated WS2 NSs exhibits the higher amount of adsorbed oxygen species on the surface of Au-decorated WS2 NSs, presumably due to catalytic effect of Au on the surface of WS2. Accordingly, a higher gas response can be expected for this gas sensor. Furthermore, we obtained the XPS regions of W 4f, W 5p3/2 and S 2p at different UV irradiation times (0, 1, 15, 30 s) and the results are presented in Fig. S1. As shown in Fig. S1(a), the peaks related to the W 4f7/2, W 4f 5/2 and W 5p 3/2 can be seen. Also, as shown in Fig. S1(b), the peaks related to the S 2p3/2 and S 2p1/2 can be seen. Overall, it can be seen that the peaks shift to lower binding energies with increasing of UV irradiation time. Also, for the 30 s irradiated sample, the 5p3/2 peak has been disappeared in the W4f region. Therefore, it seems that with increasing UV irradiation time, more Au particles can be reduced which resulted in more intense electronic interaction between Au and WS2 and thus a shift in the XPS peaks.
whereas particle density shows a maximum for 15 s irradiation time. It seems that 1 s irradiation is not enough to produce enough Au NPs on the surface of WS2, whereas 30 s irradiation will result in producing large amount of Au NPs, being agglomerated and larger, which eventually leads to a decrease of particle density. The Raman spectra of pristine and Au-decorated WS2 NSs, respectively, are shown in Fig. 4(a) and (b). Two broad peaks around 421 cm−1 and 352 cm−1 belong to the out-of-plane A 1 g vibration mode of S atoms and in-plane vibration mode, E21g , respectively [21]. The ratio of the A1 g to E21g peaks is 0.693 for the pristine WS2 NSs, which will increase to 0.872 for Au-decoratedWS2 NSs. Fig. 5 provides the XPS survey of Au-functionalized WS2 NSs. The XPS peaks related to O 1s, W 4p, W 4d, W 4f, S 2p, C 1s, Si 2p, Au 4d and Au 4f can be seen, which demonstrates the existence of Au on WS2 NSs. Fig. 5c and d show the deconvoluted O1 s region for pristine and Au-decorated WS2 NSs, respectively. For pristine WS2 NSs, O 1s peak can be deconvoluted into two peaks located at 530.7 eV and 532.6 eV. The first peak can be assigned to adsorbed oxygen species such as CO2, O2 and adsorbed H2O [28], on the surface of WS2. Since adsorbed oxygen directly affect the gas sensing properties of the gas sensor, this peak can predict the gas sensing properties. The second peak can be attributed to WeO bonds, which is due to the surface oxidation of WS2 during the preparation process [29]. For Au-decorated WS2 NWs, locations of the first peak
3.2. Gas sensing studies Before self-heating tests, we tested the behavior of gas sensor under external heating at different temperatures. Fig. 6a shows dynamic 3
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 3. TEM images of Au-decorated WS2 NSs under different UV irradiation times, (a) 0 s, (b) 1 s, (c) 15 s and (d) 30 s. (e) Particle density and average of particle size of Au NPs as a function of UV irradiation time.
10 V, respectively. Accordingly, it can be deduced that self-heating effect works on the gas sensors and can be used for sensing studies. We investigated the effect of applied voltage on the response of WS2 NSs, with H2S gas. Fig. 8a, shows the dynamic resistance changes of pristine WS2 gas sensor to 1, 10 and 50 ppm H2S gas under 1–10 V applied voltages without external heating. The sensor shows a response to all concentrations of H2S gas under all applied voltages. The resistance decreases with increasing the applied voltage due to the fact that higher applied voltage will cause more electrons jump to conduction band of WS2 NS (Fig. 8b). As presented in Fig. 9a, the magnitude of the response greatly depends on the applied voltage and the highest response is observed under 2 V applied voltage and with further increase of the applied voltage, the response of gas sensor decreases. Since the WS2 sensor
resistance curves of pristine WS2 NSs gas sensor to 1, 10 and 50 ppm H2S gas at various temperatures. It can be seen that the gas sensor shows n-type behavior, where the resistance decreases with introduction of H2S gas. In addition, the sensor also exhibits good reversibility, where upon stoppage of H2S gas, the resistance returns to its initial value in air. As plotted in Fig. 6b, the response of sensor increases with increasing the sensing temperature, in which higher temperatures provide more energy to promote the adsorption and reaction processes on the surface of the gas sensor. The highest gas response at 120 °C was 1.35 for 50 ppm H2S gas. Fig. 7 shows the temperature of pristine WS2 gas sensor as a function of voltage along with corresponding thermographs. The temperature of gas sensor at 0 V was 23.5 °C, which can be increased to 31.6, 31.7, 31.8, 31.9 and 32.2 °C, under an applied voltages of 1, 2, 3, 5 and
4
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
catalytic performance of Au towards CO gas. Fig. 10a shows corresponding response versus applied voltage for 50 ppm CO gas. The sensor shows a good response to CO gas under different applied voltages, with the maximum response being located at 2 V. For comparison, the response of the pristine WS2 NSs gas sensor is also shown, which clearly demonstrate the promising effect of Au for CO sensing. Fig. 10b shows the initial resistance of Au-decorated WS2 (irradiated for 1 s) as a function of applied voltage. The decrease of resistance with increasing the applied voltage will be related to the increased conduction electrons. The initial resistance of the Au-decorated WS2 gas sensor is higher than that of the pristine WS2 NSs gas sensor owing to transfer of electrons from WS2 to Au NPs, which will be further discussed in sensing mechanism section. It is anticipated that increasing the applied voltage will slightly increase the sensing temperature, promoting the sensing process. However, the trend in Fig. 10a indicates that another mechanism operates. Fig. 10b indicates that higher applied voltage resulted in lower initial resistance. Low initial resistance corresponds to a high initial conduction volume/regions. In this case, the same amount of gases adsorbed/removed with the higher conduction volume will bring about a lower sensor response [30]. After finding of the optimal applied voltage (2 V), the Au-decorated WS2 NSs gas sensors irradiated at different times were exposed to 1, 10 and 50 ppm CO gas at 2 V to find optimal gas sensor in term of response to CO gas. Fig. S4 (Supplementary Information) shows corresponding dynamic resistance curves and Fig. 11a summarizes the response values versus the Au irradiation times and Au particle sizes. For all concentrations of CO gas, the Au-decorated WS2 NSs gas sensor irradiated for 15 s showed the maximum response to this gas. This can be related to the amount and size of Au NPs on the surface of WS2 which will be discussed in the sensing mechanism section. After finding of the optimal gas sensor, in order to evaluate selectivity of Au-decorated WS2 (irradiated for 15 s) gas sensor, it was exposed to 1, 10 and 50 ppm of interfering gases and dynamic resistance curves under 2 V applied voltage are shown in Fig. S5 (Supplementary Information). Fig. 11b compares the selectivity patterns of pristine and Au-decorated WS2 NSs gas sensor
Fig. 4. Raman spectra of (a) pristine WS2 and (b) Au-decorated WS2 NSs.
showed its best performance under applied voltage of 2 V, selectivity tests in the presence of interfering gases were performed under this condition. Fig. S2 (Supplementary Information) shows the dynamic resistance plots to 1, 10 and 50 ppm of different gases. As it can be seen, in the selectivity pattern provided in Fig. 9b. The response of gas sensor to 50 ppm NO2, H2, CO, C7H8, C6H6, C3H6O, C2H6O and H2S is 1.044, 1.023, 1.051, 1.029, 1.055, 1.033, 1.037 and 1.061 respectively. Even though the response to H2S gas (1.061) is higher than the response to others, the sensor suffers from lack of selectivity which can limit its application. To investigate the effect of applied voltage on the sensing behavior of Au-decorated WS2 NSs, the Au-decorated WS2 (irradiated for 1 s) gas sensor was exposed to 1, 10 and 50 ppm CO gas under different applied voltages and the dynamic resistance plots are presented in Fig. S3 (Supplementary Information). Choice of CO gas was due to well-known
Fig. 5. XPS survey spectra of (a) pristine WS2 and (b) Au-decorated WS2 NSs. (c) XPS O 1s region of pristine WS2 and (d) Au-decorated WS2 NSs. 5
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 8. (a) Dynamic resistance curves of pristine WS2 NSs to 1, 10 and 50 ppm H2S gas under different applied voltages (1–10 V). (b) Change of initial resistance with varying the applied voltage.
(irradiated for 15 s) for 50 ppm tested gases under an applied voltage of 2 V. The Au-decorated WS2 NSs gas sensor not only shows higher response to all tested gases relative to pristine WS2 NS gas sensor, but it also shows the highest response to CO gas (Ra/Rg = 1.48), which is much higher than that obtained for the pristine WS2 gas sensor (Ra/ Rg = 1.05). In the next step, the sensing behavior of Au-decorated WS2 gas sensor (irradiated for 15 s) was compared to pristine WS2 gas sensor under an applied voltage of 2 V. Fig. S6a (Supplementary Information) compares the dynamic resistance curves of pristine and Au-decorated WS2 NSs gas sensors (irradiated for 15 s) to different concentrations of CO gas and corresponding calibration curves for both gas sensors are reveled in Fig. S6b (Supplementary Information). Obviously, the response of Au-decorated gas sensor is higher than that of pristine gas sensor at all CO gas concentrations. Even though some papers reported enhanced response to NO2 gas after Au-decoration [31,32], in this study, the response to NO2 was not significantly enhanced after Audecoration. The possible reasons are explained in section 3.3 (sensing mechanism). Fig. 12a shows dynamic resistance curves of Au-decorated WS2 NSs gas sensor (irradiated for 15 s) to very low concentrations of CO gas under an applied voltage of 2 V. As shown, this sensor even shows a noticeable response to 0.2 ppm of CO gas. Fig. 12b shows the corresponding calibration histogram where the response increases with increasing of the CO concentration. In Fig. 12c, response times and recovery times for different CO concentrations are plotted. At low concentrations of CO gas, the response time is relatively long, presumably because the diffusion of CO gas to deeper parts of the sensor takes more time. Regardless of CO concentration, recovery times are
Fig. 6. (a) Dynamic resistance curves of pristine WS2 NSs gas sensor to 1, 10 and 50 ppm H2S gas at different temperatures (b) Response to 1, 10 and 50 ppm H2S gas versus operating temperature.
Fig. 7. Change of temperature of Au-decorated WS2 NSs gas sensor (irradiated for 15 s) by varying the applied voltage along with corresponding thermographs.
6
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 9. (a) Response of pristine WS2 gas sensor to different concentrations of H2S gas under different applied voltages. (b) Selectivity pattern of pristine WS2 gas sensor to 50 ppm of different gases at 2 V. Fig. 10. (a) Response of Au-WS2 gas sensor (irradiated for 1 s) to 50 ppm CO gas as a function of external voltage. For comparison, the response of pristine WS2 gas at 2 V is also shown. (b) Initial resistance of Au-WS2 gas sensor (irradiated for 1 s) as a function of external voltage. For comparison, the initial resistance of pristine WS2 gas at 2 V is also shown.
3.3. Gas sensing mechanism
shorter than response times. We surmise that injected air quickly replaces CO molecules on the surface of gas sensor during the recovery period. Fig. 13 shows performance of Au-decorated WS2 NSs gas sensor (irradiated for 15 s) in the presence of different relative humidity (RH %). Fig. 13a shows dynamic resistance plots of Au-WS2 NSs gas sensor (irradiated for 15 s) to 50 ppm CO gas under 2 V external voltage at different RH%. As shown in Fig. 13b, with increases of RH% from zero to 60%, the sensor response decreases. This is due to fact that water molecules are adsorbed on the surface of gas sensor, decreasing the available sites for adsorption of CO gas. Also, in Fig. 13c the response and recovery times in the presence of RH% are plotted. With increase of RH% both response and recovery time will be longer. When RH% increases CO molecules cannot easily reach to available sites on the surface of gas sensor. Also, with increasing RH%, desorption of CO becomes more difficult.
As it was shown in gas sensing tests, WS2 NSs gas sensor revealed ntype semiconducting behavior, in which the resistance decreased in the presence of H2S gas, which is a reducing gas. Accordingly, the main charge carriers are electrons in WS2 NSs and we will apply the mechanism of n-type gas sensors in this study using an electron depletion layer (EDL) concept. Initially, when the pristine WS2 sensor is in air at near room temperature, oxygen molecules are adsorbed on the surface of WS2 and extract electrons from conduction band of WS2 owing to their high electron affinity [8]:
O2 (air ) → O2 (ads )
(1)
O2 (ads ) + e → O2−
(2)
Therefore, adsorbed oxygen species will cause formation of an EDL on the surface of WS2 NSs and thus decreases the resistance of gas sensor. When WS2 gas sensor is exposed to H2S and CO gases gas, Fig. 11. (a) Responses of Au-decorated WS2 NSs (irradiated for 15 s) gas sensors to CO gas with varying average particle size of Au NPs and irradiation time under an applied voltage of 2 V. (b) Comparison between the selectivity of pristine and Au-WS2 (irradiated for 15 s) gas sensors to a variety of gases at the concentration of 50 ppm and under an applied voltage of 2 V.
7
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Fig. 12. (a) Dynamic resistance curves of the Au-decorated WS2 NSs gas sensor (irradiated for 15 s) to low concentrations of CO gas at 2 V and (b) Corresponding histogram of response at different concentrations. (c) Response and recovery times versus CO concentration of the Au-decorated WS2 NSs.
Fig. 13. (a) Dynamic resistance curves of Au-WS2 NSs gas sensor (irradiated for 15 s) to 50 ppm CO gas at 2 V at different RH%. (b) Histogram of response of Au-decorated WS2 NSs gas sensor (c) corresponding response and recovery times as a function of RH%.
following reaction can take place on the surface of WS2:
H2 S + 3O2− → 2H2 O + 2SO2 + 3e−
(3)
2CO + O2− → 2CO2 + e−
(4)
sensing temperature (due to self-heating), and kinetic diameter of target gases. Au NPs are known to be highly efficient CO oxidation catalysts even at low temperatures [34,35]. The good selectivity for CO gas in the presence of Au NPs is firstly related to the low oxidation barrier for complete oxidation of CO (1.20 eV). The interaction between CO and Au NPs is very complex, and depends strongly on the overlap of the 5σ and 2 π* orbitals of CO with the d orbitals of the Au NP. CO donates electrons from its σ orbitals to the dσ orbitals of Au, and the filled dπ orbitals of Au donate electrons back into the empty π orbitals of CO. Furthermore, the down-shift of the Au d-band center increases the stability of the CO/Au interface, resulting in easy conversion of CO to CO2 [25,30]. Furthermore, since target gases need an energy to overcome activation energy of adsorption, it can be surmised that excerpt for CO gas, operating temperature cannot provide enough energy for adsorption of target gases. Finally, the smaller kinetic diameter of CO (3.76 Å) relative to C6H6 (5.85 Å), C7H8 (5.9 Å) [36], H2S (3.6 Å) [37], C2H6O (4.53 Å) [38], C3H6O (4.6 Å) [39], NO2 (04.01 to 05.02 Å) [40], will result in deeper diffusion of CO gas into the sensor, bring about a higher response. To obtain work functions of Au and WS2, UPS spectra (Fig. S7 (Supplementary Information)) were used and using the same procedure presented in Ref. 30, the work functions of Au and WS2 were calculated to be 5.35 and 4.85 eV, respectively (Fig. 14a)). As shown in Fig. 14b, because of work function difference between Au and WS2, Au will act as electron acceptor and electrons are transferred from WS2 to Au.
The above reactions release captured electrons back to the WS2, thereby decreasing the width of the EDL and decreasing the resistance of WS2 gas sensor. Furthermore, in the contact area between WS2-WS2 NSs, homojunctions are formed, creating a potential barrier. In the presence of reducing gases, the height of potential barrier decreases and modulation of the resistance occurs. However, the response of the pristine WS2 gas sensor was not as high as that of the Au-decorated gas sensor. In the presence of Au NPs, owing to spill over effect of Au NPs [33], oxygen species firstly will be adsorbed on the surfaces of Au and then are dissociated to oxygen atoms and then will be transferred to the surfaces of WS2. Therefore, the number of chemisorbed oxygen ions on the surfaces of WS2 NSs increases, and the EDL will be thicker relative EDL width on pristine WS2 gas sensor. Accordingly, the initial resistance of the gas sensor increases. The existence of a high number of oxygen ion species will facilitate fast surface sensing reactions between gas molecules and surface oxygen ion species, which lead to an increase in the response of the gas sensor. Accordingly, after exposure of gas sensor to target gas molecules, a greater decrease in the resistance of gas sensor causes a higher response of the gas sensor. This mechanism works for almost all tested gases. The good selectivity to CO gas can be due to catalytic effect of Au, 8
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
than C–S NWs, due to structural reasons. Second, it is possible that more noticeable self-heating will appear at > 10 V. Generally, there is an optimal value of applied voltage where the maximum response is observed under self-heating mode of operation of gas sensor. In lower applied voltages there is no enough heat inside of the sensing layer and for higher applied voltages most of the generated heat may be dissipated to the surrounding atmosphere or substrate [42]. In this study, the maximum response was observed at 2 V. The power consumption was calculated using W = V2/R relation, where V is applied voltage and R is the measured resistance. In this study, a power consumption of 28.6 μW was obtained based on (V = 2 V) and (R = 139,865 Ω) values. Table S1 (Supporting Information) compares the power consumption of some self-heated gas sensors with our present sensor, demonstrating a relatively low-power consumption of the present sensor. Accordingly, in this particular sensors based on Au-decorated WS2 NSs, their materials, structures, and self-heating allow the sufficient sensing at a low voltage of 2 V, contributing to the low powerconsumption. The 2D materials gas sensors will have a lot of advantages, in comparison to traditional semiconductor metal oxides. 2D materials can be easily fabricated onto flat substrates and the well-developed semiconductor fabrication technologies can be synergetically employed. Also, they can be efficiently combined into flexible and stretchable substrates, enlightening their use in future devices and systems. 2D materials generally have large surface areas maximizing the interaction between the sensing layer and the target gas which is beneficial for sensing studies. [43]. With its layers being able to be exfoliated, the number of layer can be controlled, tuning the semiconducting properties [44,45]. Furthermore, 2D graphene, with its fascinating properties of zero bandgap, high carrier mobility at room temperature, and lowest noise, was revealed to be able to detect a single gas molecule [46]. Similarly, other 2D materials, such as transition metal dichalcogenides (TMD), have outstanding materials and electronic properties, being expected to overcome the limitation of current gas sensors. In particular, the WS2 has a wider working temperature range than other TMDs and it is the most thermodynamically stable. Furthermore among TMDs it has the largest mobility of the carriers, making it very useful for sensing studies [47]. However, generally the sensing layer based on the NSs will lose its most fascinating properties if the NSs aggregate during fabrication of gas sensor. In fact, the very high surface area of a single NS will dramatically decreases due to aggregation, reducing the active sites and adsorption capability for gas molecules [48]. Even though the response of 2D materials may be lower that corresponding nanowires, in some aspects they are superior. For example, Tonezzer et al. reported gas sensing properties of 1D and 2D ZnO [49]. At the same operative conditions, ZnO NWs showed a higher response compared to similar thickness NSs but 2D NSs demonstrate an improved limit of detection, due to having a larger cross-section, which increased the base current, and lowered the limit of detection. In spite of several drawbacks mentioned above, 2D materials will have enough advantages to offset the above shortcomings.
Fig. 14. (a) Band structures of WS2 and Au before contact and after contact in air. (b) Change of height of Schottky barrier formed between Au and WS2 after exposure to CO gas.
Accordingly, resulted higher initial resistance in WS2 will cause a higher sensor response because the relative change of resistance will become larger upon the adsorption/desorption of the same amount of gas species. In addition, a potential energy barrier will be developed at the interface between Au and WS2 due to the band bending which is caused by the difference in the Fermi levels of the Au and WS2. Electrons must overcome this potential energy barrier in order to cross this interface [33]. Owing to presence of potential barrier, the number of electrons in EDL of WS2 decreases, resulting in increase of the initial resistance of the gas sensor. Thus, In CO ambient, the released electrons will transfer to the WS2, leading to a decrease the height of Schottky barriers (Fig. 14b). This can contribute to the decrease of the sensor`s resistance, ultimately enhancing the sensor response. An optimal Au density (irradiation for 15 s) led to the best sensing performance. When Au density on the surfaces of WS2 was low (irradiation for 1 s and 30 s), the number of Au/WS2 heterojunctions was not enough to significantly modulate the resistance of gas sensor. Also, when the size of Au NPs is high (irradiation for 30 s), their catalytic activity decreases due to lower surface area relative to finer Au NPs. Au NPs irradiated for 1 s have the finest particle size, however, their amounts is low, being relative to other samples. Au NPs irradiated at 15 s not only have fine size but their density on the surfaces of WS2 in the highest, which causes the best sensing performance of this gas sensor. Self-heating or Joule heating effect in gas sensors means heat generation inside of the sensing layer under influence of applied voltage. WS2 NSs have an intrinsic resistivity, contributing to heat generation within the gas sensor. Indeed, since the charge carriers flows through the WS2, heat can be generated inside WS2 NSs grains. Furthermore, a considerable number of the WS2 NSs are in direct contact with each other and therefore, Joule heating in WS2-WS2 homojunctions also occurs, where the resistance for flow of electrons is high. However, in comparison to Pt-functionalized [41] and Au-functionalized [30] SnO2–ZnO C-S NW sensors the self-heating in the rage of 1–10 V is not significant. Instead, it is noteworthy that the present sensor exhibited a significant self-heating (i.e. 8.1 °C) by elevating the voltage from 0 to 1 V, which is distinguishable larger than those of Pt-functionalized (i.e. -0.2 °C) and Au-functionalized (i.e. 1.6 °C) SnO2-ZnO C-S NW sensors. It is not clear at this moment why Au-decorated WS2 NSs gas sensor exhibits the different dependence on applied voltage, in comparison to Ptfunctionalized and Au-functionalized SnO2–ZnO C-S NW sensors. First, it is possible that Au-decorated WS2 NSs generate less Joule heating
4. Conclusions In this study, Au-decorated WS2 NSs were prepared for gas sensing application. Au-decoration was performed using UV irradiation at three different times of 1, 15 and 30 s. They were fully characterized by different characterization techniques. Au-decoration at 15 s under UV irradiation resulted in the realization of the best performance gas sensor, where CO gas sensing results showed the possibility of detection of CO gas at low applied voltage with low-power consumption and sufficient gas selectivity. The results obtained in this study demonstrated the possibility of fabricating cheap, low power consumption, selective and high-performance CO gas sensors using Au-decoration on WS2 NSs. 9
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
Acknowledgments
18952–18962. [24] J.-H. Kim, J.-H. Lee, A. Mirzaei, H.W. Kim, S.S. Kim, SnO2 (n)-NiO (p) composite nanowebs: gas sensing properties and sensing mechanisms, Sens. Actuators B Chem. 258 (2018) 204–214. [25] Z.U. Abideen, J.-H. Kim, A. Mirzaei, H.W. Kim, S.S. Kim, Sensing behavior to ppmlevel gases and synergistic sensing mechanism in metal-functionalized rGO-loaded ZnO nanofibers, Sens. Actuators B Chem. 255 (2018) 1884–1896. [26] S.-I. Naya, T. Kume, R. Akashi, M. Fujishima, H. Tada, Red-light-driven water splitting by Au (core)–CdS (shell) half-cut nanoegg with heteroepitaxial junction, J. Am. Chem. Soc. 140 (2018) 1251–1254. [27] F. Huang, J. Jian, R. Wu, Few-layer thick WS2 nanosheets produced by intercalation/exfoliation route, J. Mater. Sci. 51 (2016) 10160–10165. [28] V. Kumar, V. Kumar, S. Som, M.M. Duvenhage, O.M. Ntwaeaborwa, H.C. Swart, Effect of Eu doping on the photoluminescence properties of ZnO nanophosphors for red emission applications, Appl. Surf. Sci. 308 (2014) 419–430. [29] J. Ren, Z. Wang, F. Yang, R.P. Ren, Y.K. Lv, Freestanding 3D single-wall carbon nanotubes/WS2 nanosheets foams as ultra-long-life anodes for rechargeable lithium ion batteries, Electroch. Acta 267 (2018) 133–140. [30] J.-H. Kim, A. Mirzaei, H.W. Kim, S.S. Kim, Low power-consumption CO gas sensors based on Au-functionalized SnO2-ZnO core-shell nanowires, Sens. Actuators B Chem. 267 (2018) 597–607. [31] J. Liang, K. Zhu, R. Yang, M. Hu, Room temperature NO2 sensing properties of Audecorated vanadium oxide nanowires sensor, Interceram - Int. Ceram. Rev. 44 (2018) 2261–2268. [32] L. Yuan, M. Hu, Y. Wei, W. Ma, Enhanced NO2 sensing characteristics of Au modified porous silicon/thorn-sphere-like tungsten oxide composites, Appl. Surf. Sci. 389 (2016) 824–834. [33] D.R. Miller, S.A. Akbar, P.A. Morris, Nanoscale metal oxide-based heterojunctions for gas sensing: a review, Sens. Actuators B Chem. 204 (2014) 250–272. [34] I.X. Green, W. Tang, M. Neurock, J.T.Jr. Yates, Spectroscopic observation of dual catalytic sites during oxidation of CO on a Au/TiO2 catalyst, Science 333 (2011) 736–739. [35] J. Saavedra, H.A. Doan, C.J. Pursell, L.C. Grabow, B.D. Chandler, The critical role of water at the gold–titania interface in catalytic CO oxidation, Science 345 (2014) 1599–1602. [36] N.O. Chisholm, H.H. Funke, R.D. Noble, J.L. Falconer, Effect of toluene adsorption on permeation through SAPO-34 membranes, J. Mem. Sci. 560 (2018) 108–114. [37] L. Sigot, G. Ducom, P. Germain, Adsorption of hydrogen sulfide (H2S) on zeolite (Z): retention mechanism, Chem. Eng. J. 287 (2016) 47–53. [38] H. Qiu, M. Xue, C. Shen, Z. Zhang, W. Guo, Graphynes for water desalination and gas separation, Adv. Mater. (2019) 1803772. [39] T. Zhou, Y. Sang, X. Wang, C. Wu, D. Zeng, C. Xie, Pore size dependent gas-sensing selectivity based on ZnO@ ZIF nanorod arrays, Sens. Actuators B Chem. 258 (2018) 1099–1106. [40] L. Xie, F. Liu, K. Liu, X. Shi, H. He, Inhibitory effect of NO2 on the selective catalytic reduction of NOx with NH3 over one-pot-synthesized Cu–SSZ-13 catalyst, Catal. Sci. Tech. 4 (2014) 1104–1110. [41] J.-H. Kim, H.W. Kim, S.S. Kim, Self-heating effects on the toluene sensing of Ptfunctionalized SnO2-ZnO core-shell nanowires, Sens. Actuators B Chem. 251 (2017) 781–794. [42] K. Chikkadi, M. Muoth, V. Maiwald, C. Roman, C. Hierold, Ultra-low power operation of self-heated, suspended carbon nanotube gas sensors, Appl. Phys. Lett. 103 (2013) 223109. [43] S. Yang, C. Jiang, S.H. Wei, Gas sensing in 2D materials, Appl. Phys. Rev. 4 (2017) 021304. [44] K.F. Mak, C. Lee, J. Hone, J. Shan, T.F. Heinz, Atomically thin MoS2: a new directgap semiconductor, Phys. Rev. Lett. 105 (2010) 136805. [45] H. Liu, A.T. Neal, Z. Zhu, Z. Luo, X. Xu, D. Tomanek, P.D. Ye, Phosphorene: an explored 2D semiconductor with a high hole mobility, ACS Nano 8 (2014) 4033–4041. [46] F. Schedin, A.K. Geim, S.V. Morozov, E.W. Hill, P. Blake, M.I. Katsnelson, K.S. Novoselov, Detection of individual gas molecules adsorbed on graphene, Nat. Mater. 6 (2007) 652–655. [47] K. Sarvazad, M. Elahi, F. Ahmadian, A. Boochani, Hydrogen adsorption on WS2 decorated by Na and Li, Superlat. Micros. 128 (2018) 392–401. [48] X. Liu, T. Ma, N. Pinna, J. Zhang, Two‐dimensional nanostructured materials for gas sensing, Adv. Fun. Mater. 27 (2017) 1702168. [49] M. Tonezzer, T.T. Le Dang, N. Bazzanella, V.H. Nguyen, S. Iannotta, Comparative gas-sensing performance of 1D and 2D ZnO nanostructures, Sens. Actuators B Chem. 220 (2015) 1152–1160.
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2017R1A6A3A11030900 and 2016R1A6A1A03013422). Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.snb.2019.126659. References [1] M. Chhowalla, H.S. Shin, G. Eda, L.-J. Li, K.P. Loh, H. Zhang, The chemistry of twodimensional layered transition metal dichalcogenide nanosheets, Nat. Chem. 5 (2013) 263. [2] W. Yan, A. Harley-Trochimczyk, H. Long, L. Chan, T. Pham, M. Hu, Y. Qin, A. Zettl, C. Carraro, M.A. Worsley, Conductometric gas sensing behavior of WS2 aerogel, FlatChem 5 (2017) 1–8. [3] N. Huo, S. Yang, Z. Wei, S.-S. Li, J.-B. Xia, J. Li, Photoresponsive and gas sensing field-effect transistors based on multilayer WS2 nanoflakes, Sci. Rep. 4 (2014) 05209. [4] K. Kalantar‐zadeh, J.Z. Ou, T. Daeneke, M.S. Strano, M. Pumera, S.L. Gras, Two‐dimensional transition metal dichalcogenides in Biosystems, Adv. Funct. Mater. 25 (2015) 5086–5099. [5] A.S. Pawbake, R.G. Waykar, D.J. Late, S.R. Jadkar, Highly transparent wafer-scale synthesis of crystalline WS2 nanoparticle thin film for photodetector and humiditysensing applications, ACS Appl. Mater. Interfaces 8 (2016) 3359–3365. [6] Z. Jin, M. Chao, K. Xiao, L. Zhang, X.-L. Liu, Effect of WS2 particle size on mechanical properties and tribological behaviors of Cu-WS2 composites sintered by SPS, Trans. Nonferrous Met. Soc. China 28 (2018) 1176–1185. [7] M. Pumera, A.H. Loo, Layered transition-metal dichalcogenides (MoS2 and WS2) for sensing and biosensing, TrAC, Trends Anal. Chem. 61 (2014) 49–53. [8] Z. Qin, C. Ouyang, J. Zhang, L. Wan, S. Wang, C. Xie, D. Zeng, 2D WS2 nanosheets with TiO2 quantum dots decoration for high-performance ammonia gas sensing at room temperature, Sens. Actuators B Chem. 253 (2017) 1034–1042. [9] Z. Qin, D. Zeng, J. Zhang, C. Wu, Y. Wen, B. Shan, C. Xie, Effect of layer number on recovery rate of WS2 nanosheets for ammonia detection at room temperature, Appl. Surf. Sci. 414 (2017) 244–250. [10] W. Yan, M.A. Worsley, T. Pham, A. Zettl, C. Carraro, R. Maboudian, Effects of ambient humidity and temperature on the NO2 sensing characteristics of WS2/ graphene aerogel, Appl. Surf. Sci. 450 (2018) 372–379. [11] T. Xu, Y. Liu, Y. Pei, Y. Chen, Z. Jiang, Z. Shi, J. Xu, D. Wu, Y. Tian, X. Li, The ultrahigh NO2 response of ultra-thin WS2 nanosheets synthesized by hydrothermal and calcination processes, Sens. Actuators B Chem. 259 (2018) 789–796. [12] S. Leonardi, W. Wlodarski, Y. Li, N. Donato, Z. Sofer, M. Pumera, G. Neri, A highly sensitive room temperature humidity sensor based on 2D-WS2 nanosheets, FlatChem 9 (2018) 21–26. [13] Y. Chen, Y. Pei, Z. Jiang, Z. Shi, J. Xu, D. Wu, T. Xu, Y. Tian, X. Wang, X. Li, Humidity sensing properties of the hydrothermally synthesized WS2-modified SnO2 hybrid nanocomposite, Appl. Surf. Sci. 447 (2018) 325–330. [14] H.-Y. Qi, W.-T. Mi, H.-M. Zhao, T. Xue, Y. Yang, T.-L. Ren, A large-scale spray casting deposition method of WS2 Films for high-sensitive, flexible and transparent sensor, Mater. Lett. 201 (2017) 161–164. [15] M.S. Rahman, M.R. Hasan, K.A. Rikta, M. Anower, A novel graphene coated surface plasmon resonance biosensor with tungsten disulfide (WS2) for sensing DNA hybridization, Opt. Mater. (Amst) 75 (2018) 567–573. [16] S. Zhang, Y. Zhao, C. Zhang, S. Jiang, C. Yang, X. Xiu, C. Li, Z. Li, X. Zhao, B. Man, In-situ growth of AuNPs on WS2@U-bent optical fiber for evanescent wave absorption sensor, Appl. Surf. Sci. 441 (2018) 1072–1078. [17] X. Li, X. Li, Z. Li, J. Wang, J. Zhang, WS2 nanoflakes based selective ammonia sensors at room temperature, Sens. Actuators B Chem. 240 (2017) 273–277. [18] C.C. Mayorga‐Martinez, A. Ambrosi, A.Y.S. Eng, Z. Sofer, M. Pumera, Metallic 1T‐WS2 for selective impedimetric vapor sensing, Adv. Funct. Mater. 25 (2015) 5611–5616. [19] J.H. Cha, S.J. Choi, S. Yu, I.D. Kim, 2D WS2-edge functionalized multi-channel carbon nanofibers: effect of WS2 edge-abundant structure on room temperature NO2 sensing, J. Mater. Chem. A Mater. Energy Sustain. 5 (2017) 8725–8732. [20] F. Perrozzi, S.M. Emamjomeh, V. Paolucci, G. Taglieri, L. Ottaviano, C. Cantalini, Thermal stability of WS2 flakes and gas sensing properties of WS2/WO3 composite to H2, NH3 and NO2, Sens. Actuators B Chem. 243 (2017) 812–822. [21] K.Y. Ko, J.-G. Song, Y. Kim, T. Choi, S. Shin, C.W. Lee, K. Lee, J. Koo, H. Lee, J. Kim, T. Lee, J. Park, H. Kim, Improvement of gas-sensing performance of large-area tungsten disulfide nanosheets by surface functionalization, ACS Nano 10 (2016) 9287–9296. [22] D.-H. Baek, J. Kim, MoS2 gas sensor functionalized by Pd for the detection of hydrogen, Sens. Actuators B Chem. 250 (2017) 686–691. [23] A.Y. Polyakov, D.A. Kozlov, V.A. Lebedev, R.G. Chumakov, A.S. Frolov, L.V. Yashina, M.N. Rumyantseva, E.A. Goodilin, Gold decoration and photoresistive response to nitrogen dioxide of WS2 nanotubes, Chem. -A Eur. J. 24 (2018)
Jae-Hun Kim received his B.S. degree from Gyeongsang National University, Republic of Korea in 2013. In February 2015, he received M.S. degree from Inha University, Republic of Korea. He is now working as a Ph.D. candidate at Inha University, Republic of Korea. He has been working on oxide nanowire gas sensors. Ali Mirzaei received his Ph.D. degree in Materials Science and Engineering from Shiraz University in 2016. He was visiting student at Messina University, Italy in 2015. Since 2016 he is postdoctoral fellow at Hanyang University in Seoul. He is interested in the synthesis and characterization of nanocomposites for gas sensing applications. Hyoun Woo Kim joined the Division of Materials Science and Engineering atHanyang University as a full professor in 2011. He received his B.S. and M. S. degreesfrom Seoul National University and his Ph.D. degree from Massachusetts Instituteof Technology (MIT) in electronic materials in 1986, 1988, and 1994, respectively.He was a senior
10
Sensors & Actuators: B. Chemical 296 (2019) 126659
J.-H. Kim, et al.
University and his M.S and Ph.D. degrees from Pohang University of Science and Technology (POSTECH) in Material Science and Engineering in 1987, 1990, and 1994, respectively. He was a visiting researcher at the National Research in Inorganic Materials (currently NIMS), Japan for 2 years each in 1995 and in 2000. In 2006, he was a visiting professor at Department of Chemistry, University of Alberta, Canada. In 2010, he also served as a cooperative professor at Nagaoka University of Technology, Japan. His research interests include the synthesis and applications of nanomaterials such as nanowires and nanofibers, functional thin films, and surface and interfacial characterizations.
researcher in the Samsung Electronics Co., Ltd. from 1994 to 2000.He has been a professor of materials science and engineering at Inha Universityfrom 2000 to 2010. He was a visiting professor at the Department of Chemistryof the Michigan State University, in 2009. His research interests include the one-dimensional nanostructures, nanosheets, and gas sensors. Sang Sub Kim joined the Department of Materials Science and Engineering, Inha University, in 2007 as a full professor. He received his B.S. degree from Seoul National
11