Temperature-Based Selective Detection of Hydrogen Sulfide and Ethanol with MoS2/WO3 Composite.

A sensitive and temperature-based selective sensor toward hydrogen sulfide and ethanol using MoS2/WO3 composite as a sensing surface was developed in this work. The MoS2/WO3 nanocomposite was successfully obtained using a facile two-step method. Structural analysis revealed the successful formation of the composite. Further, the n-type semiconducting nature as revealed in the initial gas-sensing measurements was also confirmed via Mott-Schottky plots. The composite-based sensor showed preferential detection of ethanol (260 °C) and hydrogen sulfide (320 °C) by simply modulating the temperature of the sensor device. The device also displayed repeatability and long-term stability at respective operating temperatures. Improved sensitivity and selectivity are ascribed to synergistic effects arising from the formation of n-n type heterostructures. The present work indicates the potential use of composite-based heterojunctions to tune the sensing parameters and provide new possibilities to enhance the applications of MoS2 and metal-oxide semiconductor-based composites.

Introduction

In recent years, with fast industrial development, the increasing level of air pollution has become a serious issue, and hence, the necessity of gas sensors has become increasingly important in a variety of applications, such as medical diagnosis, food quality control, and environmental monitoring for pollutant tracking.[1,2] Hydrogen sulfide (H2S), a colorless, highly toxic, and flammable gas produced by numerous industrial operations such as coal mining, natural gas generation, and petroleum refining, is a byproduct of many industrial processes. H2S has a strong rotten egg smell at low concentrations and becomes odorless beyond 250 ppm, when it starts to paralyze human olfactory nerves.[3] An increase in the concentration of H2S up to 700 ppm may also cause human death.[4] Moreover, H2S may be used as a biomarker for asthma, airway inflammation, and oral and dental health. On the other hand, though consumable, ethanol’s long-term exposure may cause nose and throat irritation, nausea, liver failure, poor vision, nerve disease, and even cancer.[5,6] Furthermore, these volatile organic compounds (VOCs) can react with other chemicals in the atmosphere to form acid rain, which is a major harmful ingredient to the ecosystem. Because of such a negative impact on both human beings and the environment, there exists a need for highly selective detection and real-time monitoring of hazardous gases like H2S and various VOCs such as isopropyl, ethanol, acetone, etc.

Chemical field-effect transistor or chemiresistive and electrochemical sensors based on metal-oxides, polymers, and two-dimensional materials are among the most commonly used sensors for the detection of various toxic gases present in the environment and for biomedical purposes.[7−9] Among these, metal-oxide-based chemiresistive sensors are ideal candidates because of their low cost, simple structure, facile integration with electronics technology, and excellent sensitivity.[10−13] Their variable morphologies have governed enormous applications in the area of sensing.[12,14,15] One of the common problems related to metal-oxide semiconductor (MOS)-based sensors is their higher operating temperature and inferior selectivity. Among various MOS, WO3, an n-type semiconductor, possesses a larger bandgap, high thermal and chemical stability, and better sensing features,[16−18] except for its poor selectivity and low sensing performance at higher humidity levels.[14,19] Various other strategies, such as doping, grain size reduction, or composite formation with semiconducting materials, have been adopted to address some of these issues.[20−22] For instance, MoS2, a semiconducting transition metal dichalcogenide (TMDC), has been successfully employed in various sensing applications.[23,24] Inferior sensitivity, slow response and recovery to gas molecules, and the negative impact of humidity on sensing performance hinder its practical applications.[25] It is to be noted that MOS and layered TMDCs possess complementary properties. The heterostructures of metal-oxides (SnO2, WO3, MoO3, and TiO2) with 2D materials appear to be a promising approach for detecting various toxic gases at ambient temperature.[26−32] In a few cases, composite-based sensors have shown selective detection of a particular gas;[25] on the other hand, temperature modulation of certain MOS-TMDC composite-based sensors has also been found to be selective even for chemically identical compounds,[33] and a single sensing platform has been found to be extremely useful in distinguishing between two different analytes but at different temperatures. This technique has satisfactory performance in complex environments having simultaneous presence of various gases in the background air.[34] Details of various hybrid materials used for the detection of H2S and ethanol are given in Table . These studies were focused on the use of one particular sensing surface for selective detection of the H2S or ethanol. Herein, we demonstrate that a single MoS2/WO3 composite-based sensing surface can be used for superior and selective detection of H2S and ethanol at two different temperatures.

Table 1

Performances Comparison of Different Semiconducting Materials with Present Work

materials	synthesis route	response (%)	T (°C)	target gas	concentration	ref
WO3	electrodeposition	85	300	H2S	10 ppm	(14)
WO3/CuO	hydrothermal	105	85	H2S	5 ppm	(35)
Au-doped WO3	heat-treatment	12.4	300	H2S	2 ppm	(36)
VO2–WO3	ball milling	9.1	300	H2S	5 ppm	(37)
Ce-doped WO3	hydrothermal	12.3	350	ethanol	1 ppm	(38)
WO3-nanorods	pulsed laser deposition		400	ethanol	12.5 ppm	(15)
SnO2@MoS2	hydrothermal	160	280	ethanol	500 ppm	(39)
MoS2/TiO2	anodization-hydrothermal	10	150	ethanol	50 ppm	(40)
MoS2/WO3	hydrothermal	17	260	ethanol	5 ppm	this work
MoS2/WO3	hydrothermal	15	320	H2S	0.5 ppm	this work

In this article, we report the synthesis of MoS2/WO3 composite via a two-step facile approach. This article is divided into two major sections. The first section provides a detailed description of the synthesis and structural properties of the composite. The second part focuses on the electrical and gas sensing features of the composite. In this part, we will discuss how the same device, when operated at different temperatures, can be used for selective detection of H2S and ethanol against various reducing and oxidizing gases. We have obtained excellent sensitivity and selectivity of this composite toward hydrogen sulfide and ethanol among different analytes at optimal temperatures. The present work justifies the potential use of such hybrids for selective detection of various hazardous gases at their optimum temperatures.

Results and Discussion

Structural Analysis of MoS2/WO3 Composite

The XRD patterns of MoS2, WO3, and MoS2/WO3 composites are shown in Figure S2. The XRD pattern of MoS2 displays hexagonal symmetry (space group P63/mmc, a = 3.161 Å, c = 12.299 Å, JCPDS No. 037-1492). It consists of broad peaks centered at approximately 17.40°, 33.30°, and 57.40° which correspond to (002), (101), and (110) planes, respectively. It is to be noted that the peak width is larger than usually observed for pristine MoS2.[25,41] The XRD spectrum of WO3 consists of intense, narrow peaks matching well with the monoclinic phase of WO3 (JCPDS No. 83-0951). The spectrum for the composite consists of peaks corresponding to MoS2 as well as WO3. Some of the low-intensity peaks corresponding to WO3 have been masked in the spectrum. Further, scanning electron microscopy and high-resolution transmission electron microscopy were used to obtain detailed insight into morphology and structural properties. SEM images for different samples are shown in Figure S3. Panels a and b correspond to MoS2 and WO3, respectively. Typical particle size of the as-synthesized material lies in the micrometer range. Panels c and d display surface properties of the composite. As can be seen, MoS2 particles are uniformly distributed over WO3 and have wide variation in shape.

Figure displays the TEM images for the MoS2/WO3 composite. The HRTEM image in panel a shows parallel running planes with interplanar spacing of 3 Å corresponding to (004) planes of MoS2. The lower inset displaying the hexagonal symmetry of MoS2 is a digitally filtered image from the area highlighted in panel a. Panel b is another HRTEM image corresponding to WO3. Panel c represents a filtered image of the highlighted region in panel b. In panel c, the planes with a spacing of 3.2 (Å) correspond to (002) planes of WO3. We also note the presence of edge dislocations (lower box) in WO3. On the left side, the interplanar spacing is close to 2.7 (Å), whereas after merging of two lines the spacing increases to ∼3.1 (Å), implying the presence of strain in the processed sample.

Figure 1

(a and b) HRTEM images of MoS2/WO3 composite. (c) The highlighted area is a digitally filtered image of panel b.

Raman spectroscopy offers a nondestructive method to probe the phase changes, defects, and chemical modifications that may take place in a material during synthesis. Figure displays the Raman spectra of MoS2/WO3 composite obtained with λexc = 514.5 nm excitation at room temperature. The spectra clearly depicts two prominent Raman-active modes at 378.5 and 404.3 cm–1 corresponding to MoS2. These modes are labeled as E2g1 and A1g, respectively. The former corresponds to sulfur and molybdenum atoms moving in-plane, whereas the latter, i.e., A1g, represents sulfur atoms moving out-of-plane, which are consistent with previously reported results of hexagonal MoS2.[25,42] The difference between these prominent peaks of ∼25.8 cm–1 indicates that MoS2 has a thick layered structure. Raman peaks at 665.2, 772.8, 819.8, and 849.4 cm–1 have been attributed to stretching of O–W–O bonds, whereas the low-intensity modes at 337.2, 288.2, 242.3, 197.6, and 155 cm–1 represent O–W–O bending modes of WO3.[43,44] The Raman spectra, therefore, indicate that the MoS2/WO3 composite is successfully formed.

Figure 2

Raman spectrum of MoS2/WO3 composite with an excitation wavelength of 514.5 nm acquired at room temperature.

The specific surface area and porosity of the sensing layer are widely established as important parameters in influencing the gas-sensing characteristics of semiconductor-type gas sensors.[33] Nitrogen adsorption–desorption isotherms were used to investigate the specific surface area and porosity of MoS2 and MoS2/WO3 samples. The isotherms of MoS2 and MoS2/WO3 (1:1) are shown in Figure , whereas those for 1:3 and 3:1 compositions are shown in the Supporting Information (Figure S4). On the basis of International Union of Pure and Applied Chemistry (IUPAC) classification, both samples exhibit type-IV isotherms with a small H3 hysteresis loop,[45] implying small pore volumes are supplied by mesopores, thus providing efficient channels for mass transport, larger for the case of MoS2/WO3.[25,41,42] The physical parameters obtained from the isotherms are summarized in Table . As can be seen, the composite with a weight ratio of 1:1 has shown nearly 4-fold enhancement in specific surface area compared with MoS2 alone. The observed enhancement is larger when compared with other composites. Whereas the 1:1 composite has shown the largest enhancement in specific surface area, the average pore size was found to be maximum for composite with a 1:3 weight ratio. Above all, the 1:1 composite has shown the best results. These improvements may result in better sensing response from composite-based sensors.[41,42,46]

Figure 3

Nitrogen adsorption–desorption isotherms obtained at 77 K of (a) MoS2 and (b) MoS2/WO3 (1:1) composite.

Table 2

Specific Surface Area Sbet and Average Pore Size of MoS2 and MoS2/WO3 Composite with Different Ratios

sample	MoS2	MoS2/WO3 (1:1)	MoS2/WO3 (3:1)	MoS2/WO3 (1:3)
Sbet (m2 g–1)	4.86	18.4	4.79	13.5
average pore size (Å)	99.5	119.0	204.7	446.8

XPS measurements were used to analyze the surface composition and elemental chemical state of the composite. Figure S5 (Supporting Information) shows the entire survey spectra of the MoS2/WO3 composite indicating the presence of all elements. Panels a–d of Figure represent the high-resolution XPS spectra of Mo, S, W, and O in the composite. The Mo 3d XPS can be deconvoluted into five peaks.

Figure 4

High-resolution XPS spectra and peak positions of (a) Mo 3d, (b) S 2p, (c) W 4f, and (d) O 1s spectrum of MoS2/WO3 nanocomposite.

The spectrum in panel a consists of peaks at 228.6 and 231.5 eV, corresponding to Mo4+ 3d5/2 and Mo4+ 3d3/2, respectively. These belong to +4 oxidation state of Mo and govern the highest signal in the spectra, while the other low-intensity peak at 232.1 and another one at 235.6 eV corresponds to Mo6+, implying the presence of MoO3 or MoO42–, which may be formed during the synthesis of the materials.[21,33,42] The Mo4+ peaks are at 228.6 and 231.5 eV, and the S 2p region matches the 2H-phase of MoS2.[21,42] In panel b, the two major peaks of S 2p centered at 161.4 and 163.0 eV due to S2– 2p3/2 and S2– 2p1/2 can be seen.[21,42] The low-intensity high-energy component at 168.5 eV can be assigned to S6+ species in sulfate groups (SO42–).[9] Panel c displays the prominent peaks for the W6+ oxidation state, which correspond to W 4f7/2 and W 4f5/2 at 38.7 and 36.0 eV, respectively.[41,47] In addition to these two peaks of W6+, one more low-intensity peak at the binding energy of 32.4 is detected, confirming the presence of the W5+ oxidation state in the hybrid materials.[48] Two peaks of the O 1s spectra in panel d correspond to oxygen in WO3. Deconvolution of the O 1s singlet indicates two main components: lattice oxygen Olattice at 530.6 eV and surface adsorbed water peak at 532.1 eV.[16] In conclusion, the peaks corresponding to Mo 3d, S 2p, W 4f, and O 1s in the composite are moved to higher binding energies as compared to individual MoS2 and WO3. This indicates successful formation of the composite and the presence of strong electronic interaction between MoS2 and WO3.[41]

Electrical Characteristics

Current–voltage (I–V) characteristics on two-terminal devices at various temperatures were used to examine the physical properties of the MoS2/WO3 composite. Figure S6 (Supporting Information) shows that the resistance ratio (R(T)/R(T0)) decreases with increasing temperature, resulting in a negative first-order temperature coefficient, α = −5.83 × 10–3 °C–1, which differs slightly from hydrothermally grown MoS2[49] and liquid exfoliated MoS2/WO3 nanosheets.[41] The negative temperature coefficient indicates the semiconducting nature of the composite material.

Ethanol and Hydrogen Sulfide Sensing Performance of MoS2/WO3 Composite

Temperature plays an important role in determining the sensing properties of a sensing device. Because of the high energy barrier for the reaction, adsorption of gas molecules and, consequently, their reaction with adsorbed oxygen on the sensing surface cannot occur when temperature is lower than a certain optimum value.[50] On the other hand, at sufficiently high temperatures, the accelerated desorption of gas molecules from the sensing surface gives rise to a reduced sensing response.[51] Only at the optimum operating temperature does it have adsorption and desorption equilibrium to achieve the highest response. The optimum temperature was confirmed by measuring the response of different samples toward 100 ppm ethanol and H2S at different intervals of temperatures (50–400 °C), and the data are shown in Figure S7 (Supporting Information). As can be seen, with temperature rises, the relative response increases up to 260 °C (320 °C) for ethanol (hydrogen sulfide), and beyond that it decreases further. It is to be noted that Figure S7 displays data for three different weight ratios (3:1, 1:1, and 1:3) of MoS2/WO3. Increased WO3 content results in a slight shift of operating point toward higher temperature. On the other hand, the MoS2/WO3 composite with weight ratio 3:1 does not exhibit significant change in optimum temperature but has a low sensitivity. Among three composites, the one with a weight ratio of 1:1 (MoS2:WO3) has better sensing response at optimum temperature for both ethanol and hydrogen sulfide. A plausible explanation for the observed behavior could be the availability of effective surface area for making an intimate contact between MoS2 and WO3. From this point onward, the discussion is limited to only MoS2/WO3 composite with a weight ratio of 1:1.

Figure shows one of the representative response transients (resistance vs time) obtained at 320 °C with 50 ppm hydrogen sulfide. The sensor resistance changes from Ra = 58.7 kΩ (base resistance in air) to 30.2 kΩ (in hydrogen sulfide). This diagram defines three major sensor device parameters. The relative response of a sensor is defined as ΔR/Ra = (Rg – Ra)/Ra, where Ra and Rg are the device resistances in air and air–gas mixture, respectively. The sensor response time (tresponse = 49 s) refers to the time taken by the sensor to achieve 90% of minimum change in resistance (ΔR) with respect to base resistance in air. Identically, recovery time (trecovery = 57 s) refers to the time it takes to recover from the maximum resistance value Rg (in H2S) to 10% below the base resistance (Ra) value.

Figure 5

Representative measurement for 50 ppm hydrogen sulfide at 320 °C. This figure defines various parameters related to a sensor device.

The physical properties of a material in nanostructured form differ from those in bulk. For example, in nanostructured form they have a high surface-to-volume ratio, numerous surface-active sites, a high surface reactivity, a change in optical band gap, and processing-induced defects. These modifications may have an impact on the gas-sensing capabilities of a sensor device. Furthermore, as previously discussed and illustrated in Figure S8 (Supporting Information), composite devices have demonstrated better gas-sensing characteristics as compared to individual materials at their optimal temperatures for 100 ppm ethanol and H2S, respectively. It is to be noted that as-obtained pristine and hydrothermally synthesized MoS2 have shown different physical properties when they were used as gas sensors. They possess different morphological features; for instance, the former possesses well-organized crystalline and layered features, whereas the latter has been found to have variable morphology depending upon synthesis conditions.[41,42] Surprisingly, hydrothermally synthesized MoS2 exhibits n-type conduction, whereas pristine MoS2 displays p-type conduction.[33,41,42] To gain insight into conductivity type, Mott–Schottky graphs have also supported the conductivity behavior observed in sensing measurements. Different types of defects in the same material which is synthesized using two different processes may cause variations in conductivity type. But such theoretical investigations have remained elusive to date. For obtaining MS plots, three distinct working electrodes of MoS2, WO3, and MoS2/WO3 were utilized. Figures S9 and S10 depict the obtained MS plots. As can be seen, MoS2, WO3, and the composite (MoS2/WO3) with different weight ratios exhibit dominating n-type behaviors (positive slope), thus supporting the conductivity type observed in gas sensing.

The sensor’s response to ethanol concentrations ranging from 5 to 90 ppm is shown in Figure a,b. The Freundlich isotherm model can be used to represent the relationship between sensing response and gas concentration (C) in ppm.[52] A regression equation of sensing response (ΔR/Ra%) = A·C with good coefficients of R2 = 0.98 for temperatureof 260 °C (A = 11.2, b = 0.37) can be used to fit the experimental data. The obtained tresponse lies between 80 and 100 s, whereas the trecovery has a slightly wider variations over the investigated range of ethanol concentrations.

Figure 6

(a) Response-recovery transients for various ethanol concentrations and (b) corresponding absolute relative response vs ethanol concentration (ppm). (c) Stability test for continuous five cycles with 70 ppm ethanol. (d) Change in sensor response with different RH levels (20–95%) at constant ethanol concentration of 100 ppm. All measurements performed at 260 °C.

Real-world uses of a sensor device are subjected to repeatability and variations in relative response over time. The sensor was tested for repeatability at 260 °C with 70 ppm ethanol. The results for five sequential response–recovery cycles presented in Figure c reveal variations of less than ±4%, verifying the stability of MoS2/WO3 composite-based sensor devices.

Furthermore, the effect of humidity or the presence of water molecules has a major effect on the sensing performance of the devices. The increased humidity level has a number of consequences. Excess water molecules resulting from increased humidity level not only limit the ethanol adsorption but also impact the gas-sensing mechanism by removing the surface adsorbed oxygen. This is known as water poisoning, and it reduces the sensor’s sensitivity.[53] The device was tested for varying RH levels (20–95%) at a fixed (100 ppm) ethanol concentration. As can be seen in Figure d, the sensor’s response decreases (13% down) with increase in humidity level, thus indicating that the sensor’s performance slightly declines with increased RH level.[54] However, at a constant humidity level, the sensor exhibited nearly reliable response over the measured range of ethanol concentration.

Figure a shows the real-time resistance variations (in kΩ) of a MoS2/WO3 composite sensor for various H2S concentrations (0.5–90 ppm) at an optimum temperature of 320 °C. The relative response follows a Freundlich isotherm model (regression coefficient R2 = 0.97) with the increasing level of H2S concentration (panel b). However, the associated response (90 ± 20 s) and recovery times (100 ± 20 s) have a different range when compared with the previous situation. Moreover, the stability test with 50 ppm H2S at 320 °C gave less than ±3% fluctuations in relative response as shown in Figure c. Further, a change of 16% in relative response was observed when RH changed from 20 to 95% (panel d). This change is a bit larger when compared to ethanol sensing with varying RH level. Furthermore, long-term stability is an important aspect that determines the sensing performance of a sensing device. Figure a,b shows the response transients for 100 ppm ethanol and 50 ppm hydrogen sulfide measured with a one-week interval. No apparent decline of response within 56 days was seen, implying satisfactory long-term stability for practical use as gas sensors.

Figure 7

(a) Sensor response (resistance) curves for different H2S concentrations (ppm). (b) Corresponding absolute relative response vs H2S concentration (ppm). (c) Stability test for continuous six cycles with 50 ppm H2S. (d) Change in relative response with RH (20–95% RH) for fixed H2S concentration (100 ppm). All measurements performed at 320 °C.

Figure 8

Long-term durability of MoS2/WO3 composite-based sensor for 8 weeks: (a) 100 ppm ethanol and (b) 50 ppm H2S at a temperature of 260 and 320 °C, respectively.

Selectivity of Sensing Device

Selectivity refers to the strong adsorption of target gases in an air environment, while being insensitive to other gases. Because the device has shown different optimum temperatures for ethanol and hydrogen sulfide, the selectivity test was carried at 260 and 320 °C, respectively, for these analytes. The sensor was separately exposed to seven different reducing gases such as hydrogen sulfide, ammonia, formaldehyde, isopropyl, ethanol, methanol, and acetone and two oxidizing gases, nitrogen dioxide (NO2) and carbon monoxide (CO). Contrary to the universal viewpoint about CO, which behaves like reducing gas, here a different trend was seen with the MoS2/WO3 composite-based sensor. It is not surprising as an identical situation exists with H2O molecules, which behave like oxidizing gas in graphene and MoS2-based sensors, whereas it behaves like a reducing gas in other studies involving metal oxides.[55,56] Oxidizing or reducing properties of the adsorbed gas molecules and hence the charge transfer largely depend on the nature of interaction between sensing surface and target gas molecules. The obtained results are shown in Figure a,b. The sensor exhibited a higher response for 100 ppm ethanol (at 260 °C) and H2S (320 °C) as compared to other analytes (200 ppm each) at respective temperatures. Contrary to reducing gases, the sensor showed an increase in resistance when it was exposed to the oxidizing gases. This is because that when the n-type composite is exposed to oxidizing gases such as NO2 and CO, the electron transfer from MoS2/WO3 to the gas molecules gives rise to a decreased carrier density in the composite material, thus increasing its electrical resistance. Various factors, for instance, the gas molecule adsorption on the sensing surface at different operating temperatures and lowest unoccupied molecule orbit (LUMO) energy of the gas molecule, have been found to influence the selectivity of a sensor device.[42] The lower LUMO energies facilitate gas molecule detection at lower operating temperatures.[57] The orbital energy of gas molecules influences its electron affinity, whereas lower LUMO energy enhances the electron-capturing capability of the gas molecules. As a result, the electron transfer between the sensing surface and the adsorbed gas molecule is enhanced and the sensitivity of the sensor device increases. For ethanol and H2S, corresponding LUMO energies are 0.125 and 0.685 eV, respectively.[58] The lower LUMO energy for ethanol enables its detection at lower temperatures.

Figure 9

Selectivity test of the MoS2/WO3 composite sensor toward 100 ppm ethanol and hydrogen sulfide: (a) T = 260 °C and (b) T = 320 °C, respectively.

Sensing Mechanism

In one of our previous reports on MoS2/WO3-based composites, we employed density functional theory simulations to calculate the adsorption energy (Eads) for various reducing gases. The computed Eads values for hydrogen sulfide and ethanol were −0.291 and −0.070 eV, respectively.[41] Thus, the adsorption energies on the MoS2/WO3 surface suggest a spontaneous exothermic process. Typically, WO3 is an n-type wide band gap semiconductor material, whose gas sensitivity is mainly considered to be due to absorbed oxygen on its surface. Adsorbed oxygen on the WO3 surface captures the electrons from the conduction band of WO3 and forms oxyanions (O2–, O–, and O2–). This results in a decrease of conductivity of the oxide. The adsorbed gas molecules of reducing gases react with oxyanions to release electrons back to WO3, and hence, conductivity increases.[50,59] Similarly, for MoS2, the adsorption of reducing gas is followed by the transfer of electrons to MoS2, thereby resulting in a decrease in the resistance.[25,60] For MoS2/WO3 composite forming n–n heterojunctions, the gas-sensing mechanism appears to be different. The superior sensing response of the composite may be attributed to the enhanced specific surface area which in turn provides more adsorption sites for the gas molecules and a synergistic effect that arises from the formation of an n–n heterostructure between MoS2 and WO3. The space charge or depletion layer model has been found to be useful to explain the gas-sensing mechanism in such cases. The proposed sensing mechanism is displayed in Figure .

Figure 10

Energy band diagram of n-MoS2/n-WO3 heterojunction after making a contact.

The band gaps of MoS2 and WO3 are 1.8 and 3.1 eV, respectively.[41] The work function represents the amount of energy required to remove an electron from the Fermi level to vacuum. Environmental effects have been found to influence these intrinsic material parameters. For instance, air exposed (O2 adsorbed) MoS2 has a work function of 4.47 eV, whereas bare MoS2 possesses a value of 4.04 eV.[61] Therefore, for air-exposed MoS2 the work function (4.47 eV) is lower than that of WO3 (5.7 eV).[62,63] After heterostructure formation, the work function difference causes electron migration from MoS2 to WO3 until equilibrium is reached.[62,63] This process results in a depletion layer between two different n-type materials. Any charge-transfer event between the composite surface and adsorbed gas molecule will result in modulation of the depletion layer and hence influence the electrical conduction across it. The depletion layer formation is accompanied by enhanced electron concentration on the sensor (composite surface). This in turn improves oxygen adsorption. The air-exposed MoS2/WO3 surface, after interaction with ambient oxygen, forms oxygen ions. This process may involve charge transfer between oxygen and WO3 or the composite surface.[21,64] For the oxygen adsorbed composite surface, upon interacting with reducing gas molecules like H2S or ethanol, electron transfer from the adsorbed gas molecule to the adsorbed oxygen takes place. Subsequently, the electron released by the oxygen back to the metal-oxide semiconductor causes a reduction in the depletion layer width. This causes a decrease in the measured electrical resistance as noted in the measurements.

Conclusion

We have demonstrated a facile two-step synthesis method for making MoS2/WO3 composite. The XPS and Raman spectroscopy have verified the formation of the composite. Further, the temperature-based gas-sensing performance of the as-prepared MoS2/WO3 hybrid toward ethanol and hydrogen sulfide were investigated at 260 and 320 °C, respectively. The composite-based sensor displayed excellent sensing performance in terms of long-term durability, repeatability, and selectivity toward ethanol and hydrogen sulfide. Importantly, one single sensing surface was used for selective detection of ethanol and H2S at different optimum temperatures. The improved sensing performance is attributed to synergistic effects caused by the formation of n–n heterostructures between MoS2 and WO3. Too low and extremely large device resistance inhibit the gas detection process. Therefore, device resistance optimization is necessary to minimize the operating temperature. Although the operating temperature is a bit on the higher side, grain size control may help in optimizing the device resistance and hence the operating temperature of the device. The present work indicates the potential use of composites in obtaining sensors with superior performance.

Experimental Section

Precursors and Synthesis of MoS2/WO3 Composite

Pure tungsten powder, hydrazine (N2H4), and ammonium tetrathiomolybdate ((NH4)2MoS4) were acquired from Sigma-Aldrich, India. Other chemicals, for instance ethanol, hydrogen peroxide (H2O2), and sulfuric acid (H2SO4), were purchased from Loba Chemicals, India. All of the precursors were used as obtained without any further processing. A two-step synthesis technique as shown in Figure was used to make the MoS2/WO3 composites. Initially, to obtain WO3 powder, 0.5 g of pure tungsten powder was mixed with 10 mL of ethanol and stirred for 0.5 h. This resulted in a transparent solution. Upon the addition of 5 mL of H2O2 and 5 h of continuous stirring at 100 °C, a bright yellow solution was obtained.[30,65,66] This solution was evaporated for 6 h in a universal oven at 60 °C. The resultant yellowish WO3 powder was dried at 65 °C for 10 h in a vacuum oven.

Figure 11

Schematic representation of two-step synthesis route for MoS2/WO3 nanocomposite.

The next step involved the synthesis of MoS2/WO3 composites with three distinct weight ratios (1:1, 1:3, and 3:1). The various synthesis steps are depicted in Figure . For obtaining a composite, 300 mg of WO3 powder (obtained in step 1) was mixed with 50 mL of distilled water to make a WO3 suspension. After that, 0.5 g of (NH4)2MoS4 and 12 mL of hydrazine were added and stirred for another 2 h. The resultant solution was transferred to a Teflon-lined stainless steel autoclave with a capacity of 100 mL and was kept at 200 °C for 12 h. Black MoS2/WO3 composite precipitates were collected after cooling naturally. The obtained precipitates were washed several times with ethanol and distilled water before being dried in a vacuum oven at 60 °C for 8 h. Therefore, a total of five samples—hydrothermally synthesized MoS2, WO3, and MoS2/WO3 composites with weight ratios of 1:1, 1:3, and 3:1—were used for comparing their gas-sensing performance.

Characterization, Sensor Fabrication, and Measurements

For investigating the structural properties and identifying the phases of the processed samples, X-ray diffraction (XRD) patterns were acquired using a Bruker D8-discover diffractometer with a Cu Kα radiation source (λ = 1.5406 Å). Morphology of the samples was investigated using scanning electron microscopy (FE-SEM, Supra 55, Carl Zeiss operating at 20 keV). A high-resolution transmission electron microscope (JEOL, JEM-2100, Japan) was used to investigate the structural properties of the synthesized material. The high-resolution Raman spectrometer (Micro-Raman spectrometer, Renishaw) with laser excitation at 514.5 nm was used to examine vibrational characteristics. The chemical states of the as-prepared MoS2/WO3 composite were confirmed using X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific K-Alpha). The surface area and pore size distribution of the samples were measured using a Micromeritics (ASAP 2020) adsorption/desorption analyzer at liquid nitrogen temperature (77 K). An electrochemical workstation was used to perform the Mott–Schottky (MS) studies (M204 Autolab, The Netherlands). The detailed procedure for performing MS type measurements and working electrode preparation is described in our previous work.[41] In short, a three electrode system employing a Ag/AgCl electrode, a platinum wire, and MoS2/WO3 as a reference electrode, counter electrode, and working electrode, respectively, was used. The MS test was carried out in a 0.5 M H2SO4 solution. For sensor fabrication, 0.2 mg of black dried powder of MoS2/WO3 was mixed with 5 μL of DI water to make a homogeneous paste. A quartz substrate with predeposited silver electrodes and a separation of approximately 2 mm was used for making two-terminal devices. With the help of a paintbrush, the black paste was applied on a quartz substrate. This was followed by vacuum drying at 80 °C for 10 h.

A Keithley-2612A SourceMeter was used to measure two terminal current–voltage (I–V) characteristics. A home-built measurement setup (Figure S1), as detailed elsewhere[21,25,41,42,67] and briefly described in the Supporting Information, was used to investigate the gas-sensing characteristics. The relative humidity level inside the measurement chamber was maintained at a set temperature using an Omron Ultrasonic Nebulizer (model NE-U17) and measured with the help of a digital hygrometer.