Abstract

In this paper, g-C3N4-WO3 composite materials were prepared by hydrothermal processing. The composites were characterized by means of X-ray powder diffraction (XRD), scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), and N2 adsorption-desorption, respectively. The gas sensing properties of the composites were investigated. The results indicated that the addition of appropriate amount of g-C3N4 to WO3 could improve the response and selectivity to acetone. The sensor based on 2 wt% g-C3N4-WO3 composite showed the best gas sensing performances. When operating at optimum temperature of 310°C, the responses to 1000 ppm and 0.5 ppm acetone were 58.2 and 1.6, respectively, and the ratio of the 1000 ppm acetone to 1000 ppm ethanol reached 3.7.

1. Introduction

Graphitic carbon nitride (g-C3N4) nanomaterial exhibits a stable layered structure and п-conjugated s-triazine unit composed of sp2 hybridized carbon atoms and sp2 hybridized nitrogen atom. g-C3N4 nanosheets have attracted the attention of researchers in recent years for its peculiar properties as a semiconductor such as immense specific surface area [1]. Wang et al. [2] prepared g-C3N4 by thermal treatment of glucose and urea, and the p-type sensor based on g-C3N4 exhibited good response to NO2 at room temperature.

As a gas sensing material, WO3 has been paid much attention in the past decade. Cho et al. [3] used ultrasonic spray pyrolysis to prepare WO3 hollow spheres using a citric acid-containing precursor solution; the WO3 hollow spheres exhibited a high response and good gas sensing selectivity to trimethylamine, but the sensor exhibited a depressed response to NO2. Kida et al. [4] used acidification of Na2WO4 with H2SO4 solution to prepare lamellar-structured WO3 particles which had a high response (–280) even to dilute NO2 (50–1000 ppb) in air at 200°C. A study by Ma et al. [5] showed that WO3 nanoplates obtained through a topochemical transformation of the corresponding H2WO4 precursor exhibited high response to ethanol while operating at 300°C.

The photocatalytic activity of g-C3N4-WO3 nanocomposites also has been reported by many researchers [69]; the photocatalytic performances of WO3/g-C3N4 nanocomposites were higher than those of pure WO3 and pure g-C3N4, which were attributed to the synergistic effect of WO3 and pure g-C3N4. A study by Zhang et al. [10] showed that the ethanol sensing performance of α-Fe2O3/g-C3N4 nanocomposites was better than that of pure α-Fe2O3 and g-C3N4, which could be caused by porous α-Fe2O3 nanotubes wrapped by lamellar g-C3N4 nanostructures resulting in the formation of heterojunction. Cao et al. [11] reported that the gas sensing response and selectivity to ethanol could also be enhanced by modification of g-C3N4 nanosheets. The combination of WO3 and g-C3N4 may exhibit good gas sensing properties, which to the best of our knowledge has not been reported to date.

In this paper, we report the preparation of g-C3N4-WO3 nanocomposites through a hydrothermal method and the investigation of their gas sensing properties. Analysis showed that 2 wt% g-C3N4-WO3 nanocomposite responded highly and selectively to acetone.

2. Materials and Methods

2.1. Material Preparation and Characterization

g-C3N4 was prepared by heating 2.0 g melamine in an oven at 520°C for 5 hours, while keeping the heating rate at 5°C/min, which was similar to that reported in the literature [12]. After cooling to room temperature naturally, the product was purified with ethanol, following which it was dried at 60°C for 24 h succeeded by milling.

For preparing the nanocomposites, a certain amount of as-prepared g-C3N4 was added to 40 mL deionized water and sonicated for 1 hour to obtain a g-C3N4 suspension. 0.0025 mol Na2WO4·2H2O was dissolved in 20 mL deionized water, and 4 mL concentrated hydrochloric acid was added dropwise in the Na2WO4 solution slowly while stirring resulting in the formation of H2WO4; the g-C3N4 suspension was added slowly to H2WO4 while stirring. The mixture was sealed in a 100 mL Teflon-lined stainless steel autoclave and heated at 200°C for 24 h; the obtained precipitate was filtered and washed with distilled water and ethanol, followed by drying in air at 80°C for 24 hours; finally, the g-C3N4-WO3 composite was obtained. The weight ratios of g-C3N4 powders/WO3 (the weight of WO3 was calculated according to the weight of Na2WO4·2H2O) were 0 wt%, 1 wt%, 2 wt%, 3 wt%, and 4 wt% (the samples were labeled as S-0, S-1, S-2, S-3, and S-4, respectively).

X-ray diffraction (XRD, Bruker D8 Advance, Cu-Kα radiation:  nm), operating at 40 kV and 30 mA in a range from 10° to 70° at room temperature, was used to analyze the crystal structure of g-C3N4-WO3 nanocomposites. A scanning electron microscope (SEM, Hitachi S-4800 microscope), with an accelerating voltage of 10 kV, was used to characterize the surface morphology of the samples. Fourier transform infrared spectroscopy (FTIR, Nicolet 6700 FTIR Spectrometer) spectra were recorded by the KBr pellet technique in the range 400–4000 cm-1. The chemical species of elements were analyzed by X-ray photoelectron spectroscopy (XPS, Thermo ESCALAB250Xi) with all of the binding energies corrected according to contaminant carbon (). The specific surface areas were characterized by the Brunauer-Emmett-Teller (BET, ASAP2010C) method using N2 adsorption-desorption measurement.

2.2. Gas Sensing Measurement

The sensor device preparation process and the gas sensing measurement have been explained in the previous work [13]. The gas sensing response of the gas sensor was defined through the ratio of the resistance of the gas sensor in air () to that in the test gases (). Figure S1, representing the gas sensor, is shown in supplementary materials.

3. Results and Discussion

3.1. Characterization

Figure 1 shows the XRD patterns of pure g-C3N4, g-C3N4-WO3 (S-1, S-2, S-3, and S-4), and WO3. Two significant diffraction peaks were observed at 13.23° and 27.86° in the XRD pattern of g-C3N4, indicating the (100) and (002) planes of layered g-C3N4; the weaker peak at 13.23° indicates the in-planar tris-s-triazine structural packing, and the stronger peak at 27.86° corresponds to the stacked of the aromatic systems between layers [68, 14]. All the diffraction peaks observed in XRD patterns of S-0, S-1, S-2, S-3, and S-4 could be indexed to the standard data of WO3 (JCPDS 43-1035). The diffraction peaks of g-C3N4 could not be found because of the low concentration of g-C3N4 in the composites [14]. It was reported that the diffraction peaks of g-C3N4 did not appear in the XRD patterns of g-C3N4-WO3 composites when the content of WO3 was higher than 10 wt% [15]. The average crystallite sizes were calculated by the Scherrer formula: where is the crystallite size, is the instrument correction factor, is the Cu-Kα wavelength (0.151418 nm), is the full width at half maximum of the peaks, and is the position of the peaks. The average crystallite sizes of WO3 in S-0, S-1, S-2, S-3, and S-4 were 54.9, 62.6, 44.4, 44.3, and 50.2 nm, respectively. The calculated particle size of g-C3N4 was about 5.7 nm.

Figure 2 exhibits the SEM images of g-C3N4, pure WO3 (S-0), and g-C3N4-WO3 (S-2). The morphology of g-C3N4 was a sheet that consisted of small particles, and the particle sizes were about 200 nm. The WO3 particle sizes in WO3 and g-C3N4-WO3 were between 100 and 400 nm; the addition of g-C3N4 in the composite had no obvious influence on the particle size of WO3. The particle sizes obtained from SEM images were larger than those calculated from the Scherrer formula, which manifested that the particles observed by SEM in g-C3N4, pure WO3, and g-C3N4-WO3 were aggregates of smaller particles.

The FTIR spectra of WO3 and g-C3N4-WO3 (S-2) are shown in Figure 3. The absorption peaks at 763, 822, and 935 cm-1 in the spectra of two samples originated from stretching vibrations of O-W-O in WO3 [9, 15, 16]; the peaks at 1632 and 3436 cm-1 in the FTIR spectra of WO3 and g-C3N4-WO3 (S-2) resulted from the vibration of bended H-O-H and stretched O-H of absorbed H2O on the material surface [17]; and absorption peaks in the FTIR spectrum of g-C3N4-WO3 (S-2), other than those of WO3, at 1244, 1321, 1411, 1567, and 1632 cm-1 were ascribed to stretching vibration of C-NH-C and C=N of heterocycles [15], which proved the existence of g-C3N4 in the composite. The XPS results were discussed in the supplementary materials, which could prove the formation of g-C3N4-WO3.

3.2. Gas Sensing Characterization

The gas sensing responses of pure WO3, S-1, S-2, S-3, and S-4 to 1000 ppm concentration of acetone at different operating temperatures are shown in Figure 4. An increase in the response with an increase in the amount of g-C3N4 in the g-C3N4-WO3 composite was observed while the content of g-C3N4 was lower than 2 wt%; when the contents of g-C3N4 in the g-C3N4-WO3 composite were 3 wt% and 4 wt%, the responses decreased significantly; the optimal operating temperatures for S-1 and S-2 were all 310°C, which were better than that for pure WO3. The responses of pure WO3, S-1, S-2, S-3, and S-4 to 1000 ppm acetone at 310°C were 3.7, 16.2, 58.2, 1.0, and 1.3, respectively. It has been reported that pure g-C3N4 exhibited very little response to acetone; the response improved significantly for a particular content of g-C3N4 in a series of g-C3N4-SnO2 nanocomposites [18]. The response of α-Fe2O3/g-C3N4 to ethanol was also reported [11]; an appropriate amount of g-C3N4 in the composites was propitious to the dispersion of α-Fe2O3 in the composites and the formation of better heterojunctions; the reasons for the enhancement of gas sensing response were attributed to the larger specific surface area, better permeability, and heterojunction. A sensor based on the S-2 composite exhibited higher response to acetone compared with pure WO3. The N2 adsorption-desorption results are shown in Figure S3 of supplementary materials; the average pore size difference between S-0 and S-2 was not obvious, and the gas diffusion rates in the inner sections of S-0 and S-2 were approximate; the enhancement of gas sensing response of S-2 was probably attributable to the larger specific surface area of S-2 and heterojunction. Many literatures have reported the acetone sensing mechanism of WO3 [19, 20], and the conductance of the WO3 sensor is influenced by the changes in chemisorbed oxygen present on the surface of the gas sensing material; on exposure of a sensor to air, oxygen is adsorbed on the surface of WO3, which in turn captured electrons from the conduction band of WO3, resulting in the decrease in electron concentration in the conduction band; oxygen molecules changes into O2-(ads), O-(ads), and O2-(ads) with variation of temperature. When the sensor was placed in an acetone vapor atmosphere, acetone reacted with O2-(ads), O-(ads), and O2-(ads), releasing the electrons captured by oxygen molecules to the conduction band of WO3 and decreasing the resistance of the sensor. The reaction is as follows:

Figure 5 depicts the responses of an S-2-based sensor to 1000 ppm acetic acid, acetone, formaldehyde, ethanol, acetaldehyde, and ammonia at different temperatures. The optimal operating temperatures for formaldehyde, ethanol, and acetaldehyde were 360°C, 240°C, and 360°C, respectively; the maximum responses for acetic acid, acetone, and ammonia all appeared at 310°C. At an operating temperature of 310°C, the responses to 1000 ppm acetic acid, acetone, formaldehyde, ethanol, acetaldehyde, and ammonia were 8.9, 58.2, 6.0, 15.4, 4.3, and 6.0, respectively; the S-2 sensor showed significant gas sensing selectivity to acetone, with the response ratio of 1000 ppm acetone/1000 ppm ethanol reaching 3.8.

Figure 6 shows the responses of S-0- and S-2-based sensors to six kinds of gases, while keeping the concentration 1000 ppm, at 310°C. The responses of the S-2 sensor to acetaldehyde, ethanol, and formaldehyde were lower than those of the pure WO3 sensor. But the responses of the S-2 sensor to ammonia, acetone, and acetic acid were higher than those of the pure WO3 gas sensor; especially, the response of the S-2 sensor to acetone was 15 times that of the pure WO3 gas sensor, proving the role of g-C3N4 in improving the selectivity of the gas sensor.

The response time and recovery time were calculated using the formula defined in a previous literature [21]. The response curve transients of the sensor based on the sample S-2 composite to acetone (1000 ppm, 500 ppm, 100 ppm, 10 ppm, 1 ppm, and 0.5 ppm) at 310°C are shown in Figure 7. The responses to 1000 ppm, 500 ppm, 100 ppm, 10 ppm, 1 ppm, and 0.5 ppm acetone were 58.2, 36.6, 17.2, 3.0, 1.8, and 1.6, respectively; the detection limit of the S-2 composite-based sensor to acetone was 0.5 ppm. The response times for 1000, 500, 100, 10, 1, and 0.5 ppm acetone were 53, 24, 10, 15, 7, and 5 s, respectively, while the recovery times for 1000, 500, 100, 10, 1, and 0.5 ppm acetone were 29, 25, 6, 7, 12, and 3 s, respectively. The acetone concentration in the breath varies from 0.3 to 0.9 ppm for healthy people, but the acetone concentration exceeds 1.8 ppm for diabetic patients [22]. The S-2 composite-based sensor had a response of 1.6 to 0.5 ppm acetone which meant that it has the potential for application in diabetes detection.

The gas sensor stability is a significant parameter for a gas sensor, and the curve of gas sensing response versus time of the S-2 composite-based sensor is shown in Figure 8. The gas sensing response decreased significantly in seven days and then achieved stability between the seventh and thirtieth days. The stability of the sensor can be improved further. The g-C3N4-WO3 composite (S-2) has proven itself to be a potential candidate for application as an acetone sensor, if the stability of the sensor can be improved.

Table 1 compares the gas sensing properties of different acetone sensors. Overall, the g-C3N4-WO3 acetone gas sensor showed good sensitivity to acetone and also had significant selectivity. Besides, the detection limit is as low as 0.5 ppm which is the second lowest in the listed acetone sensor.

4. Conclusions

It can be observed that the content of g-C3N4 in g-C3N4-WO3 composites influences the response and selectivity of g-C3N4-WO3 composite-based sensors to acetone. 2 wt% g-C3N4-WO3 composite (S-2) showed the best gas sensing performances in the series of g-C3N4-WO3 composites, when operating at an optimum temperature of 310°C; the responses to 1000 ppm and 0.5 ppm acetone were 58.2 and 1.6, respectively, with the ratio of the 1000 ppm acetone to 1000 ppm ethanol reaching 3.7; the S-2 composite-based sensor was able to detect acetone at concentrations as low as 0.5 ppm. The sensor took 5 s and 3 s to respond to 0.5 ppm acetone and to recover; the g-C3N4-WO3 composite (S-2) has proven to be a potential candidate for application as an acetone sensor if the stability of the sensor can be improved.

Data Availability

The underlying data related to this manuscript is available on request.

Conflicts of Interest

The authors declare no conflict of interest.

Acknowledgments

This research was funded by the National Natural Science Foundation of China (Nos. 61671019 and 61971003).

Supplementary Materials

Figure S1: gas sensor used for characterizing gas sensing behavior of samples. Figure S2: XPS spectra of g-C3N4-WO3 (S-2): (a) full spectrum. (b) C1s. (c) N1s. (d) O1s. (e) W. Figure S3: (a) N2 adsorption-desorption isotherm of WO3. (b) Pore size distribution curve of WO3. (c) N2 adsorption-desorption isotherm of g-C3N4-WO3 (S-2). (d) Pore size distribution curve of g-C3N4-WO3 (S-2). (Supplementary Materials)