- Email: [email protected]
B2O3 gas sensor to hydrogen using different organic binder
Materials Science in Semiconductor Processing 99 (2019) 140–148
Contents lists available at ScienceDirect
Materials Science in Semiconductor Processing journal homepage: www.elsevier.com/locate/mssp
Response of TiO2/MWCNT/B2O3 gas sensor to hydrogen using different organic binder
T
Siti Amaniah Mohd Chachulia,d,∗, Mohd Nizar Hamidona, Md Shuhazlly Mamatb, Mehmet Ertugrulc, Nor Hapishah Abdullaha a
Institute of Advanced Technology, Universiti Putra Malaysia, 43400, Serdang, Selangor, Malaysia Faculty of Sciences, Universiti Putra Malaysia, 43400, Serdang, Selangor, Malaysia c Engineering Faculty, Ataturk University, 25240, Erzurum, Turkey d Fakulti Kejuruteraan Elektronik & Kejuruteraan Komputer, Universiti Teknikal Malaysia Melaka, Hang Tuah Jaya, 76100, Durian Tunggal, Melaka, Malaysia b
A R T I C LE I N FO
A B S T R A C T
Keywords: Organic binder Linseed oil Paste Gas sensor TiO2/MWCNT/B2O3
A binder influences the sensitivity, resistivity and optimal operating temperature of a gas sensor, which plays an important role in gas sensing. This work compared the sensitivity of the TiO2/MWCNT/B2O3 gas sensor to hydrogen with the addition of different organic binders, namely linseed oil and ethyl cellulose, to TiO2/ MWCNT/B2O3 paste. Both pastes were deposited on alumina substrate using the screen-printing method and annealed at 500 °C. The sensing films of gas sensor, OBL and OBE were characterized by field emission scanning electron microscopy (FESEM), Energy dispersive x-ray (EDX), X-ray diffraction (XRD) Raman Spectroscopy and Brunauer-Emmett-Teller (BET). The gas sensors were also exposed to different concentrations of hydrogen (100–1000 ppm) at various operating temperature (100 °C, 200 °C and 300 °C). The obtained results revealed that ethyl cellulose-based gas sensor achieves better sensitivity, whereas linseed oil-based gas sensor has better conductivity and recovery characteristic.
1. Introduction The metal-oxide gas sensor has gained growing interest given its capability to detect different types of gases at minimal cost preparation. There are two types of metal-oxide gas sensor, namely n-type and ptype. Some of the most commonly used n-type metal-oxide gas sensors include tin dioxide (SnO2), zinc oxide (ZnO), titanium dioxide (TiO2), tungsten oxide (WO3), gallium oxide (Ga2O3) and indium oxide (In2O3). Meanwhile, nickel oxide (NiO), copper oxide (CuO), chromium oxide (Cr2O3), cobalt oxide (Co3O4) and manganese oxide (Mn3O4) are some of the most commonly used p-type metal-oxide gas sensors. The p-type gas sensors have attracted little attention compared to the n-type gas sensors [1]. Meanwhile, TiO2, among most metal-oxides, has shown good response to hydrogen [2–6]. It was revealed that TiO2 can be detected with hydrogen of as low as 1 ppm at room temperature [3]. Generally, TiO2 exists in anatase phase, rutile phase and brookite phase. The gas sensors often involve anatase phase and rutile phase [6,7]. Besides that, carbon-based gas sensor such as multi-walled carbon nanotube (MWCNT) and graphene has also recently gained growing interest given its unique electrical, optical and mechanical properties [8]. For instance, MWCNT reportedly identifies hydrogen [9,10],
∗
ammonia and methane [11] whereas graphene detects various types of gases such as carbon dioxide [12] and nitrogen dioxide [13]. Among the carbon-based gas sensors, MWCNT has shown good response to hydrogen at room temperature [9,10]. The addition of MWCNT in TiO2 was revealed to reduce the high resistivity of titanium and improve the response to hydrogen [14]. Therefore, this work added MWCNT in TiO2 for enhanced properties of gas sensor to hydrogen. The binder in thick film gas sensor enhances the sensor response and reduces the resistivity of a gas sensor [15]. Besides that, the addition of binder in a paste also improves the cohesion of particles after the firing process and enhances the adhesion of the sensing film to the substrate [16]. Most binders in thick film gas sensor are based on tetraethyl orthosilicate [15,17] and ethyl cellulose [16,18]. The addition of organic binder, specifically linseed oil in TiO2 paste for gas sensor was previously proposed [19], which showed good response to the hydrogen at elevated temperature. Other than that, linseed oil has been used in diverse fields; for examples, as diesel engine [20], to synthesize carbon nanofiber [21], as a varnish to mitigate wood degradation [22] and as anti-corrosive coating [23]. In view of the above, this work fabricated TiO2/MWCNT/B2O3 gas sensor with different binder and exposed to various concentrations of
Corresponding author. Institute of Advanced Technology, Universiti Putra Malaysia, 43400, Serdang, Selangor, Malaysia E-mail address: [email protected] (S.A. Mohd Chachuli).
https://doi.org/10.1016/j.mssp.2019.04.009 Received 28 December 2018; Received in revised form 29 March 2019; Accepted 5 April 2019 1369-8001/ © 2019 Elsevier Ltd. All rights reserved.
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
observe the structure of TiO2 and verify the elements of MWCNT after the heat treatment. The element composition and mapping of the elements were examined by the Energy Dispersive X-ray (EDX) inside the FESEM. The X-ray Diffraction (XRD) (Brand: Philips, Model: PW 3040/ 60 MPD X'pert High Pro Panalytical) studies were then carried out for powder and thick film in 5 min and 30 min, respectively, over a 2θ range between 20° and 80°. Raman spectroscopy (Brand: WITec, Model: Alpha 300 R) was carried out using a laser excitation line at 532 nm for powder and thick film. The absorption of both thick films was measured by UV–Visible spectrophotometer (Brand: Perkin Elmer, Model: Lambda 35) within the range of 280 nm–800 nm to calculate the optical band gap using Tauc plot. Surface analysis for both sensing material of gas sensor, OBL and OBE were characterized by Brunauer-EmmettTeller (BET) (Brand: Micromeritics, Model: TriStar II Plus) using nitrogen adsorption-desorption.
hydrogen at different operating temperatures. Different binders were applied to the TiO2/MWCNT/B2O3 paste to assess the properties of the gas sensor in terms of resistivity, band gap, sensitivity and optimal operating temperature. 2. Experimental details 2.1. Preparation of organic binder and TiO2/MWCNT/B2O3 pastes This work employed two different binders to prepare TiO2/ MWCNT/B2O3 paste for gas sensing purpose. One of binders used in this work was the conventional binder, which was based on ethyl cellulose and α-terpineol. The conventional binder was prepared according to the method presented in prior studies [24,25]. Besides that, the proposed organic binder was prepared by mixing the linseed oil, m-xylene and αterpineol using the magnetic stirrer to obtain homogeneous binder [19]. The proposed binder is labelled as OB-1 while the conventional binder is labelled as OB-2. The linseed oil, which is extracted from flax seeds oil, is eco-friendly and organic. Moreover, a homogeneous binder can be easily attained when linseed oil is used compared to the ethyl cellulose since the linseed oil is produced in liquid form, which allows better mixing. TiO2 (Aeroxide® P25) and MWCNT (OD = 10–20 nm, length = 10–30 μm) were obtained from Sigma-Aldrich and MaterialsA2Z, respectively. To prepare the sensing film of gas sensor, 95 wt % of TiO2 was mixed with 5 wt % of MWCNT using m-xylene as the medium in an ultrasonic bath for 90 min. Next, it was dried in an oven and ground in a mortar. Glass powder, specifically boron oxide (B2O3) was added to the paste to ensure good adhesion of sensing film on the alumina substrate. Following that, 10 wt % of B2O3 was mixed with 90 wt % of TiO2/MWCNT using m-xylene as the medium in an ultrasonic bath for 90 min. Then, it was dried in an oven and the produced powder was ground in a mortar. The addition of glass powder not affect the nanoparticles structure of TiO2 [19]. Both pastes were then prepared by mixing 30 wt % of TiO2/MWCNT/B2O3 and 70 wt % of organic binder using the magnetic stirrer until a homogeneous and viscous paste was obtained. A similar ratio was applied for both pastes to compare the performance of gas sensor to hydrogen. Accordingly, paste based on OB-1 was labelled as OBL and paste based on OB-2 was labelled as OBE.
2.4. Measurement of gas sensor The schematic diagram of experimental setup of gas sensor in a gas chamber can be referred in prior study [19]. 500 sccm of nitrogen was initially flowed in 600 s, then, followed by the flow of various concentrations of hydrogen between 100 ppm and 1000 ppm. Each cycle of hydrogen was subsequently flowed in 600 s. The nitrogen as carrier gas, flowed after the exposure to hydrogen in each cycle. Both gas sensors were tested at different operating temperatures: 100 °C, 200 °C and 300 °C and the measurement of gas sensor was conducted after achieving a stabilizing current at elevated operating temperature. 3. Results and discussion 3.1. Characterizations of binders by Fourier-transform infrared spectroscopy (FTIR) Fig. 1 presents the FTIR spectra of OB-1 and OB-2 in the range of 4000–400 cm−1. The apparent difference between both binders where OB-1 demonstrated a peak in the range of 2000–1500 cm−1 whereas OB-2 demonstrated the presence of OeH bond in the range of 4000–2500 cm−1 may have occurred due to the different mixture in both binders. Specifically, OB-2 did not contain m-xylene and as well as the respective functional groups of ethyl cellulose and linseed oil were included in the binders. In the case of OB-2, the bands centre of OeH bond was at 3500 cm−1 due to the presence of ethyl cellulose which belongs to the hydroxyl group. As for the case of OB-1, the peaks was at 1750 cm−1, which display the characteristic of C]O. The observed peak that represents the CeH stretching band in the region of 2800 cm−1 to 2900 cm−1 for both samples suggests the presence of
2.2. Fabrication of TiO2/MWCNT/B2O3 gas sensor To fabricate the gas sensor, an interdigitated electrode (IDE) based on silver conductive paste (Sigma-Aldrich) was deposited using screenprinting on alumina substrate as a first layer of gas sensor and annealed at 120 °C for 15 min. Following that, TiO2/MWCNT/B2O3 paste was printed as the second layer of the gas sensor and it was annealed in the furnace at 500 °C for 30 min. Accordingly, 500 °C was chosen as the annealing temperature because the paste based on linseed oil (OB-1) was completely evaporated at this temperature [19]. Both annealing processes were carried out using air as the carrier gas. The size of IDE used in this work was approximately 9.5 mm × 4.0 mm while the size of the sensing film was approximately 4.2 mm × 4.2 mm. The thickness of IDE and sensing film were approximately between 8 μm and 12 μm and the size of alumina substrate was about 20 mm × 15 mm. 2.3. Materials characterization The characterization of binder was defined by Fourier-Transform Infrared Spectroscopy (FTIR), (Brand: Thermo Nicolet, Model: Nicolet 6700) to identify the functional group for both binders (OB-1 and OB2). Meanwhile, the surface morphology of both thick films after annealed under air at 500 °C was characterized by the Field Emission Scanning Electron Microscopy (FESEM), (Brand: Thermo Fisher Scientific, Model: Nova Nanosem 230) at 100 k magnification to
Fig. 1. FTIR spectra of OB-1 and OB-2. 141
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 2. FESEM image of TiO2/MWCNT/B2O3 thick film annealed at 500 °C for (a) OBL and (b) OBE.
plan (002) of graphite hexagonal structure [27]. After sonication process of TiO2 with MWCNT for 90 min, carbon peak was not detected in XRD spectra of TiO2/MWCNT. However, after addition of B2O3 and sonication process in 90 min, it was found that the carbon peak was detected and shifted from 2θ = 25.95° to 2θ = 28.17° in TiO2/ MWCNT/B2O3, in which better dispersion of MWCNT was achieved after 3 h of sonication process as seen in Fig. 5. However, there is no carbon peak was detected after the annealing treatment at 500 °C for OBL and OBE (Fig. 6). This may due to the combination of carbon and oxygen during the annealing process under air in the furnace. In another study, Contreras et al. oxidized MWCNT under air at different temperature (between 270 – 600 °C) and found reduced weight loss of MWCNT when the temperature increased [28]. It was observed that the crystallinity of anatase phase at 2θ = 25.44° and rutile phase at 2θ = 27.61° increased after MWCNT was added to TiO2 and subsequently increased after B2O3 was added (Fig. 5). These results revealed that an extended sonication process enhances the crystallinity of anatase and rutile phase. After annealed treatment at 500 °C, the crystallinity of anatase phase at 2θ = 25.39° and rutile phase at 2θ = 27.45° for OBE was slightly higher than OBL (Fig. 6). Additionally, B2O3 peak was only detected after the annealing treatment at 500 °C, which at similar phase of rutile phase at 2θ = 36.25°. This also suggests that the crystallinity of B2O3 was achieved after the melting process was occurred. Raman characterization was carried out for original powder (TiO2 and MWCNT) and thick film (OBL and OBE). Raman spectrum for the samples in the range of 0–3500 cm−1 is shown in Fig. 7. As seen in Raman spectra, anatase peak was identified at 140 cm−1, 404 cm−1, 505 cm−1 and 638 cm−1 in TiO2 (P25). Whereas, MWCNT has three typical bands, D-band, G-band and 2D-band that correspond to 1343 cm−1, 1589 cm−1 and 2688 cm−1, respectively. The pattern of MWCNT also was observed in Ref. [29]. The peak of D-band was seen
organic compounds in both binders. Accordingly, the observed peaks at 1500 cm−1, 1150 cm−1, 900 cm−1 and 800 cm−1 in both binders were attributed to the CeH, CeO, C]C and CeH bond, respectively. 3.2. Characterization of TiO2/MWCNT/B2O3 thick films by FESEM, EDX, XRD and Raman Spectroscopy Fig. 2 presents the surface morphology of TiO2/MWCNT/B2O3 sensing film for OBL and OBE after undergone annealed treatment at 500 °C. The nanoparticles structure of TiO2 was clearly seen in both samples at 100 k magnification. Besides that, the structure of MWCNT was also apparent for both samples. However, sample based on OBE demonstrated higher quantity of MWCNT compared to the sample based on OBL. Average sizes of TiO2 nanoparticles ranged between 40 nm and 70 nm for both samples. As shown in Fig. 3, the EDX results revealed the presence of titanium, oxygen, carbon and aluminium (substrate). Both thick films generated high peak of titanium which indicates the crystallization of TiO2 at 500 °C. Weight of titanium for OBL and OBE was approximately 42.73% and 40.85%, respectively. In addition, the EDX results revealed that, the weight of carbon in OBL was approximately 4.29%, which was lower than the weight of carbon in OBE (4.41%). This implies the more carbon loss in OBL compared to the OBE during annealing at 500 °C, which was verified by the mapping in EDX (Fig. 4). The EDX analysis demonstrates the ability of OB-2 to hold more of carbon elements rather than OB-1. Fig. 5 displays XRD spectra of MWCNT, TiO2, TiO2/MWCNT and TiO2/MWCNT/B2O3 without the heat treatment, while Fig. 6 displays XRD spectra of TiO2/MWCNT/B2O3 after annealed treatment at 500 °C. Both figures showed high crystallinity of anatase phases and rutile phases in TiO2, TiO2/MWCNT and TiO2/MWCNT/B2O3, which are well consistent with TiO2 nanoparticles [26]. As for the case of MWCNT, high carbon peak was detected at 2θ = 25.95°, which was related to the
Fig. 3. EDX results of TiO2/MWCNT/B2O3 thick film on alumina substrate for (a) OBL and (b) OBE. 142
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 4. Mapping of MWCNT in sensing film on alumina substrate by EDX for (a) OBL and (b) OBE.
Fig. 7. Raman spectrum of TiO2, MWCNT, OBL and OBE.
Fig. 5. XRD spectra of MWCNT, TiO2, TiO2/MWCNT and TiO2/MWCNT/B2O3 without heat treatment.
TiO2 and MWCNT in the sensing film and also verified the crystallinity of the sensing film for gas sensor. 3.3. Surface analysis of TiO2/MWCNT/B2O3 by Brunauer-Emmett-Teller (BET) The surface area and pore size distribution of sensing material, OBL and OBE were further characterized by nitrogen adsorption-desorption isotherm using BET analysis. The sample was prepared by dropped the paste (OBL and OBE) on the alumina substrate and annealed under air at 500 °C for 30 min. Table 1 lists average pore size and BET surface area for both sensing material. It indicated that average pore size of OBE smaller than OBL, while surface area of OBE was larger than OBL. This analysis also confirmed distinctively the porosities of both sensing films. Higher surface area maybe contributed to the larger response of sensing film to the target gas due to the ability to facilitate the diffusion of the target gas [32,33]. Fig. 8 shows nitrogen gas adsorption-desorption isotherms shape of IV with a H3 hysteresis loop of the sensing material, OBL and OBE according to the IUPAC classification [34]. This hysteresis loop is usually attributed to the thermodynamic or/and network effects and it also indicative of the pores in the sensing film of gas sensor [35]. It can
Fig. 6. XRD spectra of sensing film based on OBL and OBE after annealed at 500 °C.
occurred at ∼1340 cm−1 and G-band at ∼1580 cm−1 [30,31]. Degree of graphitization can be calculated using ID/IG ratio, where the value is 0.85 for MWCNT, OBL and OBE. This ratio indicated higher value of graphitization based on ID/IG value is less than 1 [29]. The ID/IG ratio for OBL and OBE showed similar value with MWCNT, which indicates no defects occurred during synthesis of TiO2/MWCNT/B2O3. Peaks in the Raman spectrum for OBL and OBE also confirmed the presence of
Table 1 Average pore size and BET surface area of OBL and OBE.
143
Sensing material
Average pore size (nm)
BET surface area (m2/g)
OBL OBE
5.6473 5.4275
52.8219 57.3509
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 8. Nitrogen gas adsorption-desorption isotherms and pore size distribution of (a) OBL and (b) OBE.
Fig. 9. Optical band gap of (a) TiO2 (P25); (b) TiO2/B2O3 thick film; (c) OBL and; (d) OBE.
where α is absorption coefficient, Eg is absorption band gap, A is a constant (depending on the transition probability), hv is photon energy and n depends on the nature of transition [37]. Eg was obtained by draw a tangent line on graph of (αhv)n versus hv. For indirect band gap, n equals to ½ and as for direct band gap, n equals to 2 [37]. Fig. 9 shows the band gap values of TiO2 (P25), TiO2/B2O3 thick film and TiO2/MWCNT/B2O3 thick film for both samples of OBL and OBE. Band gap of TiO2 (P25) was measured in liquid form by dispersion of TiO2 (P25) powders in ethanol. It was found that the calculated band gap of TiO2 (P25) was based on direct band gap which approximately 3.25 eV. It has been reported in the literature the band gap of TiO2 (P25) can be existed in direct band gap or indirect band gap [38–40]. Wang et al. also calculated TiO2 (P25) powders at different calcination
be seen that, the pore size distribution of OBL is slightly higher than OBE (Fig. 8) in which similar result as tabulated in Table 1. Changes in hysteresis loop, pore size distribution and surface area of the sensing material can affect the performance of sensing film in a gas sensor [35]. 3.4. Optical band gap of TiO2/MWCNT/B2O3 by UV–Vis spectroscopy Absorption data of TiO2/MWCNT/B2O3 thick film from the UV–Vis Spectroscopy is used to calculate the optical band gap (Eg) of TiO2/ MWCNT/B2O3 according to the proposed model by Tauc [36]. The following equation was used to generate the graph of (αhv)n versus hv,
αhv n = A (hv − Eg ) 144
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 10. Resistance of TiO2/B2O3 and TiO2/MWCNT/B2O3 (OBL and OBE) gas sensor at different operating temperatures (a) T = 100 °C; (b) T = 200 °C; and (c) T = 300 °C.
Fig. 11. Response of TiO2/MWCNT/B2O3 (OBL and OBE) gas sensor to hydrogen with different binder at different operating temperature (a) T = 100 °C; (b) T = 200 °C; and (c) T = 300 °C.
revealed at this temperature. It should be noted that, similar absorption produces similar optical band gap. The similar values of band gap for both samples (OBL and OBE) also indicated that different binder does not affect the semiconductor properties of TiO2.
treatment based on direct band gap [41]. Whereas, TiO2/B2O3 thick film band gap in previous work [19] was calculated based on indirect band gap, which the value was approximately 4.1 eV (Fig. 9(b)). Prior studies also associated TiO2 (P25) thick film to the indirect band gap [42,43]. When MWCNT was added into TiO2, band gaps of TiO2/ MWCNT/B2O3 thick film (OBL and OBE) were also calculated based on indirect band gap measurement (Fig. 9(c) and (d)). With the addition of MWCNT to TiO2, the band gap was shown to reduce to 3.72 eV. With reduction of band gap, the response time of a gas sensor can become faster to detect the target gas and increase the sensitivity of gas sensor. Both thick films (OBL and OBE) produced similar band gap of approximately 3.72 eV. It may because of the mass loss of binder was completely loss at 500 °C, therefore similar absorption of UV light was
3.5. Resistance of TiO2/MWCNT/B2O3 gas sensor at different operating temperature The resistance of TiO2/B2O3 gas sensor and TiO2/MWCNT/B2O3 (OBL and OBE) gas sensor at different operating temperatures under nitrogen flowed are displayed in Fig. 10. The addition of MWCNT reduced the resistance of TiO2 gas sensor due to the higher carrier mobility of MWCNT [8]. The percentage of reduction was approximately 145
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 12. Sensitivity of TiO2/MWCNT/B2O3 (OBL and OBE) gas sensor to hydrogen with different binder at different operating temperatures (a) T = 100 °C; (b) T = 200 °C; and (c) T = 300 °C.
and OBE) responded well to 1000 ppm and 500 ppm of H2 at all operating temperatures (100–300 °C). At the operating temperature of 100 °C, both gas sensors (OBL and OBE) were unable to sense 100 ppm of H2. As the operating temperature increased to 200 °C and 300 °C, both gas sensors (OBL and OBE) showed good response to 100 ppm of H2 due to the slower kinetics of electron movement at lower operating temperature and higher adsorption at higher operating temperature [45]. The sensitivity of gas sensor of various concentrations of hydrogen at different operating temperatures is displayed in Fig. 12. As compared to the previous work [19], the sensitivity of gas sensor in this work improved with the addition of MWCNT in TiO2. At the operating temperature of 100 °C, the sensitivity for both gas sensors did not demonstrate significant difference at 500 ppm and 1000 ppm of H2. On the other hand, the OBL-based gas sensor demonstrated significantly different sensitivity from the sensitivity of OBE-based gas sensor at operating temperature of 200 °C for 500 ppm and 1000 ppm of H2. In particular, OBE-based gas sensor produced highest sensitivity compared to the OBL-based gas sensor at 500 ppm and 1000 ppm of H2. However, the sensitivity of both gas sensors to 500 ppm and 1000 ppm of H2 did not show significant difference at the operating temperature of 300 °C. These results indicated that the optimal operating temperature for both gas sensor in this work is 200 °C. Besides that, the OBL-based gas sensor, compared to OBE-based gas sensor showed better recovery characteristic at 200 °C and 300 °C where its sensitivity value was able to achieve the original value which is one. Sensing mechanism of a gas sensor can be influenced by surfacecontrolled mode, grain-controlled mode and neck-controlled mode [46]. For OBL-based gas sensor and OBE-based gas sensor, surfacecontrolled mode are similar due to the similar compact layer structure and thickness of gas sensor produced by screen-printing technique.
99.86%, 95.50% and 91.79% at operating temperature of 100 °C, 200 °C and 300 °C, respectively. Overall, the resistance of the gas sensor decreased as temperature increased. The resistance of OBL-based gas sensor and the resistance of OBE-based gas sensor demonstrated significant difference at operating temperature of 100 °C due to the presence of OeH bond in the OBE-based gas sensor. In other words, the occurrence of higher oxidation results in higher resistance. Moreover, the weight of oxygen in the both samples varied, where OBL and OBE contributed 51.92% and 54.45% (including Al2O3), respectively. However, the resistance of both gas sensors did not exhibit significant difference at operating temperature of 200 °C and 300 °C, which may due to the decomposition of silver oxide (electrode) at operating temperature of above 200 °C [44], thus resulting in similar resistance. 3.6. Response of TiO2/MWCNT/B2O3 gas sensor to hydrogen at different operating temperature The response of TiO2/MWCNT/B2O3 gas sensor to hydrogen with different binder at different operating temperatures is shown in Fig. 11. From the response, the gas sensor behaved as a p-type gas sensor based on the decreased current when it was exposed to hydrogen and the increased current when it was exposed to nitrogen. This behaviour also reported in Ref. [14]. Besides that, the initial current of OBL-based gas sensor was slightly higher than the observed current OBE-based gas sensor at different operating temperatures. These obtained results indicated that the linseed oil-based gas sensor produces better conductivity and low resistivity compared to the ethyl cellulose-based gas sensor based. Higher conductivity of OBL-based gas sensor can be caused by its ability to hold more titanium elements after annealed treatment at 500 °C compared to the OBE-based gas sensor as referred to the EDX results (Fig. 3). In term of responses, both gas sensors (OBL 146
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Besides, similar band gap produced by OBL and OBE indicates the surface-controlled mode are similar for both gas sensors. Whereas, grain-controlled mode and neck-controlled mode contribute to the electroconductibility and performance of gas sensor [46]. As seen in Fig. 12, OBE-based gas sensor exhibits higher sensitivity at optimal operating temperature compared to the OBL-based gas sensor, this might be due to the higher surface area of OBE as tabulated in Table 1, in which ease diffusion of hydrogen to the sensing film. Higher response for larger surface area also reported in Refs. [32,33]. This results also revealed that larger grain size can improve response of a gas sensor to the target gas. It also can be seen that OBL-based gas sensor has better recovery characteristic in which ability to return to its original resistance value after exposed to the hydrogen compared to the OBEbased gas sensor. This phenomenon might because of higher pore size in OBL, thus ease for the target gas to dismiss from the surface of a gas sensor.
based hydrogen gas sensor, Sensors Actuators, A Phys. 201 (2013) 321–327. [10] S. Dhall, N. Jaggi, Highly dispersed platinum sputtered multiwall carbon nanotubes based hydrogen gas sensor at room temperature, Sensors Actuators, A Phys. 224 (2015) 50–56. [11] N. Kumar, P. Sahatiya, P. Dubey, Fabrication of CNT based gas sensor using interdigitated gold electrodes, Procedia Materials Science 6 Icmpc, 2014, pp. 1976–1980. [12] J. Wu, S. Feng, Z. Li, K. Tao, J. Chu, J. Miao, L.K. Norford, Boosted sensitivity of graphene gas sensor via nanoporous thin film structures, Sensor. Actuator. B Chem. 255 (2018) 1805–1813. [13] Y. Seekaew, D. Phokharatkul, A. Wisitsoraat, Applied Surface Science Highly sensitive and selective room-temperature NO2 gas sensor based on bilayer transferred chemical vapor deposited graphene, Appl. Surf. Sci. 404 (2017) 357–363. [14] L. De Luca, A. Donato, S. Santangelo, G. Faggio, G. Messina, N. Donato, G. Neri, Hydrogen sensing characteristics of Pt/TiO2/MWCNTs composites, Int. J. Hydrogen Energy 37 (2) (2012) 1842–1851. [15] F. Engzneenng, Effects of tetraethyl orthosilicate binder on the characteristics of an SnO2, ceramic-type semiconductor gas sensor, Sensor. Actuator. 9 (1986) 133–145. [16] V. Guidi, M.A. Butturi, M.C. Carotta, B. Cavicchi, M. Ferroni, C. Malagù, G. Martinelli, D. Vincenzi, M. Sacerdoti, M. Zen, Gas sensing through thick film technology, Sensor. Actuator. B Chem. 84 (1) (2002) 72–77. [17] S.J. kim, G. Lee, The effect on the gas selectivity of gas sensors by binder in SWNTs solution, IEEE Sens. J. 10 (1) (2010) 1–5. [18] S. Lee, G.G. Lee, J. Kim, S.J.L. Kang, A novel process for fabrication of SnO2-based thick film gas sensors, Sensor. Actuator. B Chem. 123 (1) (2007) 331–335. [19] S. Amaniah, M. Chachuli, M.N. Hamidon, M.S. Mamat, M. Ertugrul, N.H. Abdullah, A hydrogen gas sensor based on TiO2 nanoparticles on alumina substrate, Sensors 18 (2483) (2018) 1–13. [20] H. Hazar, H. Sevinc, Investigation of the effects of pre-heated linseed oil on performance and exhaust emission at a coated diesel engine, Renew. Energy 130 (2018) 961–967. [21] D.E. Kshirsagar, V. Puri, H. Dubey, M. Sharon, Giga hertz frequency absorber carbon nano fibers synthesized using linseed oil, Mater. Today Commun. 13 (August) (2017) 23–25. [22] J.J. Łucejko, J. La Nasa, C.M.A. McQueen, S. Braovac, M.P. Colombini, F. Modugno, Protective effect of linseed oil varnish on archaeological wood treated with alum, Microchem. J. 139 (2018) 50–61. [23] H. Wang, Q. Zhou, Evaluation and failure analysis of linseed oil encapsulated selfhealing anticorrosive coating, Prog. Org. Coating 118 (November 2017) (2018) 108–115. [24] M.H. Seo, M. Yuasa, T. Kida, J.S. Huh, N. Yamazoe, K. Shimanoe, Detection of organic gases using TiO2 nanotube-based gas sensors, Procedia Chem. 1 (1) (2009) 192–195. [25] M. Zhang, G. Jiang, Gas sensing properties of Co3O4-loaded SnO2 to ethanol and acetone, Chin. J. Chem. Phys. 20 (3) (2007) 315–318. [26] R. Atchudan, T. Nesakumar, J. Immanuel, S. Perumal, D. Karthikeyan, Y. Rok, Effective photocatalytic degradation of anthropogenic dyes using graphene oxide grafting titanium dioxide nanoparticles under UV-light irradiation, J. Photochem. Photobiol. A Chem. 333 (2017) 92–104. [27] M.B. Askari, Z. Tavakoli Banizi, M. Seifi, S. Bagheri Dehaghi, P. Veisi, Synthesis of TiO2 nanoparticles and decorated multi-wall carbon nanotube (MWCNT) with anatase TiO2 nanoparticles and study of optical properties and structural characterization of TiO2/MWCNT nanocomposite, Opt. - Int. J. Light Electron Opt. 149 (2017) 447–454. [28] C. Wongchoosuk, A. Wisitsoraat, D. Phokharatkul, A. Tuantranont, T. Kerdcharoen, Multi-walled carbon nanotube-doped tungsten oxide thin films for hydrogen gas sensing, Sensors 10 (8) (2010) 7705–7715. [29] R. Atchudan, B.G. Cha, N. Lone, J. Kim, J. Joo, Synthesis of high-quality carbon nanotubes by using monodisperse spherical mesoporous silica encapsulating iron oxide nanoparticles, Kor. J. Chem. Eng. 36 (1) (2019) 157–165. [30] N. Muthuchamy, R. Atchudan, T. Nesakumar, J. Immanuel, High-performance glucose biosensor based on green synthesized zinc oxide nanoparticle embedded nitrogen-doped carbon sheet, J. Electroanal. Chem. 816 (2018) 195–204 March. [31] R. Atchudan, T. Nesakumar, J. Immanuel, S. Perumal, R. Vinodh, Y. Rok, In-situ green synthesis of nitrogen-doped carbon dots for bioimaging and TiO2 nanoparticles @ nitrogen-doped carbon composite for photocatalytic degradation of organic pollutants, J. Alloy. Comp. 766 (2018) 12–24. [32] Q. Rong, Z. By, H. Jc, Z. Zq, J. Zhang, L. Qj, Methanol sensing performance of gas sensor based on Ag-LaFeO3 modified with graphene, J. Mol. Sci. 2 (1:2) (2018) 1–5. [33] C. Li, M. Lv, J. Zuo, X. Huang, SnO2 Highly Sensitive CO Gas Sensor Based on QuasiMolecular-Imprinting Mechanism Design, (2015), pp. 3789–3800. [34] M. Thommes, K. Kaneko, A. V Neimark, J.P. Olivier, F. Rodriguez-reinoso, J. Rouquerol, K.S.W. Sing, Physisorption of gases , with special reference to the evaluation of surface area and pore size distribution (IUPAC Technical Report), Pure Appl. Chem. 87 (2015) 1051–1069. [35] W. Tan, X. Ruan, Q. Yu, Z. Yu, X. Huang, Fabrication of a SnO2-Based Acetone Gas Sensor Enhanced by Molecular Imprinting, (2015), pp. 352–364. [36] J. Tauc, R. Grigorovici, A. Vancu, Optical properties and electronic structure of amorphous germanium, 15 (2) (1966) 627–637. [37] W. Saidi, N. Hfaidh, M. Rasheed, M. Girtan, A. Megriche, M. El Maaoui, Effect of B2O3 addition on optical and structural properties of TiO2 as a new blocking layer for multiple dye sensitive solar cell application (DSSC), RSC Adv. 00 (January) (2013) 1–3. [38] M. Scarsella, M.P. Bracciale, B. De Caprariis, P. De Filippis, A. Petrullo, L. Pronti, M.L. Santarelli, Improved photocatalytic properties of doped titanium-based nanometric oxides, Chem. Eng. Trans. 60 (2017) 133–138.
4. Conclusions The addition of MWCNT in TiO2 improves the sensitivity and reduces the band gap, resistance, and the optimal operating temperature of TiO2 gas sensor. The TiO2/MWCNT/B2O3 gas sensor with different binders to various concentrations of hydrogen also demonstrates good response at different operating temperatures. In terms of sensitivity, the ethyl cellulose-based gas sensor exhibits better response to hydrogen. However, linseed oil-based gas sensor has better conductivity and better recovery characteristic after its exposure to hydrogen. Adding to that, the optimal operating temperature for TiO2/MWCNT/B2O3 gas sensor for OBL and OBE in this work is 200 °C given the significant difference in the observed sensitivity at 100 ppm, 500 ppm and 1000 ppm of hydrogen. This work also reveals that linseed oil–based gas sensor can be developed as an alternative of ethyl cellulose-based gas sensor in which it offers low cost fabrication, good response to hydrogen and better recovery characteristics of gas sensor. Acknowledgment The author would like to thank the members of the Physics Laboratory, Science Faculty, Ataturk University, Turkey for the facilities provided to conduct gas sensor measurement. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.mssp.2019.04.009. References [1] Z. Wen, L. Tian-mo, Gas-sensing properties of SnO2-TiO2-based sensor for volatile organic compound gas and its sensing mechanism, Phys. B Condens. Matter 405 (5) (2010) 1345–1348. [2] X. Peng, Z. Wang, P. Huang, X. Chen, X. Fu, W. Dai, Comparative study of two different TiO2 film sensors on response to H2 under UV light and room temperature, Sensors (Switzerland) 16 (8) (2016). [3] X. Xia, W. Wu, Z. Wang, Y. Bao, Z. Huang, Y. Gao, A hydrogen sensor based on orientation aligned TiO2 thin films with low concentration detecting limit and short response time, Sensor. Actuator. B Chem. 234 (2016) 192–200. [4] A. Haidry, P. Schlosser, P. Durina, M. Mikula, M. Tomasek, T. Plecenik, T. Roch, A. Pidik, M. Stefecka, J. Noskovic, M. Zahoran, P. Kus, A. Plecenik, Hydrogen gas sensors based on nanocrystalline TiO2 thin films, Open Phys. 9 (5) (2011). [5] O. Krško, T. Plecenik, T. Roch, B. Grančič, L. Satrapinskyy, M. Truchlý, P. Ďurina, M. Gregor, P. Kúš, A. Plecenik, Flexible highly sensitive hydrogen gas sensor based on a TiO2 thin film on polyimide foil, Sensor. Actuator. B Chem. 240 (2017) 1058–1065. [6] E. Şennik, Z. Çolak, N. Kilinç, Z.Z. Öztürk, Synthesis of highly-ordered TiO2 nanotubes for a hydrogen sensor, Int. J. Hydrogen Energy 35 (9) (2010) 4420–4427. [7] K. Zakrzewska, M. Radecka, “TiO2-Based nanomaterials for gas sensing—Influence of anatase and rutile contributions, Nanoscale Res. Lett. 12 (1) (2017). [8] E. Llobet, Gas sensors using carbon nanomaterials: a review, Sensor. Actuator. B Chem. 179 (2013) 32–45. [9] S. Dhall, N. Jaggi, R. Nathawat, Functionalized multiwalled carbon nanotubes
147
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
(1) (2017) 243. [43] C.-H. Wei, C.-M. Chang, Polycrystalline TiO2 thin films with different thicknesses deposited on unheated substrates using RF magnetron sputtering, Mater. Trans. 52 (3) (2011) 554–559. [44] W.E. Garner, L.W. Reeves, The thermal decomposition of silver oxide, (1953), pp. 254–260 254. [45] C. Wang, L. Yin, L. Zhang, D. Xiang, R. Gao, Metal oxide gas sensors: sensitivity and influencing factors, Sensors 10 (3) (2010) 2088–2106. [46] Y. Wang, T. Wu, Y. Zhou, C. Meng, W. Zhu, L. Liu, TiO2-based nanoheterostructures for promoting gas sensitivity performance: Designs, developments, and prospects, Sensors 17 (9) (2017) 1–35.
[39] S. Valencia, J.M. Marín, G. Restrepo, Study of the bandgap of synthesized titanium dioxide nanoparticules using the sol-gel method and a hydrothermal treatment, Open Mater. Sci. J. 4 (2010) 9–14. [40] P. Malliga, J. Pandiarajan, N. Prithivikumaran, K. Neyvasagam, Effect of Film Thickness on Structural and Optical Properties of TiO2 Thin Films, (2013), pp. 488–491. [41] G. Wang, L. Xu, J. Zhang, T. Yin, and D. Han, “Enhanced photocatalytic activity of TiO2 powders (P25) via calcination treatment,” Int. J. Photoenergy, vol. 2012, 2012. [42] Y.-J. Shi, R.-J. Zhang, H. Zheng, D.-H. Li, W. Wei, X. Chen, Y. Sun, Y.-F. Wei, H.L. Lu, N. Dai, L.-Y. Chen, Optical constants and band gap evolution with phase transition in sub-20-nm-thick TiO2 films prepared by ALD, Nanoscale Res. Lett. 12
148
Contents lists available at ScienceDirect
Materials Science in Semiconductor Processing journal homepage: www.elsevier.com/locate/mssp
Response of TiO2/MWCNT/B2O3 gas sensor to hydrogen using different organic binder
T
Siti Amaniah Mohd Chachulia,d,∗, Mohd Nizar Hamidona, Md Shuhazlly Mamatb, Mehmet Ertugrulc, Nor Hapishah Abdullaha a
Institute of Advanced Technology, Universiti Putra Malaysia, 43400, Serdang, Selangor, Malaysia Faculty of Sciences, Universiti Putra Malaysia, 43400, Serdang, Selangor, Malaysia c Engineering Faculty, Ataturk University, 25240, Erzurum, Turkey d Fakulti Kejuruteraan Elektronik & Kejuruteraan Komputer, Universiti Teknikal Malaysia Melaka, Hang Tuah Jaya, 76100, Durian Tunggal, Melaka, Malaysia b
A R T I C LE I N FO
A B S T R A C T
Keywords: Organic binder Linseed oil Paste Gas sensor TiO2/MWCNT/B2O3
A binder influences the sensitivity, resistivity and optimal operating temperature of a gas sensor, which plays an important role in gas sensing. This work compared the sensitivity of the TiO2/MWCNT/B2O3 gas sensor to hydrogen with the addition of different organic binders, namely linseed oil and ethyl cellulose, to TiO2/ MWCNT/B2O3 paste. Both pastes were deposited on alumina substrate using the screen-printing method and annealed at 500 °C. The sensing films of gas sensor, OBL and OBE were characterized by field emission scanning electron microscopy (FESEM), Energy dispersive x-ray (EDX), X-ray diffraction (XRD) Raman Spectroscopy and Brunauer-Emmett-Teller (BET). The gas sensors were also exposed to different concentrations of hydrogen (100–1000 ppm) at various operating temperature (100 °C, 200 °C and 300 °C). The obtained results revealed that ethyl cellulose-based gas sensor achieves better sensitivity, whereas linseed oil-based gas sensor has better conductivity and recovery characteristic.
1. Introduction The metal-oxide gas sensor has gained growing interest given its capability to detect different types of gases at minimal cost preparation. There are two types of metal-oxide gas sensor, namely n-type and ptype. Some of the most commonly used n-type metal-oxide gas sensors include tin dioxide (SnO2), zinc oxide (ZnO), titanium dioxide (TiO2), tungsten oxide (WO3), gallium oxide (Ga2O3) and indium oxide (In2O3). Meanwhile, nickel oxide (NiO), copper oxide (CuO), chromium oxide (Cr2O3), cobalt oxide (Co3O4) and manganese oxide (Mn3O4) are some of the most commonly used p-type metal-oxide gas sensors. The p-type gas sensors have attracted little attention compared to the n-type gas sensors [1]. Meanwhile, TiO2, among most metal-oxides, has shown good response to hydrogen [2–6]. It was revealed that TiO2 can be detected with hydrogen of as low as 1 ppm at room temperature [3]. Generally, TiO2 exists in anatase phase, rutile phase and brookite phase. The gas sensors often involve anatase phase and rutile phase [6,7]. Besides that, carbon-based gas sensor such as multi-walled carbon nanotube (MWCNT) and graphene has also recently gained growing interest given its unique electrical, optical and mechanical properties [8]. For instance, MWCNT reportedly identifies hydrogen [9,10],
∗
ammonia and methane [11] whereas graphene detects various types of gases such as carbon dioxide [12] and nitrogen dioxide [13]. Among the carbon-based gas sensors, MWCNT has shown good response to hydrogen at room temperature [9,10]. The addition of MWCNT in TiO2 was revealed to reduce the high resistivity of titanium and improve the response to hydrogen [14]. Therefore, this work added MWCNT in TiO2 for enhanced properties of gas sensor to hydrogen. The binder in thick film gas sensor enhances the sensor response and reduces the resistivity of a gas sensor [15]. Besides that, the addition of binder in a paste also improves the cohesion of particles after the firing process and enhances the adhesion of the sensing film to the substrate [16]. Most binders in thick film gas sensor are based on tetraethyl orthosilicate [15,17] and ethyl cellulose [16,18]. The addition of organic binder, specifically linseed oil in TiO2 paste for gas sensor was previously proposed [19], which showed good response to the hydrogen at elevated temperature. Other than that, linseed oil has been used in diverse fields; for examples, as diesel engine [20], to synthesize carbon nanofiber [21], as a varnish to mitigate wood degradation [22] and as anti-corrosive coating [23]. In view of the above, this work fabricated TiO2/MWCNT/B2O3 gas sensor with different binder and exposed to various concentrations of
Corresponding author. Institute of Advanced Technology, Universiti Putra Malaysia, 43400, Serdang, Selangor, Malaysia E-mail address: [email protected] (S.A. Mohd Chachuli).
https://doi.org/10.1016/j.mssp.2019.04.009 Received 28 December 2018; Received in revised form 29 March 2019; Accepted 5 April 2019 1369-8001/ © 2019 Elsevier Ltd. All rights reserved.
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
observe the structure of TiO2 and verify the elements of MWCNT after the heat treatment. The element composition and mapping of the elements were examined by the Energy Dispersive X-ray (EDX) inside the FESEM. The X-ray Diffraction (XRD) (Brand: Philips, Model: PW 3040/ 60 MPD X'pert High Pro Panalytical) studies were then carried out for powder and thick film in 5 min and 30 min, respectively, over a 2θ range between 20° and 80°. Raman spectroscopy (Brand: WITec, Model: Alpha 300 R) was carried out using a laser excitation line at 532 nm for powder and thick film. The absorption of both thick films was measured by UV–Visible spectrophotometer (Brand: Perkin Elmer, Model: Lambda 35) within the range of 280 nm–800 nm to calculate the optical band gap using Tauc plot. Surface analysis for both sensing material of gas sensor, OBL and OBE were characterized by Brunauer-EmmettTeller (BET) (Brand: Micromeritics, Model: TriStar II Plus) using nitrogen adsorption-desorption.
hydrogen at different operating temperatures. Different binders were applied to the TiO2/MWCNT/B2O3 paste to assess the properties of the gas sensor in terms of resistivity, band gap, sensitivity and optimal operating temperature. 2. Experimental details 2.1. Preparation of organic binder and TiO2/MWCNT/B2O3 pastes This work employed two different binders to prepare TiO2/ MWCNT/B2O3 paste for gas sensing purpose. One of binders used in this work was the conventional binder, which was based on ethyl cellulose and α-terpineol. The conventional binder was prepared according to the method presented in prior studies [24,25]. Besides that, the proposed organic binder was prepared by mixing the linseed oil, m-xylene and αterpineol using the magnetic stirrer to obtain homogeneous binder [19]. The proposed binder is labelled as OB-1 while the conventional binder is labelled as OB-2. The linseed oil, which is extracted from flax seeds oil, is eco-friendly and organic. Moreover, a homogeneous binder can be easily attained when linseed oil is used compared to the ethyl cellulose since the linseed oil is produced in liquid form, which allows better mixing. TiO2 (Aeroxide® P25) and MWCNT (OD = 10–20 nm, length = 10–30 μm) were obtained from Sigma-Aldrich and MaterialsA2Z, respectively. To prepare the sensing film of gas sensor, 95 wt % of TiO2 was mixed with 5 wt % of MWCNT using m-xylene as the medium in an ultrasonic bath for 90 min. Next, it was dried in an oven and ground in a mortar. Glass powder, specifically boron oxide (B2O3) was added to the paste to ensure good adhesion of sensing film on the alumina substrate. Following that, 10 wt % of B2O3 was mixed with 90 wt % of TiO2/MWCNT using m-xylene as the medium in an ultrasonic bath for 90 min. Then, it was dried in an oven and the produced powder was ground in a mortar. The addition of glass powder not affect the nanoparticles structure of TiO2 [19]. Both pastes were then prepared by mixing 30 wt % of TiO2/MWCNT/B2O3 and 70 wt % of organic binder using the magnetic stirrer until a homogeneous and viscous paste was obtained. A similar ratio was applied for both pastes to compare the performance of gas sensor to hydrogen. Accordingly, paste based on OB-1 was labelled as OBL and paste based on OB-2 was labelled as OBE.
2.4. Measurement of gas sensor The schematic diagram of experimental setup of gas sensor in a gas chamber can be referred in prior study [19]. 500 sccm of nitrogen was initially flowed in 600 s, then, followed by the flow of various concentrations of hydrogen between 100 ppm and 1000 ppm. Each cycle of hydrogen was subsequently flowed in 600 s. The nitrogen as carrier gas, flowed after the exposure to hydrogen in each cycle. Both gas sensors were tested at different operating temperatures: 100 °C, 200 °C and 300 °C and the measurement of gas sensor was conducted after achieving a stabilizing current at elevated operating temperature. 3. Results and discussion 3.1. Characterizations of binders by Fourier-transform infrared spectroscopy (FTIR) Fig. 1 presents the FTIR spectra of OB-1 and OB-2 in the range of 4000–400 cm−1. The apparent difference between both binders where OB-1 demonstrated a peak in the range of 2000–1500 cm−1 whereas OB-2 demonstrated the presence of OeH bond in the range of 4000–2500 cm−1 may have occurred due to the different mixture in both binders. Specifically, OB-2 did not contain m-xylene and as well as the respective functional groups of ethyl cellulose and linseed oil were included in the binders. In the case of OB-2, the bands centre of OeH bond was at 3500 cm−1 due to the presence of ethyl cellulose which belongs to the hydroxyl group. As for the case of OB-1, the peaks was at 1750 cm−1, which display the characteristic of C]O. The observed peak that represents the CeH stretching band in the region of 2800 cm−1 to 2900 cm−1 for both samples suggests the presence of
2.2. Fabrication of TiO2/MWCNT/B2O3 gas sensor To fabricate the gas sensor, an interdigitated electrode (IDE) based on silver conductive paste (Sigma-Aldrich) was deposited using screenprinting on alumina substrate as a first layer of gas sensor and annealed at 120 °C for 15 min. Following that, TiO2/MWCNT/B2O3 paste was printed as the second layer of the gas sensor and it was annealed in the furnace at 500 °C for 30 min. Accordingly, 500 °C was chosen as the annealing temperature because the paste based on linseed oil (OB-1) was completely evaporated at this temperature [19]. Both annealing processes were carried out using air as the carrier gas. The size of IDE used in this work was approximately 9.5 mm × 4.0 mm while the size of the sensing film was approximately 4.2 mm × 4.2 mm. The thickness of IDE and sensing film were approximately between 8 μm and 12 μm and the size of alumina substrate was about 20 mm × 15 mm. 2.3. Materials characterization The characterization of binder was defined by Fourier-Transform Infrared Spectroscopy (FTIR), (Brand: Thermo Nicolet, Model: Nicolet 6700) to identify the functional group for both binders (OB-1 and OB2). Meanwhile, the surface morphology of both thick films after annealed under air at 500 °C was characterized by the Field Emission Scanning Electron Microscopy (FESEM), (Brand: Thermo Fisher Scientific, Model: Nova Nanosem 230) at 100 k magnification to
Fig. 1. FTIR spectra of OB-1 and OB-2. 141
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 2. FESEM image of TiO2/MWCNT/B2O3 thick film annealed at 500 °C for (a) OBL and (b) OBE.
plan (002) of graphite hexagonal structure [27]. After sonication process of TiO2 with MWCNT for 90 min, carbon peak was not detected in XRD spectra of TiO2/MWCNT. However, after addition of B2O3 and sonication process in 90 min, it was found that the carbon peak was detected and shifted from 2θ = 25.95° to 2θ = 28.17° in TiO2/ MWCNT/B2O3, in which better dispersion of MWCNT was achieved after 3 h of sonication process as seen in Fig. 5. However, there is no carbon peak was detected after the annealing treatment at 500 °C for OBL and OBE (Fig. 6). This may due to the combination of carbon and oxygen during the annealing process under air in the furnace. In another study, Contreras et al. oxidized MWCNT under air at different temperature (between 270 – 600 °C) and found reduced weight loss of MWCNT when the temperature increased [28]. It was observed that the crystallinity of anatase phase at 2θ = 25.44° and rutile phase at 2θ = 27.61° increased after MWCNT was added to TiO2 and subsequently increased after B2O3 was added (Fig. 5). These results revealed that an extended sonication process enhances the crystallinity of anatase and rutile phase. After annealed treatment at 500 °C, the crystallinity of anatase phase at 2θ = 25.39° and rutile phase at 2θ = 27.45° for OBE was slightly higher than OBL (Fig. 6). Additionally, B2O3 peak was only detected after the annealing treatment at 500 °C, which at similar phase of rutile phase at 2θ = 36.25°. This also suggests that the crystallinity of B2O3 was achieved after the melting process was occurred. Raman characterization was carried out for original powder (TiO2 and MWCNT) and thick film (OBL and OBE). Raman spectrum for the samples in the range of 0–3500 cm−1 is shown in Fig. 7. As seen in Raman spectra, anatase peak was identified at 140 cm−1, 404 cm−1, 505 cm−1 and 638 cm−1 in TiO2 (P25). Whereas, MWCNT has three typical bands, D-band, G-band and 2D-band that correspond to 1343 cm−1, 1589 cm−1 and 2688 cm−1, respectively. The pattern of MWCNT also was observed in Ref. [29]. The peak of D-band was seen
organic compounds in both binders. Accordingly, the observed peaks at 1500 cm−1, 1150 cm−1, 900 cm−1 and 800 cm−1 in both binders were attributed to the CeH, CeO, C]C and CeH bond, respectively. 3.2. Characterization of TiO2/MWCNT/B2O3 thick films by FESEM, EDX, XRD and Raman Spectroscopy Fig. 2 presents the surface morphology of TiO2/MWCNT/B2O3 sensing film for OBL and OBE after undergone annealed treatment at 500 °C. The nanoparticles structure of TiO2 was clearly seen in both samples at 100 k magnification. Besides that, the structure of MWCNT was also apparent for both samples. However, sample based on OBE demonstrated higher quantity of MWCNT compared to the sample based on OBL. Average sizes of TiO2 nanoparticles ranged between 40 nm and 70 nm for both samples. As shown in Fig. 3, the EDX results revealed the presence of titanium, oxygen, carbon and aluminium (substrate). Both thick films generated high peak of titanium which indicates the crystallization of TiO2 at 500 °C. Weight of titanium for OBL and OBE was approximately 42.73% and 40.85%, respectively. In addition, the EDX results revealed that, the weight of carbon in OBL was approximately 4.29%, which was lower than the weight of carbon in OBE (4.41%). This implies the more carbon loss in OBL compared to the OBE during annealing at 500 °C, which was verified by the mapping in EDX (Fig. 4). The EDX analysis demonstrates the ability of OB-2 to hold more of carbon elements rather than OB-1. Fig. 5 displays XRD spectra of MWCNT, TiO2, TiO2/MWCNT and TiO2/MWCNT/B2O3 without the heat treatment, while Fig. 6 displays XRD spectra of TiO2/MWCNT/B2O3 after annealed treatment at 500 °C. Both figures showed high crystallinity of anatase phases and rutile phases in TiO2, TiO2/MWCNT and TiO2/MWCNT/B2O3, which are well consistent with TiO2 nanoparticles [26]. As for the case of MWCNT, high carbon peak was detected at 2θ = 25.95°, which was related to the
Fig. 3. EDX results of TiO2/MWCNT/B2O3 thick film on alumina substrate for (a) OBL and (b) OBE. 142
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 4. Mapping of MWCNT in sensing film on alumina substrate by EDX for (a) OBL and (b) OBE.
Fig. 7. Raman spectrum of TiO2, MWCNT, OBL and OBE.
Fig. 5. XRD spectra of MWCNT, TiO2, TiO2/MWCNT and TiO2/MWCNT/B2O3 without heat treatment.
TiO2 and MWCNT in the sensing film and also verified the crystallinity of the sensing film for gas sensor. 3.3. Surface analysis of TiO2/MWCNT/B2O3 by Brunauer-Emmett-Teller (BET) The surface area and pore size distribution of sensing material, OBL and OBE were further characterized by nitrogen adsorption-desorption isotherm using BET analysis. The sample was prepared by dropped the paste (OBL and OBE) on the alumina substrate and annealed under air at 500 °C for 30 min. Table 1 lists average pore size and BET surface area for both sensing material. It indicated that average pore size of OBE smaller than OBL, while surface area of OBE was larger than OBL. This analysis also confirmed distinctively the porosities of both sensing films. Higher surface area maybe contributed to the larger response of sensing film to the target gas due to the ability to facilitate the diffusion of the target gas [32,33]. Fig. 8 shows nitrogen gas adsorption-desorption isotherms shape of IV with a H3 hysteresis loop of the sensing material, OBL and OBE according to the IUPAC classification [34]. This hysteresis loop is usually attributed to the thermodynamic or/and network effects and it also indicative of the pores in the sensing film of gas sensor [35]. It can
Fig. 6. XRD spectra of sensing film based on OBL and OBE after annealed at 500 °C.
occurred at ∼1340 cm−1 and G-band at ∼1580 cm−1 [30,31]. Degree of graphitization can be calculated using ID/IG ratio, where the value is 0.85 for MWCNT, OBL and OBE. This ratio indicated higher value of graphitization based on ID/IG value is less than 1 [29]. The ID/IG ratio for OBL and OBE showed similar value with MWCNT, which indicates no defects occurred during synthesis of TiO2/MWCNT/B2O3. Peaks in the Raman spectrum for OBL and OBE also confirmed the presence of
Table 1 Average pore size and BET surface area of OBL and OBE.
143
Sensing material
Average pore size (nm)
BET surface area (m2/g)
OBL OBE
5.6473 5.4275
52.8219 57.3509
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 8. Nitrogen gas adsorption-desorption isotherms and pore size distribution of (a) OBL and (b) OBE.
Fig. 9. Optical band gap of (a) TiO2 (P25); (b) TiO2/B2O3 thick film; (c) OBL and; (d) OBE.
where α is absorption coefficient, Eg is absorption band gap, A is a constant (depending on the transition probability), hv is photon energy and n depends on the nature of transition [37]. Eg was obtained by draw a tangent line on graph of (αhv)n versus hv. For indirect band gap, n equals to ½ and as for direct band gap, n equals to 2 [37]. Fig. 9 shows the band gap values of TiO2 (P25), TiO2/B2O3 thick film and TiO2/MWCNT/B2O3 thick film for both samples of OBL and OBE. Band gap of TiO2 (P25) was measured in liquid form by dispersion of TiO2 (P25) powders in ethanol. It was found that the calculated band gap of TiO2 (P25) was based on direct band gap which approximately 3.25 eV. It has been reported in the literature the band gap of TiO2 (P25) can be existed in direct band gap or indirect band gap [38–40]. Wang et al. also calculated TiO2 (P25) powders at different calcination
be seen that, the pore size distribution of OBL is slightly higher than OBE (Fig. 8) in which similar result as tabulated in Table 1. Changes in hysteresis loop, pore size distribution and surface area of the sensing material can affect the performance of sensing film in a gas sensor [35]. 3.4. Optical band gap of TiO2/MWCNT/B2O3 by UV–Vis spectroscopy Absorption data of TiO2/MWCNT/B2O3 thick film from the UV–Vis Spectroscopy is used to calculate the optical band gap (Eg) of TiO2/ MWCNT/B2O3 according to the proposed model by Tauc [36]. The following equation was used to generate the graph of (αhv)n versus hv,
αhv n = A (hv − Eg ) 144
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 10. Resistance of TiO2/B2O3 and TiO2/MWCNT/B2O3 (OBL and OBE) gas sensor at different operating temperatures (a) T = 100 °C; (b) T = 200 °C; and (c) T = 300 °C.
Fig. 11. Response of TiO2/MWCNT/B2O3 (OBL and OBE) gas sensor to hydrogen with different binder at different operating temperature (a) T = 100 °C; (b) T = 200 °C; and (c) T = 300 °C.
revealed at this temperature. It should be noted that, similar absorption produces similar optical band gap. The similar values of band gap for both samples (OBL and OBE) also indicated that different binder does not affect the semiconductor properties of TiO2.
treatment based on direct band gap [41]. Whereas, TiO2/B2O3 thick film band gap in previous work [19] was calculated based on indirect band gap, which the value was approximately 4.1 eV (Fig. 9(b)). Prior studies also associated TiO2 (P25) thick film to the indirect band gap [42,43]. When MWCNT was added into TiO2, band gaps of TiO2/ MWCNT/B2O3 thick film (OBL and OBE) were also calculated based on indirect band gap measurement (Fig. 9(c) and (d)). With the addition of MWCNT to TiO2, the band gap was shown to reduce to 3.72 eV. With reduction of band gap, the response time of a gas sensor can become faster to detect the target gas and increase the sensitivity of gas sensor. Both thick films (OBL and OBE) produced similar band gap of approximately 3.72 eV. It may because of the mass loss of binder was completely loss at 500 °C, therefore similar absorption of UV light was
3.5. Resistance of TiO2/MWCNT/B2O3 gas sensor at different operating temperature The resistance of TiO2/B2O3 gas sensor and TiO2/MWCNT/B2O3 (OBL and OBE) gas sensor at different operating temperatures under nitrogen flowed are displayed in Fig. 10. The addition of MWCNT reduced the resistance of TiO2 gas sensor due to the higher carrier mobility of MWCNT [8]. The percentage of reduction was approximately 145
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Fig. 12. Sensitivity of TiO2/MWCNT/B2O3 (OBL and OBE) gas sensor to hydrogen with different binder at different operating temperatures (a) T = 100 °C; (b) T = 200 °C; and (c) T = 300 °C.
and OBE) responded well to 1000 ppm and 500 ppm of H2 at all operating temperatures (100–300 °C). At the operating temperature of 100 °C, both gas sensors (OBL and OBE) were unable to sense 100 ppm of H2. As the operating temperature increased to 200 °C and 300 °C, both gas sensors (OBL and OBE) showed good response to 100 ppm of H2 due to the slower kinetics of electron movement at lower operating temperature and higher adsorption at higher operating temperature [45]. The sensitivity of gas sensor of various concentrations of hydrogen at different operating temperatures is displayed in Fig. 12. As compared to the previous work [19], the sensitivity of gas sensor in this work improved with the addition of MWCNT in TiO2. At the operating temperature of 100 °C, the sensitivity for both gas sensors did not demonstrate significant difference at 500 ppm and 1000 ppm of H2. On the other hand, the OBL-based gas sensor demonstrated significantly different sensitivity from the sensitivity of OBE-based gas sensor at operating temperature of 200 °C for 500 ppm and 1000 ppm of H2. In particular, OBE-based gas sensor produced highest sensitivity compared to the OBL-based gas sensor at 500 ppm and 1000 ppm of H2. However, the sensitivity of both gas sensors to 500 ppm and 1000 ppm of H2 did not show significant difference at the operating temperature of 300 °C. These results indicated that the optimal operating temperature for both gas sensor in this work is 200 °C. Besides that, the OBL-based gas sensor, compared to OBE-based gas sensor showed better recovery characteristic at 200 °C and 300 °C where its sensitivity value was able to achieve the original value which is one. Sensing mechanism of a gas sensor can be influenced by surfacecontrolled mode, grain-controlled mode and neck-controlled mode [46]. For OBL-based gas sensor and OBE-based gas sensor, surfacecontrolled mode are similar due to the similar compact layer structure and thickness of gas sensor produced by screen-printing technique.
99.86%, 95.50% and 91.79% at operating temperature of 100 °C, 200 °C and 300 °C, respectively. Overall, the resistance of the gas sensor decreased as temperature increased. The resistance of OBL-based gas sensor and the resistance of OBE-based gas sensor demonstrated significant difference at operating temperature of 100 °C due to the presence of OeH bond in the OBE-based gas sensor. In other words, the occurrence of higher oxidation results in higher resistance. Moreover, the weight of oxygen in the both samples varied, where OBL and OBE contributed 51.92% and 54.45% (including Al2O3), respectively. However, the resistance of both gas sensors did not exhibit significant difference at operating temperature of 200 °C and 300 °C, which may due to the decomposition of silver oxide (electrode) at operating temperature of above 200 °C [44], thus resulting in similar resistance. 3.6. Response of TiO2/MWCNT/B2O3 gas sensor to hydrogen at different operating temperature The response of TiO2/MWCNT/B2O3 gas sensor to hydrogen with different binder at different operating temperatures is shown in Fig. 11. From the response, the gas sensor behaved as a p-type gas sensor based on the decreased current when it was exposed to hydrogen and the increased current when it was exposed to nitrogen. This behaviour also reported in Ref. [14]. Besides that, the initial current of OBL-based gas sensor was slightly higher than the observed current OBE-based gas sensor at different operating temperatures. These obtained results indicated that the linseed oil-based gas sensor produces better conductivity and low resistivity compared to the ethyl cellulose-based gas sensor based. Higher conductivity of OBL-based gas sensor can be caused by its ability to hold more titanium elements after annealed treatment at 500 °C compared to the OBE-based gas sensor as referred to the EDX results (Fig. 3). In term of responses, both gas sensors (OBL 146
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
Besides, similar band gap produced by OBL and OBE indicates the surface-controlled mode are similar for both gas sensors. Whereas, grain-controlled mode and neck-controlled mode contribute to the electroconductibility and performance of gas sensor [46]. As seen in Fig. 12, OBE-based gas sensor exhibits higher sensitivity at optimal operating temperature compared to the OBL-based gas sensor, this might be due to the higher surface area of OBE as tabulated in Table 1, in which ease diffusion of hydrogen to the sensing film. Higher response for larger surface area also reported in Refs. [32,33]. This results also revealed that larger grain size can improve response of a gas sensor to the target gas. It also can be seen that OBL-based gas sensor has better recovery characteristic in which ability to return to its original resistance value after exposed to the hydrogen compared to the OBEbased gas sensor. This phenomenon might because of higher pore size in OBL, thus ease for the target gas to dismiss from the surface of a gas sensor.
based hydrogen gas sensor, Sensors Actuators, A Phys. 201 (2013) 321–327. [10] S. Dhall, N. Jaggi, Highly dispersed platinum sputtered multiwall carbon nanotubes based hydrogen gas sensor at room temperature, Sensors Actuators, A Phys. 224 (2015) 50–56. [11] N. Kumar, P. Sahatiya, P. Dubey, Fabrication of CNT based gas sensor using interdigitated gold electrodes, Procedia Materials Science 6 Icmpc, 2014, pp. 1976–1980. [12] J. Wu, S. Feng, Z. Li, K. Tao, J. Chu, J. Miao, L.K. Norford, Boosted sensitivity of graphene gas sensor via nanoporous thin film structures, Sensor. Actuator. B Chem. 255 (2018) 1805–1813. [13] Y. Seekaew, D. Phokharatkul, A. Wisitsoraat, Applied Surface Science Highly sensitive and selective room-temperature NO2 gas sensor based on bilayer transferred chemical vapor deposited graphene, Appl. Surf. Sci. 404 (2017) 357–363. [14] L. De Luca, A. Donato, S. Santangelo, G. Faggio, G. Messina, N. Donato, G. Neri, Hydrogen sensing characteristics of Pt/TiO2/MWCNTs composites, Int. J. Hydrogen Energy 37 (2) (2012) 1842–1851. [15] F. Engzneenng, Effects of tetraethyl orthosilicate binder on the characteristics of an SnO2, ceramic-type semiconductor gas sensor, Sensor. Actuator. 9 (1986) 133–145. [16] V. Guidi, M.A. Butturi, M.C. Carotta, B. Cavicchi, M. Ferroni, C. Malagù, G. Martinelli, D. Vincenzi, M. Sacerdoti, M. Zen, Gas sensing through thick film technology, Sensor. Actuator. B Chem. 84 (1) (2002) 72–77. [17] S.J. kim, G. Lee, The effect on the gas selectivity of gas sensors by binder in SWNTs solution, IEEE Sens. J. 10 (1) (2010) 1–5. [18] S. Lee, G.G. Lee, J. Kim, S.J.L. Kang, A novel process for fabrication of SnO2-based thick film gas sensors, Sensor. Actuator. B Chem. 123 (1) (2007) 331–335. [19] S. Amaniah, M. Chachuli, M.N. Hamidon, M.S. Mamat, M. Ertugrul, N.H. Abdullah, A hydrogen gas sensor based on TiO2 nanoparticles on alumina substrate, Sensors 18 (2483) (2018) 1–13. [20] H. Hazar, H. Sevinc, Investigation of the effects of pre-heated linseed oil on performance and exhaust emission at a coated diesel engine, Renew. Energy 130 (2018) 961–967. [21] D.E. Kshirsagar, V. Puri, H. Dubey, M. Sharon, Giga hertz frequency absorber carbon nano fibers synthesized using linseed oil, Mater. Today Commun. 13 (August) (2017) 23–25. [22] J.J. Łucejko, J. La Nasa, C.M.A. McQueen, S. Braovac, M.P. Colombini, F. Modugno, Protective effect of linseed oil varnish on archaeological wood treated with alum, Microchem. J. 139 (2018) 50–61. [23] H. Wang, Q. Zhou, Evaluation and failure analysis of linseed oil encapsulated selfhealing anticorrosive coating, Prog. Org. Coating 118 (November 2017) (2018) 108–115. [24] M.H. Seo, M. Yuasa, T. Kida, J.S. Huh, N. Yamazoe, K. Shimanoe, Detection of organic gases using TiO2 nanotube-based gas sensors, Procedia Chem. 1 (1) (2009) 192–195. [25] M. Zhang, G. Jiang, Gas sensing properties of Co3O4-loaded SnO2 to ethanol and acetone, Chin. J. Chem. Phys. 20 (3) (2007) 315–318. [26] R. Atchudan, T. Nesakumar, J. Immanuel, S. Perumal, D. Karthikeyan, Y. Rok, Effective photocatalytic degradation of anthropogenic dyes using graphene oxide grafting titanium dioxide nanoparticles under UV-light irradiation, J. Photochem. Photobiol. A Chem. 333 (2017) 92–104. [27] M.B. Askari, Z. Tavakoli Banizi, M. Seifi, S. Bagheri Dehaghi, P. Veisi, Synthesis of TiO2 nanoparticles and decorated multi-wall carbon nanotube (MWCNT) with anatase TiO2 nanoparticles and study of optical properties and structural characterization of TiO2/MWCNT nanocomposite, Opt. - Int. J. Light Electron Opt. 149 (2017) 447–454. [28] C. Wongchoosuk, A. Wisitsoraat, D. Phokharatkul, A. Tuantranont, T. Kerdcharoen, Multi-walled carbon nanotube-doped tungsten oxide thin films for hydrogen gas sensing, Sensors 10 (8) (2010) 7705–7715. [29] R. Atchudan, B.G. Cha, N. Lone, J. Kim, J. Joo, Synthesis of high-quality carbon nanotubes by using monodisperse spherical mesoporous silica encapsulating iron oxide nanoparticles, Kor. J. Chem. Eng. 36 (1) (2019) 157–165. [30] N. Muthuchamy, R. Atchudan, T. Nesakumar, J. Immanuel, High-performance glucose biosensor based on green synthesized zinc oxide nanoparticle embedded nitrogen-doped carbon sheet, J. Electroanal. Chem. 816 (2018) 195–204 March. [31] R. Atchudan, T. Nesakumar, J. Immanuel, S. Perumal, R. Vinodh, Y. Rok, In-situ green synthesis of nitrogen-doped carbon dots for bioimaging and TiO2 nanoparticles @ nitrogen-doped carbon composite for photocatalytic degradation of organic pollutants, J. Alloy. Comp. 766 (2018) 12–24. [32] Q. Rong, Z. By, H. Jc, Z. Zq, J. Zhang, L. Qj, Methanol sensing performance of gas sensor based on Ag-LaFeO3 modified with graphene, J. Mol. Sci. 2 (1:2) (2018) 1–5. [33] C. Li, M. Lv, J. Zuo, X. Huang, SnO2 Highly Sensitive CO Gas Sensor Based on QuasiMolecular-Imprinting Mechanism Design, (2015), pp. 3789–3800. [34] M. Thommes, K. Kaneko, A. V Neimark, J.P. Olivier, F. Rodriguez-reinoso, J. Rouquerol, K.S.W. Sing, Physisorption of gases , with special reference to the evaluation of surface area and pore size distribution (IUPAC Technical Report), Pure Appl. Chem. 87 (2015) 1051–1069. [35] W. Tan, X. Ruan, Q. Yu, Z. Yu, X. Huang, Fabrication of a SnO2-Based Acetone Gas Sensor Enhanced by Molecular Imprinting, (2015), pp. 352–364. [36] J. Tauc, R. Grigorovici, A. Vancu, Optical properties and electronic structure of amorphous germanium, 15 (2) (1966) 627–637. [37] W. Saidi, N. Hfaidh, M. Rasheed, M. Girtan, A. Megriche, M. El Maaoui, Effect of B2O3 addition on optical and structural properties of TiO2 as a new blocking layer for multiple dye sensitive solar cell application (DSSC), RSC Adv. 00 (January) (2013) 1–3. [38] M. Scarsella, M.P. Bracciale, B. De Caprariis, P. De Filippis, A. Petrullo, L. Pronti, M.L. Santarelli, Improved photocatalytic properties of doped titanium-based nanometric oxides, Chem. Eng. Trans. 60 (2017) 133–138.
4. Conclusions The addition of MWCNT in TiO2 improves the sensitivity and reduces the band gap, resistance, and the optimal operating temperature of TiO2 gas sensor. The TiO2/MWCNT/B2O3 gas sensor with different binders to various concentrations of hydrogen also demonstrates good response at different operating temperatures. In terms of sensitivity, the ethyl cellulose-based gas sensor exhibits better response to hydrogen. However, linseed oil-based gas sensor has better conductivity and better recovery characteristic after its exposure to hydrogen. Adding to that, the optimal operating temperature for TiO2/MWCNT/B2O3 gas sensor for OBL and OBE in this work is 200 °C given the significant difference in the observed sensitivity at 100 ppm, 500 ppm and 1000 ppm of hydrogen. This work also reveals that linseed oil–based gas sensor can be developed as an alternative of ethyl cellulose-based gas sensor in which it offers low cost fabrication, good response to hydrogen and better recovery characteristics of gas sensor. Acknowledgment The author would like to thank the members of the Physics Laboratory, Science Faculty, Ataturk University, Turkey for the facilities provided to conduct gas sensor measurement. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.mssp.2019.04.009. References [1] Z. Wen, L. Tian-mo, Gas-sensing properties of SnO2-TiO2-based sensor for volatile organic compound gas and its sensing mechanism, Phys. B Condens. Matter 405 (5) (2010) 1345–1348. [2] X. Peng, Z. Wang, P. Huang, X. Chen, X. Fu, W. Dai, Comparative study of two different TiO2 film sensors on response to H2 under UV light and room temperature, Sensors (Switzerland) 16 (8) (2016). [3] X. Xia, W. Wu, Z. Wang, Y. Bao, Z. Huang, Y. Gao, A hydrogen sensor based on orientation aligned TiO2 thin films with low concentration detecting limit and short response time, Sensor. Actuator. B Chem. 234 (2016) 192–200. [4] A. Haidry, P. Schlosser, P. Durina, M. Mikula, M. Tomasek, T. Plecenik, T. Roch, A. Pidik, M. Stefecka, J. Noskovic, M. Zahoran, P. Kus, A. Plecenik, Hydrogen gas sensors based on nanocrystalline TiO2 thin films, Open Phys. 9 (5) (2011). [5] O. Krško, T. Plecenik, T. Roch, B. Grančič, L. Satrapinskyy, M. Truchlý, P. Ďurina, M. Gregor, P. Kúš, A. Plecenik, Flexible highly sensitive hydrogen gas sensor based on a TiO2 thin film on polyimide foil, Sensor. Actuator. B Chem. 240 (2017) 1058–1065. [6] E. Şennik, Z. Çolak, N. Kilinç, Z.Z. Öztürk, Synthesis of highly-ordered TiO2 nanotubes for a hydrogen sensor, Int. J. Hydrogen Energy 35 (9) (2010) 4420–4427. [7] K. Zakrzewska, M. Radecka, “TiO2-Based nanomaterials for gas sensing—Influence of anatase and rutile contributions, Nanoscale Res. Lett. 12 (1) (2017). [8] E. Llobet, Gas sensors using carbon nanomaterials: a review, Sensor. Actuator. B Chem. 179 (2013) 32–45. [9] S. Dhall, N. Jaggi, R. Nathawat, Functionalized multiwalled carbon nanotubes
147
Materials Science in Semiconductor Processing 99 (2019) 140–148
S.A. Mohd Chachuli, et al.
(1) (2017) 243. [43] C.-H. Wei, C.-M. Chang, Polycrystalline TiO2 thin films with different thicknesses deposited on unheated substrates using RF magnetron sputtering, Mater. Trans. 52 (3) (2011) 554–559. [44] W.E. Garner, L.W. Reeves, The thermal decomposition of silver oxide, (1953), pp. 254–260 254. [45] C. Wang, L. Yin, L. Zhang, D. Xiang, R. Gao, Metal oxide gas sensors: sensitivity and influencing factors, Sensors 10 (3) (2010) 2088–2106. [46] Y. Wang, T. Wu, Y. Zhou, C. Meng, W. Zhu, L. Liu, TiO2-based nanoheterostructures for promoting gas sensitivity performance: Designs, developments, and prospects, Sensors 17 (9) (2017) 1–35.
[39] S. Valencia, J.M. Marín, G. Restrepo, Study of the bandgap of synthesized titanium dioxide nanoparticules using the sol-gel method and a hydrothermal treatment, Open Mater. Sci. J. 4 (2010) 9–14. [40] P. Malliga, J. Pandiarajan, N. Prithivikumaran, K. Neyvasagam, Effect of Film Thickness on Structural and Optical Properties of TiO2 Thin Films, (2013), pp. 488–491. [41] G. Wang, L. Xu, J. Zhang, T. Yin, and D. Han, “Enhanced photocatalytic activity of TiO2 powders (P25) via calcination treatment,” Int. J. Photoenergy, vol. 2012, 2012. [42] Y.-J. Shi, R.-J. Zhang, H. Zheng, D.-H. Li, W. Wei, X. Chen, Y. Sun, Y.-F. Wei, H.L. Lu, N. Dai, L.-Y. Chen, Optical constants and band gap evolution with phase transition in sub-20-nm-thick TiO2 films prepared by ALD, Nanoscale Res. Lett. 12
148