Sputtered SnO 2 / ZnO Heterostructures for Improved NO 2 Gas Sensing Properties

: A highly sensitive and selective NO 2 gas sensor dependent on SnO 2 / ZnO heterostructures was fabricated using a sputtering process. The SnO 2 / ZnO heterostructure thin ﬁlm samples were characterized by ﬁeld emission scanning electron microscopy (FESEM), X-ray di ﬀ raction (XRD), Energy-dispersive X-ray spectroscopy (EDS), and X-ray photoelectron spectroscopy (XPS). Sensors fabricated with heterostructures attained higher gas response (S = 66.9) and quicker response-recovery (20 s, 45 s) characteristics at 100 ◦ C operating temperature towards 100 ppm NO 2 gas e ﬃ ciently in comparison to sensors based on their mono-counterparts. The selectivity and stability of SnO 2 / ZnO heterostructures were studied. The more desirable sensing mechanism of SnO 2 / ZnO heterostructures towards NO 2 was described in detail.


Introduction
The effective detection of numerous harmful and toxic gases has become extremely vital for human health safety. Nitrogen oxide (NO 2 ), mostly released by fossil fuel combustion and motor vehicles, is one of the most hazardous air pollutants. It can cause various fatal diseases even at very low concentrations. Hence, the development of the NO 2 gas sensor is extremely important for health protection and environmental applications [1,2]. Metal oxide semiconductors (MOS) have been studied and utilized as effective and dominant methods in the detection of explosive or toxic gases and environmental monitoring [3,4]. Consequently, the variety of semiconductor metal oxides such as WO 3 [5], In 2 O 3 [6], NiO [7], ZnO [8], and SnO 2 [2] with numerous morphologies have been studied for gas sensing applications to detect various gases that include reducing and oxidizing gases. In spite of all its advantages, the development of a highly selective gas sensor based on MOS is still a challenge [9]. The sensing characteristics of the MOS are dependent on its morphology, composition, and crystalline size.
To improve the selectivity and sensor response of MOS many approaches are utilized that includes the doping of the transition metal, loading of the noble metal catalyst, and developing binary metal oxides. Previously, several studies confirmed that the sensing materials consisting of two MOS exhibited better sensing characteristics than their mono counterparts [10]. Thus, many hybrid materials such as SnO 2 /Fe 2 O 3 , CeO 2 /ZnO, In 2 O 3 /ZnO, and SnO 2 /ZnO with numerous morphologies have been considered for gas sensing applications and attained improved sensor response [11][12][13][14]. As the important gas sensing materials, SnO 2 and ZnO having band gaps of 3.6 eV and 3.4 eV, individually, previously reported for gas sensors. Recently, several studies have shown that sensing performance SnO 2 or ZnO can be highly enhanced by the formation of SnO 2 /ZnO heterostructures [15].
In the present study, NO 2 gas response was highly enhanced by the use of the SnO 2 /ZnO heterostructures synthesized by RF-sputtering. So far, SnO 2 /ZnO heterostructures have been broadly studied to sense different oxidizing gases. The SnO 2 /ZnO heterostructures were used to fabricate an NO 2 gas sensor, and gas sensing characteristics were studied by varying operating temperature and gas concentrations. The improvement in sensing performance may be credited to the formation of SnO 2 /ZnO heterostructures.

Experimental
For the formation of gas sensors, an SiO 2 -deposited silicon substrate was prepared with inter-digitated electrodes including that of a Pt (200 nm) layer that was coated on a sensor platform through the sputtering process. The deposition of thin-film SnO 2 and ZnO was done by utilizing RF-sputtering from SnO 2 and ZnO targets. The co-sputtering of the two targets was done at room temperature (RT). The base pressure inside the sputtering chamber was held around 5.5 × 10 −6 mbar. The thin films were deposited at a O 2 /Ar flow ratio of 50:50. The sputtering powers utilized for SnO 2 and ZnO were 50 and 80 W, respectively. The working pressure of the sputtering chamber was sustained at 9.0 × 10 −3 mbar to attain oxide films. Lastly, heat treatments at 500 • C were done for 3 h in air to confirm the stability of the gas sensor. The morphological properties of the sensing film were characterized by utilizing FE-SEM and EDS that was combined with FE-SEM. The structural characteristics of the sensing film were examined through XRD using Cu Kα X-radiation with a wavelength of 1.55178 Å and an analysis of surface elements was made by XPS.
The sensor characteristics were examined by using the flow-through technique that is described elsewhere [16]. In the gas sensing chamber, two tungsten wires were utilized for an electrical connection to the gas sensor. A sequences of mass flow controllers (MFCs) were utilized for maintaining the insertion of NO 2 gas inside the sensing chamber. Figure 1 shows the schematic for the gas sensing set-up along with the dimensions of the gas sensor. Initially, dry air was inserted inside until the saturation in the baseline sensor resistance was touched. The sensor electrical resistance was uninterruptedly noted by a Keithley multimeter (model 2600A) coupled to a laptop for continuously switching the NO 2 and dry air on and off through every cycle. The flow rate of gases was maintained around 200 sccm. Gas response is considered as S = R a /R g , where R g and R a are the electrical resistances in NO 2 gas and dry air, respectively. From this experiment, a gas concentration of 100 ppm NO 2 diluted nitrogen (N 2 ) and diluted with dry air as a carrier utilizing MFCs to attain a lower concentration. The concentration of the gas was premeditated as NO 2 (ppm) = NO 2std (ppm) × f/(f + F), where F and f are the flow rates of dry air and NO 2 gas, respectively. The selectivity of the sensor was measured with other interfering gases such as acetone, hydrogen, methanol, ammonia, and ethanol.

Results and discussions
The surface morphology of the SnO2/ZnO was examined by FESEM. Figure 2a displays the FESEM image of the thin film displayed uniformly distributed columnar microstructure, conforming to the zone structure model. Similar nanostructures were observed in another report and also in [17].

Results and Discussions
The surface morphology of the SnO 2 /ZnO was examined by FESEM. Figure 2a displays the FESEM image of the thin film displayed uniformly distributed columnar microstructure, conforming to the zone structure model. Similar nanostructures were observed in another report and also in [17]. The thickness of SnO 2 /ZnO thin film was around 10 nm, verified by cross-sectional view as shown in inset of Figure 2a. The elemental distribution of the SnO 2 /ZnO thin film was studied by EDS analysis.

Results and discussions
The surface morphology of the SnO2/ZnO was examined by FESEM. Figure 2a displays the FESEM image of the thin film displayed uniformly distributed columnar microstructure, conforming to the zone structure model. Similar nanostructures were observed in another report and also in [17]. The thickness of SnO2/ZnO thin film was around 10 nm, verified by cross-sectional view as shown in inset of Figure 2a. The elemental distribution of the SnO2/ZnO thin film was studied by EDS analysis.  The XRD pattern shows the crystallographic structure and phase purity of the SnO2/ZnO thin film, as shown in Figure 3. The diffraction displayed sharp, clear, and strong diffraction peaks, that are well matched with the rutile-like structure of SnO2 (JCPD # 41-1445) and hexagonal wurtzite phase of ZnO (JCPD # 36-1451), respectively [18]. Moreover, no impurity peaks were observed; that is a clear indication for the formation of SnO2/ZnO heterostructures. The XRD pattern shows the crystallographic structure and phase purity of the SnO 2 /ZnO thin film, as shown in Figure 3. The diffraction displayed sharp, clear, and strong diffraction peaks, that are well matched with the rutile-like structure of SnO 2 (JCPD # 41-1445) and hexagonal wurtzite phase of ZnO (JCPD # 36-1451), respectively [18]. Moreover, no impurity peaks were observed; that is a clear indication for the formation of SnO2/ZnO heterostructures. XPS study was done to inspect the chemical states of Sn and Zn and the surface composition in SnO2/ZnO heterostructures. As presented in Figure 4a, the XPS survey spectrum indicates that the SnO2/ZnO heterostructures show the existence of Sn, Zn, and O elements. Figure 4b presented the high-resolution spectra for Sn 3d. Two peaks at binding energies 492.5 eV and 484 eV can be credited to the binding energies of Sn 3d3/2 and 3d5/2, respectively, which are assigned Sn 4+ cations [19]. The doublet peaks conforming to Zn 2p1/2 and Zn 2p3/2 were spotted, as shown in Figure 3c. The peak positions of Zn 2p1/2 and Zn 2p3/2 were situated at 1043.3 eV and 1020.5 eV, respectively, and the binding energy distance between Zn 2p1/2 and Zn 2p3/2 is 23.2 eV, signifying the Zn species were XPS study was done to inspect the chemical states of Sn and Zn and the surface composition in SnO 2 /ZnO heterostructures. As presented in Figure 4a, the XPS survey spectrum indicates that the SnO 2 /ZnO heterostructures show the existence of Sn, Zn, and O elements. Figure 4b presented the high-resolution spectra for Sn 3d. Two peaks at binding energies 492.5 eV and 484 eV can be credited to the binding energies of Sn 3d 3/2 and 3d 5/2 , respectively, which are assigned Sn 4+ cations [19]. The doublet peaks conforming to Zn 2p 1/2 and Zn 2p 3/2 were spotted, as shown in Figure 4c. The peak positions of Zn 2p 1/2 and Zn 2p 3/2 were situated at 1043.3 eV and 1020.5 eV, respectively, and the binding energy distance between Zn 2p 1/2 and Zn 2p 3/2 is 23.2 eV, signifying the Zn species were present in the chemical state Zn 2+ [20]. The O 1s spectra is displayed in Figure 4d; the broad peak of O 1s is irregular and can be fitted into two peaks at 528.6 eV and 529.7 eV that corresponds to chemical sates for O in the SnO 2 /ZnO heterostructures. The peak at 528.6 eV binding energy corresponds to lattice oxygen denoted by O I in Figure 4d, whereas the peak at 529.7 eV is related to the chemisorbed oxygen denoted by O II in Figure 4d. Therefore, the surface oxygen (O 2 ) absorbed capacity was significantly enhanced and improved reacting with the target gas species that resulted in the attainment of a high gas response as a sensing material [21].  For MOS gas sensors, operating temperature plays a dynamic role in determining the gas response of gas sensor. It is because of the dependence of desorption, and adsorption procedures of O2 molecules takes place on the surface of the sensor [22]. Hence, first we examined the gas response of pure SnO2, ZnO, and SnO2/ZnO heterostructure samples towards NO2 gas at various operating temperatures fluctuating from 25-400 °C, and the result is presented in Figure 5a. The gas response is increased in all the three cases, reached to the maximum value and then rapidly decreased. The optimum operating temperature was about 100 °C, as can be seen in the Figure 5a. Moreover, compared to their mono-counterparts the SnO2/ZnO exhibits about a six times higher gas response towards NO2 gas. The enhanced gas response in the instance of SnO2/ZnO may be credited to the formation of heterostructures. At lower working temperatures gas kinetics is low and results in a lower sensor response. Similarly, at higher working temperatures, more than optimal working temperature, the kinetics of gas species more than that of molecules, might seepage from their active sites of the surface, formerly the reaction, and would disturb the amount of adsorbed gas. Henceforth, they are subsequent in a low sensor response. At optimal working temperature (100 °C) the sensor response is higher which might be because of more surface interactions and oxygen vacancies. The consequences also display the outcome of crystallinity reliant on sensing performances for thin-film gas sensors wherever higher crystallinity achieves an enhanced sensor response. The dynamic gas response of the SnO2/ZnO heterostructures sensor to various NO2 concentrations varies from 5-50 ppm, as displayed in Figure 5b. The gas response is increased with a rise in NO2 gas concentration. The sensor exhibits a noticeable gas response from 26.4 towards 5 ppm NO2 with very rapid response-recovery times of around 20 s and 45 s, respectively. For MOS gas sensors, operating temperature plays a dynamic role in determining the gas response of gas sensor. It is because of the dependence of desorption, and adsorption procedures of O 2 molecules takes place on the surface of the sensor [22]. Hence, first we examined the gas response of pure SnO 2 , ZnO, and SnO 2 /ZnO heterostructure samples towards NO 2 gas at various operating temperatures fluctuating from 25-400 • C, and the result is presented in Figure 5a. The gas response is increased in all the three cases, reached to the maximum value and then rapidly decreased. The optimum operating temperature was about 100 • C, as can be seen in the Figure 5a. Moreover, compared to their mono-counterparts the SnO 2 /ZnO exhibits about a six times higher gas response towards NO 2 gas. The enhanced gas response in the instance of SnO 2 /ZnO may be credited to the formation of heterostructures. At lower working temperatures gas kinetics is low and results in a lower sensor response. Similarly, at higher working temperatures, more than optimal working temperature, the kinetics of gas species more than that of molecules, might seepage from their active sites of the surface, formerly the reaction, and would disturb the amount of adsorbed gas. Henceforth, they are subsequent in a low sensor response. At optimal working temperature (100 • C) the sensor response is higher which might be because of more surface interactions and oxygen vacancies. The consequences also display the outcome of crystallinity reliant on sensing performances for thin-film gas sensors wherever higher crystallinity achieves an enhanced sensor response. The dynamic gas response of the SnO 2 /ZnO heterostructures sensor to various NO 2 concentrations varies from 5-50 ppm, as displayed in Figure 5b. The gas response is increased with a rise in NO 2 gas concentration. The sensor exhibits a noticeable gas response from 26.4 towards 5 ppm NO 2 with very rapid response-recovery times of around 20 s and 45 s, respectively.  Figure 6a displays the reproducibility test of the SnO2/ZnO sensor by measuring the gas response for five successive cycles after the insertion to NO2 gas at even intermissions. It is clear from Figure 6a, that the SnO2/ZnO sensor exhibits variation in resistance from Ra to Rg after insertion and injection to NO2 gas; the sensor recovers its initial value Ra from Rg. This result shows excellent reproducibility of the SnO2/ZnO sensor. Selectivity is a significant factor to study gas sensing performance of the gas sensor. Hence, for the practical opinion of application, the gas sensor must present high selectivity. Thus, a superior consideration has been given in this study, to the cautious valuation of selectivity of the sensor. The SnO2/ZnO heterostructure sensor is examined by measuring numerous interfering gases such as acetone, hydrogen, methanol, ammonia, and ethanol at 100 °C towards 100 ppm NO2 gas. The sensor displays low gas response to other interfering gases excluding NO2, signifying its high specific adsorption capability towards NO2, as revealed in Figure 6b. High gas response to NO2 is credited to its high electron-withdrawing ability, in comparison to other interfering gases. The sensing mechanism of n-type MOS sensors have been studied through the space-charge layer method [23,24]. The electrical resistance of MOS will vary after being exposed to various  Figure 6a displays the reproducibility test of the SnO 2 /ZnO sensor by measuring the gas response for five successive cycles after the insertion to NO 2 gas at even intermissions. It is clear from Figure 6a, that the SnO 2 /ZnO sensor exhibits variation in resistance from R a to R g after insertion and injection to NO 2 gas; the sensor recovers its initial value R a from R g . This result shows excellent reproducibility of the SnO 2 /ZnO sensor. Selectivity is a significant factor to study gas sensing performance of the gas sensor. Hence, for the practical opinion of application, the gas sensor must present high selectivity. Thus, a superior consideration has been given in this study, to the cautious valuation of selectivity of the sensor. The SnO 2 /ZnO heterostructure sensor is examined by measuring numerous interfering gases such as acetone, hydrogen, methanol, ammonia, and ethanol at 100 • C towards 100 ppm NO 2 gas. The sensor displays low gas response to other interfering gases excluding NO 2 , signifying its high specific adsorption capability towards NO 2 , as revealed in Figure 6b. High gas response to NO 2 is credited to its high electron-withdrawing ability, in comparison to other interfering gases.  Figure 6a displays the reproducibility test of the SnO2/ZnO sensor by measuring the gas response for five successive cycles after the insertion to NO2 gas at even intermissions. It is clear from Figure 6a, that the SnO2/ZnO sensor exhibits variation in resistance from Ra to Rg after insertion and injection to NO2 gas; the sensor recovers its initial value Ra from Rg. This result shows excellent reproducibility of the SnO2/ZnO sensor. Selectivity is a significant factor to study gas sensing performance of the gas sensor. Hence, for the practical opinion of application, the gas sensor must present high selectivity. Thus, a superior consideration has been given in this study, to the cautious valuation of selectivity of the sensor. The SnO2/ZnO heterostructure sensor is examined by measuring numerous interfering gases such as acetone, hydrogen, methanol, ammonia, and ethanol at 100 °C towards 100 ppm NO2 gas. The sensor displays low gas response to other interfering gases excluding NO2, signifying its high specific adsorption capability towards NO2, as revealed in Figure 6b. High gas response to NO2 is credited to its high electron-withdrawing ability, in comparison to other interfering gases. The sensing mechanism of n-type MOS sensors have been studied through the space-charge layer method [23,24]. The electrical resistance of MOS will vary after being exposed to various The sensing mechanism of n-type MOS sensors have been studied through the space-charge layer method [23,24]. The electrical resistance of MOS will vary after being exposed to various oxidizing or reducing gases. In air, O 2 molecules can adsorb onto the surface that leads to surface adsorbed O 2 species (O − (ads) , O − 2(ads) , O 2− (ads) ) by taking free e − from the conduction band (CB). The reactions are defined as in Equations (1)- (4).
During the process, a wide e − depletion layer is formed that results in the decrement of the charge carrier concentration and the rise in the electrical resistance of the sensor. After exposure to the target gas, absorbed O 2 species react to target gas molecules. This results in e-trapped in O 2 species that are sent back to CB that leads to an increase in the width of the depletion layer and a rise in the electrical resistance of the as-prepared sensor. According to literature, selectivity of the sensor depends on the many aspects like the lowest unoccupied molecules orbit (LUMO) energy of target gas molecules and the adsorption of the gas molecules onto the surface of sensing materials at various operating temperatures. In addition, electron affinity can be changed by the orbital energy of the target gas molecules. The smaller the LUMO energies the higher the gas molecules' capability to capture electrons. Thus, at an operating of temperature 100 • C, LUMO energy of NO 2 is lesser compared to other interfering gases, the capability of taking e − of NO 2 will be sturdier compared to other interfering gases, and hence the sensor displays a high gas response towards NO 2 gas [25].
These results exhibited that SnO 2 /ZnO heterostructures have better gas response compared to their mono counterparts. The improvement in gas response for SnO 2 /ZnO heterostructures can be credited to subsequent features. At first, the surface of SnO 2 /ZnO heterostructures leads to both oxides being highly available for adsorption of the O 2 species, resulting in the creation of a depletion layer onto the surface. Thus, synergetic effects of the oxides perhaps donate the improvement of the gas response in comparison to their mono counterparts. Then, work functions of SnO 2 and ZnO that have stated 4.9 eV and 5.2 eV, respectively, results in the formation of heterojunctions among SnO 2 and ZnO [26]. Shown in Figure 7, is the schematic of energy band diagrams of the SnO 2 /ZnO heterojunction in dry air and NO 2 gas. The flow of electrons form SnO 2 to ZnO till the fermi levels (FE) are balanced. This generates an e − depletion layer onto the ZnO and bend SnO 2 energy band that results in a change in electrical resistance of SnO 2 /ZnO heterostructures. As a sensor was exposed to NO 2 gas at the optimal operating temperature, trapped e − were sent back to the CB of SnO 2 /ZnO heterostructures because of the reaction between adsorbed O 2 species and NO 2 molecules. Therefore, electrical resistance of SnO 2 /ZnO heterostructures was highly increased resulting in an enhanced gas response.
During the process, a wide edepletion layer is formed that results in the decrement of the charge carrier concentration and the rise in the electrical resistance of the sensor. After exposure to the target gas, absorbed O2 species react to target gas molecules. This results in e-trapped in O2 species that are sent back to CB that leads to an increase in the width of the depletion layer and a rise in the electrical resistance of the as-prepared sensor. According to literature, selectivity of the sensor depends on the many aspects like the lowest unoccupied molecules orbit (LUMO) energy of target gas molecules and the adsorption of the gas molecules onto the surface of sensing materials at various operating temperatures. In addition, electron affinity can be changed by the orbital energy of the target gas molecules. The smaller the LUMO energies the higher the gas molecules' capability to capture electrons. Thus, at an operating of temperature 100 °C, LUMO energy of NO2 is lesser compared to other interfering gases, the capability of taking eof NO2 will be sturdier compared to other interfering gases, and hence the sensor displays a high gas response towards NO2 gas [25]. These results exhibited that SnO2/ZnO heterostructures have better gas response compared to their mono counterparts. The improvement in gas response for SnO2/ZnO heterostructures can be credited to subsequent features. At first, the surface of SnO2/ZnO heterostructures leads to both oxides being highly available for adsorption of the O2 species, resulting in the creation of a depletion layer onto the surface. Thus, synergetic effects of the oxides perhaps donate the improvement of the

Conclusions
In conclusion, the NO 2 sensors formed by SnO 2 /ZnO heterostructures were fabricated and examined in this work. The morphological and structural characteristics of the SnO 2 /ZnO heterostructures are identified by numerous characterization techniques. The results for SnO 2 /ZnO heterostructures showed outstanding NO 2 sensing characteristics. The main reason for high gas performance was due to the SnO 2 and ZnO establishing N-N heterojunctions that significantly rise the electrical resistance of the sensor compared to pure SnO 2 and ZnO. This may be the foremost cause for the improved gas response towards NO 2 . Our study offers a balanced method for the fabrication and design of resistive gas sensors with enhanced sensing characteristics.