Study on room temperature gas-sensing performance of CuO film-decorated ordered porous ZnO composite by In2O3 sensitization

For the first time, ordered mesoporous ZnO nanoparticles have been synthesized by a template method. The electroplating after chemical plating method was creatively used to form copper film on the surface of the prepared ZnO, and then a CuO film-decorated ordered porous ZnO composite (CuO/ZnO) was obtained by a high-temperature oxidation method. In2O3 was loaded into the prepared CuO film–ZnO by an ultrasonic-assisted method to sensitize the room temperature gas-sensing performance of the prepared CuO/ZnO materials. The doped In2O3 could effectively improve the gas-sensing properties of the prepared materials to nitrogen oxides (NOx) at room temperature. The 1% In2O3 doped CuO/ZnO sample (1 wt% In2O3–CuO/ZnO) showed the best gas-sensing properties whose response to 100 ppm NOx reached 82%, and the detectable minimum concentration reached 1 ppm at room temperature. The prepared materials had a good selectivity, better response, very low detection limit, and high sensitivity to NOx gas at room temperature, which would have a great development space in the gas sensor field and a great research value.

For the first time, ordered mesoporous ZnO nanoparticles have been synthesized by a template method. The electroplating after chemical plating method was creatively used to form copper film on the surface of the prepared ZnO, and then a CuO film-decorated ordered porous ZnO composite (CuO/ZnO) was obtained by a high-temperature oxidation method. In 2 O 3 was loaded into the prepared CuO film-ZnO by an ultrasonicassisted method to sensitize the room temperature gas-sensing performance of the prepared CuO/ZnO materials. The doped In 2 O 3 could effectively improve the gas-sensing properties of the prepared materials to nitrogen oxides (NO x ) at room temperature. The 1% In 2 O 3 doped CuO/ZnO sample (1 wt% In 2 O 3 -CuO/ZnO) showed the best gas-sensing properties whose response to 100 ppm NO x reached 82%, and the detectable minimum concentration reached 1 ppm at room temperature. The prepared materials had a good selectivity, better response, very low detection limit, and high sensitivity to NO x gas at room temperature, which would have a great development space in the gas sensor field and a great research value. chemical plating method was used to form copper film, and the n-p heterojunction between CuO film and ZnO matrix played an important role in the enhancement of the sensor response and selectively. The In 2 O 3 played a sensitizer role to sensitize the surface reaction. In 2 O 3 was a catalyst, it reduced the activation energy of the oxidation-reduction reaction on the surface of composites, efficiently separated the oxygen molecules, improved the activity of gas-sensitive materials, increased the free electron amount and expedited the reaction rate. As such, the gas sensitivity of the composites was enhanced. The prepared material greatly reduced the working temperature of the sensor, which would have a great development prospect and great research value in the sensor field.

. Preparation of mesoporous ZnO nanoparticles
One gram of CTMAB was added into 480 ml of deionized water, and the mixture was vigorously stirred until the solution was homogeneous at 80°C. About 4.92 g of Zn(Ac) 2 · 2H 2 O was dissolved into the solution with stirring. The pH of the solution was adjusted with LiOH to alkaline; the reaction then proceeded at 80°C for 2 h. The product was filtered, washed with deionized water and dried at room temperature to get the powder sample. The original sample was calcined at 500°C for 4 h, and the ordered mesoporous ZnO material was finally obtained.

Preparation of CuO film on the mesoporous ZnO matrix
The electroless plating solution was prepared according to table 1. The prepared ordered porous ZnO material was added into the electroless plating solution, which was reacted for 1 h under vigorous stirring. The powder was filtered, washed with deionized water and dried to afford a reddish brown powder with metallic lustre.
The prepared powder was put into semipermeable membrane, and the negative pole was inserted into the powder. The powder was immersed in the plating solution (table 2). The sheet copper was used as anode, voltage was regulated at 17 V, electricity was regulated at 1 A and the plating solution was electrified for 1 h. The powder was filtered, washed with deionized water, dried, calcined at 500°C for 4 h to afford a black CuO film-ZnO powder sample.

Preparation of In 2 O 3 sensitization CuO film-ZnO composite material
In order to study the effect of different In 2 O 3 decoration amounts on the gas-sensing performance, a series of In 2 O 3 decorated CuO film-ZnO composites were synthesized. In a typical procedure, 0.1 g prepared CuO film-ZnO product was dispersed in 20 ml deionized water with the aid of ultrasonication. A certain amount of In(NO 3 ) 3 (In(NO 3 ) 3 concentration: 10 wt%, 15 wt%, 20 wt% and 25 wt%) was added in the above solution. The reaction system was dispersed in ultrasonic for 10 min, and the suspension liquid was aged for 2 h at room temperature. The suspension liquid was filtered, washed with deionized water and dried for 10 h at 60°C. The prepared samples were heated for 4 h at 500°C with a warming rate of 2°C min −1 to afford the products. The In-loaded contents were obtained with an inductively coupled   plasma-optical emission spectrometer (ICAP 6000 Series), and the corresponding doping contents of In 2 O 3 were calculated which were 0.5, 1, 3 and 6 wt%, respectively.

Characterization
Powder X-ray diffraction (XRD) patterns were collected with a Siemens D5005 diffractometer using Cu-Kα radiation (λ = 1.5418 Å and operating at 30 kV and 20 mA). Transmission electron microscopy (TEM) images were taken with a JEOL 2010 TEM instrument. The scanning election microscopy (SEM) images were taken with a JEOL JSM-5600 L. Physical adsorption of nitrogen was performed with a Micromeritics ASAP2010M volumetric adsorption analyser at 77 K. A sample was degassed in vacuum at 573 K for 12 h before measurement. Surface area was calculated based on the BET model, while pore size was computed using the Barrett-Joyner-Halenda method.

The preparation of gas sensor
In this study, an interdigitated Au electrode was selected for the gas-sensing detection. The sample was disposed as follows: about 3 mg of the prepared sample was dispersed in ethanol, which was dispersed ultrasonically for about 20 min. The sample material was spin-coated on the interdigitated electrode surface to form a thin film, which was dried to obtain the gas sensor. The parameters of interdigitated gold electrode and schematic diagram of the prepared gas sensor are shown in figure 1.
The sensor was welded onto a socket, and the electrical properties of the sensor were measured by a JF02E intelligent gas-sensing analysis system. The sensor response was defined as Here, R a and R g were the resistances of the sensors in the air and target gas, respectively. The response and recovery time were defined as the time taken by the sensors to achieve 85% of the total resistance change in the case of adsorption and desorption, respectively [38][39][40].

Results and discussion
3.1. Structure and morphology Figure 2 shows the wide-angle XRD pattern of the prepared ordered mesoporous 1 wt% In 2 O 3 -CuO/ZnO sample. The sharp diffraction peaks indicated the highly crystalline characteristic of 1 wt% In 2 O 3 -CuO/ZnO samples. All of the diffraction peaks were consistent with the standard card of CuO (JCPDS No. 44-0706), which showed that the CuO has been deposited on the surface of the ZnO. The diffraction peaks of wurtzite ZnO (JCPDS No. 36-1451) could be seen in the prepared material. In the XRD pattern, the characteristic peaks of the In 2 O 3 could also be seen. It showed that the In 2 O 3 has been doped into the prepared CuO film-ZnO matrix material. Figure 3 shows the energy-dispersive X-ray spectra of CuO/ZnO and 1 wt% In 2 O 3 -CuO/ZnO samples. From figure 3a, the peaks of C, Zn, O, Si and Cu could be seen in CuO/ZnO sample. The sample was first placed on a silicon wafer when the sample was tested, which led to the highest Si content peak. The C and O were mainly from the conducting resin. In figure 3b, the indium (In) content peaks could be seen in 1 wt% In 2 O 3 -CuO/ZnO sample, which could prove the existence of the In 2 O 3 in 1 wt% In 2 O 3 -CuO/ZnO sample, and the In 2 O 3 has been successfully assembled into the CuO film-ZnO matrix. The Pt content peak was attributed to the Pt conductive agent sprayed on the surface of the sample. There were no peaks of other elements in samples, which indicated that the prepared samples were pure. Figure 4 shows the SEM image of CuO film-ZnO and 1 wt% In 2 O 3 -CuO/ZnO samples. From figure 4a, the CuO/ZnO sample showed a whole flower-like appearance, whose surface was smooth. From figure 4b, the 1 wt% In 2 O 3 -CuO/ZnO sample showed an analogous flower-like structure, which obtained from the petals of flower-like CuO/ZnO sample fallen off, the analogous flower-like structures were heaped with irregular flaky grain. Figure 5a shows the TEM image of the 1 wt% In 2 O 3 -CuO/ZnO sample, which was taken parallel to the direction of the channels. From figure 5a, it could be seen that there were a lot of ordered pores in the sample. Figure 5b shows the high-resolution TEM (HRTEM) image of the 1% In 2 O 3 -CuO/ZnO sample. From figure 5b, it could be seen that the lattice fringes of the sample were very clear, which were the lattice fringes of CuO. In the HRTEM images, the ordered pores could be very clearly seen under the CuO lattice fringes, which indicated that the CuO thin films were formed on the surface of the prepared sample. Figure 6a shows the gas response bar charts of different In 2 O 3 -doped concentration samples to different concentrations of NO x . The study found that the sensitivity of each sample decreased when NO x concentration decreased from 100 to 5 ppm. The results showed that In 2 O 3 could significantly enhance the response of the materials to NO x , and the 1 wt% In 2 O 3 -CuO/ZnO sample had the best response to NO x among five prepared (0, 0.5, 1, 3, 6 wt%) In 2 O 3 -CuO/ZnO sensor samples in this study. At room temperature, the response and response time of the sensors with different In 2 O 3 contents to 100 ppm NO x gas are shown in the electronic supplementary material, table S1.

Gas-sensing performance
Under the current operating conditions, the response-recovery curves of 1 wt% In 2 O 3 -CuO/ZnO sensor are shown in figure 6b. The resistance sharply decreased when the NO x gas was injected, and the resistance recovered to its initial value when the NO x gas was discharged. When the prepared sensor was in air, oxygen molecules were adsorbed on the surface of the sensor; the adsorbed oxygen captured electrons from the conductance band to produce negatively charged oxygen species. These chemisorbed oxygen species would act as surface acceptors, trapping electrons and increasing surface resistance of the sensor. When the NO x gas was injected, the NO x would react with the formed oxygen ions, which released the trapped electrons back to the sensor surfaces, and the resistance of the sensor was sharply decreased. It indicated that the prepared sensing material was a p-type semiconductor material. Figure 6c shows the response and response time curve of the 1 wt% In 2 O 3 -CuO/ZnO sample to 100-1 ppm NO x at room temperature. When the NO x gas was injected into the sensing chamber, the resistance of the 1 wt% In 2 O 3 -CuO/ZnO sensor rapidly decreased, and reached the minimum resistance value in a short time. The 1 wt% In 2 O 3 -CuO/ZnO sensor exhibited a fast and reversible response to NO x , the response time being only 7 s to 100 ppm NO x which was much shorter than that reported in a lot of literature [26,39,[41][42][43], and the corresponding response reached up to 82%. The responses were weakened with the reduction in the NO x concentration from 100 to 1 ppm, and the corresponding response time increased. It was related to the gas concentration, gas diffusion and adsorption on the    surface of the nanomaterials. With decreasing of NO x concentration, the NO x gas partial pressure declined in the reaction chamber which led to gas phase reaction velocity being slower, and the response time increased. It is worth noting that the minimum detection limit of 1     relatively high response to NH 3 . When the NH 3 concentration was 100 ppm, the response reached 48%. However, the response to NH 3 of the sensor was also significantly lower than the response to the same concentration of NO x . The response of other gases (O 2 , CO and H 2 ) was closer to 0%. The experimental results showed that the 1 wt% In 2 O 3 -CuO/ZnO gas sensor had good selectivity.
The humidity is one of the significant factors in the gas-sensing process, and the degree of humidity influences the sensing performance of oxide semiconductors. However, the humidity has little effect on NO x gas response in this study. The result is shown in figure 7. Figure 8 shows the stability bar charts of 1 wt% In 2 O 3 -CuO/ZnO sample. The response of the prepared sample remained stable after six months. It indicated that the prepared 1 wt% In 2 O 3 -CuO/ZnO sample had good stability. layer will be formed. The negative surface charges will generate a higher surface potential barrier, which leads to a high electrical resistance [46,47]. The process can be expressed as follows:

Discussion of sensing mechanism
When the sensor is exposed in oxidizing NO x gas, the NO x gases attract the electrons from sensor materials, which can be attributed to their high electron affinity. The process leads to electrons of the sensor materials transfering to the surface NO x , and forming NO 2 − (from NO 2 ) and NO − (from NO).
The process traps electrons from the conduction band or donor level of materials, which leads to electron density reduction and the hole carrier density increase in the sensor. The hole carriers result in a decrease in resistance of the semiconductor layer. It will lead to a thinner space-charge layer, and reduce the potential barrier at the grain boundaries and cause a decrease in the sensor resistance [39]. The process can be expressed as follows: Also, the target gases NO 2 and NO directly adsorb on the sensor, which react with O − and generate bidentate NO 3 − (s) and NO 2 − , and release the trapped electrons back into the conduction band. The process can be expressed as follows [48]: The band gap (E g ) of ZnO is 3.4 eV, and its electron affinity (c) is 4.35 eV. The band gap of CuO is 1.35 eV, and its electron affinity is 4.07 eV. From a thermodynamic perspective, the electrons will transfer from the conduction band of CuO to ZnO, while the holes will oppositely migrate from the valence band of ZnO to CuO [49]. The process will increase the charge separation rate of electron-hole pairs under the bias potential, while the longer lifetime of electron-hole pairs of CuO/ZnO will be attributed to the formation of the p-n heterojunction. The p-n heterojunction may facilitate the adsorption of gas molecules. Thus, CuO/ZnO composite shows a higher sensing activity than pure ZnO, the pure ZnO has no gas response at room temperature. However, the formation of a p-n junction results in higher resistance [22]. The higher sensitivity of CuO/ZnO composite can be attributed to the electronic sensitization induced by CuO [50,51]. There are more active sites and more chemisorbed oxygen species on the surface of p-type semiconductors than n-type semiconductors [51,52]. The modified CuO nanostructures can act as a strong acceptor of electrons from ZnO [50]. The Fermi energy levels of n-type ZnO and p-type CuO are equalized which is due to the charge transfer [53,54]. The introduction of CuO film on the surface of the porous ZnO increases the carrier density of the ZnO, which results in the electron redistribution between the ZnO and the CuO [55]. In the air, oxygen molecules capture electrons from CuO, and chemisorbed oxygen species are formed on its surface [56]. Meanwhile, electrons transfer from ZnO to CuO, which causes the formation of an electron depletion layer (EDL) extending into ZnO [51]. The EDL significantly narrows the conducting part and dominates the conductivity of CuO/ZnO [57,58]. The decoration of CuO promotes the chemisorption of oxygen species and the chemical reactions on the surface of CuO/ZnO [51,59]. This leads to an enhancement in the sensitivity of CuO/ZnO composite [60].
In addition, high specific surface area of sensing material usually has positive effects on the sensing response [61]. The porous structure helps to increase the gas-sensing performance. The porous structure provides abundant pores, which is more beneficial to diffusion of gas molecules [62,63]. Moreover, it also provides more active sites, and improves the kinetics of chemical reactions on the surface [63]. This also results in enhanced sensing performance. Furthermore, there would be more oxygen species chemisorbed on the surface of CuO film structures due to its large surface to volume ratio [64]. Thus, the sensitivity of the CuO/ZnO composite is further enhanced. In this study, CuO is introduced onto the surface of ordered mesoporous zinc oxide by electrodeposition after chemical deposition method. The active site density is improved by the advantages of porous and large specific surface area of mesoporous materials. CuO is dispersed on the porous surface of the mesoporous ZnO matrix material to provide an effective gas diffusion channel for CuO contacting with the gas. The provided multiple active sites make NO x easier to react with adsorbed oxygen ions on the surface, which increases gas sensitivity performance.
On the surface of the prepared CuO/ZnO composite, the heterogeneous structure can be formed between CuO and ZnO by ultrasonic method, and the gas-sensitive properties can be enhanced by In 2 O 3 . Generally, the n-n (or p-n) type heterojunction-based interfacial barriers between In 2 O 3 and ZnO (or    Figure 9. A schematic diagram of sensing mechanism of 1 wt% In 2 O 3 -CuO/ZnO sensor. CuO) nanocrystals are another mechanism accounting for the enhanced sensing performance [45,[65][66][67].
In this study, the doping content is very little which forms discontinuity phase. A possible gas-sensing mechanism is as follows: in the band of interfaces, there is a spike at the In 2 O 3 side along with a notch at the host side. The discontinuity band sets barriers for electrons in both conduction band and valence band. The heterojunction interfacial barriers can strongly modify the charge transport behaviours of carriers because the carrier concentration varies exponentially with the barrier height. Further studies are required to determine the mechanism through which the In 2 O 3 -doped semiconductor interfacial barriers promote the sensing performance towards gases [68]. Also, the catalysis of the In 2 O 3 should not be overlooked. The catalytic activity of In 2 O 3 accelerates the dissociation of oxygen molecules, and causes a spillover of the adsorbed oxygen ions on the surface of 1% In 2 O 3 -CuO/ZnO composite. More adsorbed oxygen ions provide more sensing activity sites and shows high response [45,[69][70][71]. In this study, the doped quantity In 2 O 3 is very little, the p-n heterojunction or n-n/p-p homo-type heterojunction cannot be formed, and the In 2 O 3 only plays a role of sensitizer to sensitize the surface reaction. With the In 2 O 3 content increasing, the sensitization is enhanced. But, the sensitized mechanism is changed when the doped In 2 O 3 quantity exceeded the limit, which can be explained with heterojunction mechanism, and the response of the sensor will be the lower when the In 2 O 3 content is not enough to form the continuous heterojunction.
In this study, the remarkable gas-sensing performance of the prepared 1 wt% In 2 O 3 -CuO/ZnO sensor may be for the following reasons. First, the fluffy porous structure is helpful for gas diffusion and surface reaction, which lead to the sensor exhibiting faster response. Second, n-p heterojunction between CuO and ZnO plays an important role in the enhancement of the sensor response and selectively. The CuO has a lower Fermi level than ZnO, and ZnO would receive electrons from CuO, leading to the formation of an accumulation layer at the CuO/ZnO interface. The increase in electrons on the surface of ZnO is able to facilitate adsorption of oxygen molecules and a decrease in resistance. Third, the doping of In 2 O 3 is also helpful for remarkable gas-sensing performance of 1 wt% In 2 O 3 -CuO/ZnO composite. The reasons include electron interactions between In 2 O 3 and ZnO, the catalytic activity of In 2 O 3 accelerates the dissociation of oxygen molecules, and causes a spillover of the adsorbed oxygen ions on the surface of 1 wt% In 2 O 3 -CuO/ZnO composite. More adsorbed oxygen ions provide more sensing sites, and show high response. The mechanism is intuitively explained in figure 9.

Conclusion
An ordered mesoporous ZnO nanomaterial has been synthesized by a template method. The prepared porous ZnO nanomaterial is used as matrix to prepare a porous Cu film-ZnO composite by the electroplating after chemical plating method, and finally a CuO film-decorated ordered porous ZnO composite is obtained by a high-temperature oxidation method. Then, In 2 O 3 is loaded into the prepared CuO/ZnO by an ultrasonic-assisted method, affording an In 2 O 3 -CuO/ZnO composite. The prepared 1 wt% In 2 O 3 -CuO/ZnO composite shows the best gas-sensing performance, in which the In 2 O 3 doping content is only 1 wt%. For 1 wt% In 2 O 3 -CuO/ZnO sample, its response time is only 7 s to 100 ppm NO x , and the corresponding response reaches 82%. It is worth noting that the minimum detection limit of 1 wt% In 2 O 3 -CuO/ZnO gas sensor was only 1 ppm at room temperature, the corresponding response still reached 17% and the corresponding response time was 136 s. The response of the 1 wt% In 2 O 3 -CuO/ZnO gas sensor is much higher, the minimum detection limit is much lower, the selectivity and stability are much better and the corresponding response time is much shorter at room temperature than that reported in a lot of literature.