In₂O₃-decorated ordered mesoporous NiO for enhanced NO₂ sensing at room temperature

Abstract

In this work, ordered mesoporous structures of In₂O₃-decorated NiO were prepared by a two-step process, comprising of the synthesis of ordered mesoporous NiO followed by injection of In³⁺ into their pores. The pore size distribution of the as prepared samples was between 4.1 and 21.1 nm. Furthermore, their sensing performances toward NO₂ were tested systematically. The results showed the highest response about 3 towards 15 ppm NO₂ sensing at room temperature for 5.0 at.% In₂O₃-decorated NiO compared to other decorated and pure samples. Moreover, the sensor displayed excellent selectivity towards NO₂ in the presence of other interfering gases, such as carbon monoxide, ammonia, ethanol, methanol, formaldehyde, toluene, acetone. The exceptional NO₂ sensing performance of the In₂O₃-decorated mesoporous NiO may be attributed to their high specific surface area and the formation of p–n junction with modified carrier concentration caused by In³⁺ doping. This method can act as an effective strategy for enhancement of gas-sensing properties of pure metal oxides.

1 Introduction

In the current times, air pollution has caused a great deal of harm to human health. NO₂ being a major air pollutant, causes adverse effects on plants and respiratory system of human beings and animals. Moreover, as a source of photochemical smog and acid rain, it causes serious environmental damages [1]. In addition, asthma diagnosis can be performed by detecting the concentration of NO₂ in the air exhaled by patients [2]. Therefore, the detection of NO₂ has attracted tremendous interest in the context of health and environment monitoring and thus, it is necessary to develop a NO₂ gas sensor with excellent properties, including low detection limit concentration, high sensitivity, good selectivity, and low cost. Metal oxide semiconductors (MOSs) have been widely used in the detection of a variety of toxic and harmful gases owing to its high stability, low cost, and convenience [3].

Nickel oxide(NiO), a typical p-type semiconductor with a wide bandgap (3.6–4.0 eV), due to its high chemical and thermal stability, unique magnetic, optical, catalytic, and electrical properties, has been widely applied in many areas, such as chip fabrication [4], medical diagnosis [5], fabrication of electrochemical supercapacitors [6], catalytic combustion [7], and gas sensing [8]. As a matter of fact, studies on n-type semiconductors have been far more in number than those corresponding to p-type materials [9]. Therefore, it is necessary to explore and develop p-type oxide semiconductor gas sensors which possess advantages of high catalytic activity and enhanced stability.

However, to the best of our knowledge, there are only a few reports on gas sensing studies of NiO at room temperature and the response in these reports is limited [10,11,12]. Hence, it is still essential and challenging to develop a low cost and effective gas sensitive material based on NiO.

Morphology are very significant for the detection of materials. Afkhami et al. successfully attached NiFe₂O₄ nanoparticles to Gr for detecting tramadol and acetaminophen [13]. Bagheri et al. synthesized Fe₃O₄–SnO₂–Gr ternary nanocomposite with the formation of uniform and monodispersed Fe₃O₄–SnO₂ for the detection of dopamine, uric acid and ascorbic acid [14]. And Gas-sensitive materials also require proper structure.

In general, the gas sensing performance of metal oxide semiconductors is usually determined by three factors, namely, receptor function, transducer function, and utility factor [15]. In short, the first factor is related to the response of each crystal, the second to electronic transport between adjacent crystals, and the third to effective diffusion and recognition of gas molecules on the reactive surface. Therefore, enhancing adsorption, diffusion of target gas, and mobility of charge carriers would benefit the performance of these gas sensors. Li et al. synthesized 3D Ni(OH)₂ and NiO nanowalls for HCHO and NO₂ sensing [16]. Miao et al. reported the preparation of NiO flower-like microspheres with abundant nanoparticles adhering to the petals, enabling the detection of ethanol [17]. Nguyen et al. found that NO₂ detection can be achieved by mesoporous NiO nanosheets [18]. Moreover, doping other metal elements is also an effective means to enhance the gas-sensing response. Sun and coworkers reported the fabrication of Fe-doped ordered mesoporous NiO based sensors for detecting VOCs (volatile organic compounds) [19]. Tian et al. demonstrated the synthesis of Lotus pollen derived 3D hierarchically porous NiO microspheres for NO₂ sensing and Kim et al. discovered the superior ethanol sensing properties of In₂O₃-decorated NiO [20, 21]. Considering the high cost of noble metals, it makes more sense to improve gas sensing performance by doping base metals. Thirdly, composite reduced graphene and mesoporous carbon are favorable for electron transport [22, 23].

Based on the literature survey as presented above, in this work, we designed the synthesis of ordered mesoporous NiO with large specific surface area, small pore, and particle size, followed by the injection of In³⁺ ions. Hence, the In₂O₃-decorated ordered mesoporous NiO was successfully prepared via a facile two-step process. For the synthesis of ordered mesoporous NiO, 3D cubic mesoporous KIT-6 silica was used as a hard template. And it was found that it was In doping that greatly enhanced the sensitivity of pure NiO to NO₂. Furthermore, we found that the 5.0 at.% In₂O₃-decorated NiO exhibited an excellent selectivity toward NO₂ at room temperature.

2 Experimental

2.1 Material preparation

2.1.1 Synthesis of KIT-6

The synthesis of KIT-6 was carried out according to a previously reported procedure [24]. About 6 g of P-123 was dissolved in 11.8 g of HCl(35%) and 217 g of H₂O. Then, 6 g of n-butanol was added and the solution was stirred for 1 h at 35 °C. Then, 12.9 g of tetraethyl orthosilicate (TEOS) was added, and the resultant solution was stirred for 24 h at 35 °C. It was then transferred to a Teflon-lined stainless steel autoclave and placed in an oven at 70 °C for 24 h. The resulting precipitate was filtered, washed, with deionized water, and dried at 550 °C at a ramp rate of 2 °C/min and calcined at this temperature in air for 5 h.

2.1.2 Synthesis of mesoporous NiO

The ordered mesoporous NiO was synthesized with the hard template named KIT-6, which was previously reported by Sun et al [25]. Typically,0.6 g of the KIT-6 silica template was dispersed in 20 ml of ethanol solution containing 1.45 g nickel nitrate hexahydrate (Ni(NO₃)₂·6H₂O). The mixture under vigorous stirring was heated to 65 °C and kept at the temperature for 4 h. Due to the capillary condensation effect between the two substances, the Ni²⁺ can permeate into the template pores. Afterwards, the resulting sample was harvested by centrifugation and wash with distilled water for several times. Then, the as-obtained sample was transferred into a furnace and heated up to 500 °C at a ramp rate of 2 °C/min and calcined at this temperature in air for 3 h to vary the nickel nitrate precursors to NiO. To remove the silica template, the resulting NiO@KIT-6 powders were handled with 2 M NaOH solution. The precipitate was recovered by centrifugation and washing alternately with deionized water and ethanol for 3 times, respectively.

2.1.3 Synthesis of In-decorated mesoporous NiO

The periodic ordered NiO materials decorated with In₂O₃ were synthesized by an incipient-wetness impregnation method [26]. In a typical procedure, the as-prepared NiO materials were dispersed in 30 mL deionized water, then a certain amount of In(NO₃)₃·4.5H₂O (the molar ratios of In/Ni were 2.0, 5.0, 10.0 at.%, respectively) were added into the dispersion. Following impregnation, the obtained powders were kept at room temperature for 24 h, dried at 80 °C in air overnight, and then calcined at 500 °C for 3 h. The calcination temperature was risen from room temperature at a heating rate of 2 °C/min. These specimens are referred to as “2.0 at.% In–NiO” and “5.0 at.% In–NiO” and “10.0 at.% In-NiO”.

2.2 Characterization

The crystalline phases of the powder samples was analyzed by the wide-angle X-ray diffraction(XRD, Shimadzu XRD-6000) using CuKα1 radiation (λ = 0.15406 nm) at 40 kV and 20 mA. The XRD pattern were obtained from 20° to 80° with a scanning rate of 5°/min. The morphology and crystal structure were performed by transmission electron microscopy(TEM, JEOL, JEM-2010). The doped compositions of the samples were examined by the energy dispersive X-ray spectrometers(EDS, JEOL, JSM-5510LV). Moreover, the mesoporous characteristics of the pure NiO and In-doped NiO were examined by nitrogen physisorption isotherm techniques on a Micromeritics ASAP2020 system. The specific surface areas of samples were evaluated using the Brunauer–Emmett–Teller (BET) method. The electronic structure of the surface of 5.0 at.% In-dope NiO was investigated by X-ray photoelectron spectroscopy(XPS, K-Alpha).

2.3 Fabrication and gas-sensing measurement

The as-prepared products were ground into fine powders and mixed with ethanol to form paste. Then, the paste was dropped onto an Al₂O₃ ceramic plate substrate installed previously with a pair of gold electrodes on the front side to form a sensing film. Afterwards, the as-prepared sensing layer was dried at 80 °C for 2 h, followed by annealing at 500 °C for another 3 h. The gas-sensing properties were examined in a steel-made chamber. A mixture of compressed dry air and target gas flowed at atmospheric pressure and room temperature (ca. 25 °C) on the condition that the relative humidity in the chamber was about 10%. The concentration of target gas was changed by controlling the flow ratio of the two gases using mass flowmeters. The tested gas with given concentration ran into the chamber and the response of sensors were measured. Due to hardly reach a steady equilibrium signal for the sensors exposed to the certain concentration of NO₂ gas, thus every response event was fixed at 200 s. The gas response was defined as S = R_a/R_g when the gas was oxidizing. On the contrary, the gas response was defined as S = R_g/R_a when the gas was reductive. R_a is defined as the resistance of the gas sensor in the air, R_g is defined as the resistance of the gas sensor in the target gas.

3 Results and discussion

3.1 Morphology and structure

Figure 1a shows the wide-angle X-ray diffraction (XRD) patterns for the pure NiO and the 5.0 at.% In–NiO. All the diffraction peaks of the pure sample could be assigned to the face-centered cubic structure of NiO (JCPDS No.47-1049). Furthermore, the radius of In³⁺ is 0.08 nm and the radius of Ni²⁺ is 0.069 nm at a coordination number of 6, obviously, the former is larger than the latter. Therefore, the peaks shift can be attributed to In doping. Figure 1b shows the (111) and (200) diffraction peaks of 5.0 at.% In–NiO is shifted by about 0.05° to the left, which indicates the absence of In-related peaks can be explained by the entry of In³⁺ into the lattice of NiO or the existence of the secondary phase which is below the detection limit of XRD [19]. And the XRD patterns of 5.0 at.% In–NiO were almost same as those of pure NiO, suggesting that the In species might be homogenously dispersed in the NiO matrix.

Fig. 1

a XRD patterns of the pure and 5.0 at.% In-NiO samples. b Comparison of (111) and (200) peaks from XRD patterns. (Color figure online)

Figure 2a, b show the transmission electron microscopy (TEM) images of the pure NiO, and the TEM image of the 5.0 at.% In–NiO is shown in Fig. 2c, respectively. TEM was used to analyze the mesoporous structure of the samples and the dispersion of the dopants on the support. The samples studied with TEM were suspended in ethanol and drop casted on a copper grid. The TEM studies reveal the presence of highly periodic cubic (Ia3d) mesostructures in the NiO samples. The energy dispersive spectra (EDS) (Fig. 2d) recorded on the 5.0 at.% In–NiO sample indicates the presence of the main elements i.e. Ni, O, and In.

Fig. 2

TEM images of the pure (a, b) and 5.0 at.% In–NiO (c). EDS d of the 5.0 at.% In–NiO

X-ray photoelectron spectroscopy (XPS) was conducted to analyze the surface compositions and chemical states of the as-synthesized In₂O₃-decorated NiO, and the results are shown in Fig. 3. Figure 3a exhibits the XPS survey spectrum of the 5.0 at.% In–NiO nanocomposite, which clearly reflects the peaks corresponding to Ni, O, C, and In. Furthermore, the high-resolution scans of Ni 2p, In 3d, and O 1s are shown in Fig. 3b–d. As shown in Fig. 3b, the Ni 2p signal could be divided into five peaks. Consequently, the binding energies could be obtained as 854.0 and 855.8 eV corresponding to the Ni 2p_3/2 peak, and its satellite peak could be observed at 861.0 eV. The peaks corresponding to 872.8 and 869.6 eV were assigned to Ni 2p_1/2, and its satellite, respectively. The Ni 2p_3/2 peak at 854.0 eV was assigned to Ni²⁺, while the one at 855.8 eV was attributed to the presence of Ni³⁺, caused by In doping [21, 27, 28]. In addition, the Ni 2p_1/2 peaks were also attributed to Ni³⁺. The In 3d spectrum (Fig. 3c) was composed of two obvious peaks at 444.7 and 452.4 eV, corresponding to In 3d_5/2 and In 3d_3/2 of In₂O₃, respectively. Hence, combined with XRD results, it indicates that a part of In entered NiO lattice and the other part of In became In₂O₃. Figure 3d, the O 1s binding energy peaks whose shapes are asymmetric are located at 529.6 and 531.5 eV, respectively, corresponding to the lattice oxygen in crystalline NiO and chemisorbed oxygen species on the sample surface. The above results further confirmed that In³⁺ existed as the phase of In₂O₃ in the 5.0 at.% In–NiO sample.

Fig. 3

XPS spectra of the 5.0 at.% In-NiO a a survey scan, b Ni 2p, c In 3d and d O 1s. (Color figure online)

Figure 4 displays the hysteresis loops recorded at relatively high pressures, which indicate large mesoporous diameters, and the pore size distributions calculated by the BJH method also confirmed the mesoporous structure. The detailed information on the pore structure is listed in Table 1. Two well-defined steps of capillary condensation corresponding to two pore size distributions can be clearly observed. The small one is caused by the hard template replica, substantiating the ordered uniform pore structure, and the bigger arises from the piled porosity caused by the aggregation of nanoparticles [29]. The 5.0 at.% In–NiO has the largest specific surface area and pore volume.

Fig. 4

N₂ adsoption–desorption isotherms of the a undoped, b 2.0 at.%, c 5.0 at.% and d 10.0 at.% In–NiO samples

Table 1 Properties of pure and In-doped ordered mesoporous NiO

To further confirm the conversion of In(NO₃)₃ into In₂O₃, the thermal analysis was investigated. Figure 5 displays the thermogravimetric(TG) curve for the ordered mesoporous NiO after the impregnation with 5.0 at.% In(NO₃)₃, which reveals two stages of weight loss [30]. The first stage between 30 and 120 °C was ascribed to water from the In(NO₃)₃ sample, and the weight loss of this stage was about 4%. The second stage between 120 and 600 °C was attributed to the decomposition of In(NO₃)₃ into In₂O₃, and the weight loss of this stage was about 7.5%, which is close to the theoretical value (10.1%) within the error. Therefore, the formation of In₂O₃ phase can be completed by calcination of In(NO₃)₃.

Fig. 5

TG curve of the as-made ordered mesoporous NiO impregnated with 5.0 at.% In(NO₃)₃

3.2 Gas sensing properties

Figure 6 shows the room temperature dynamic response-recovery curves of the as-prepared sensor corresponding to 5.0 at.% In₂O₃-decorated mesoporous NiO as a function of different NO₂ concentrations measured at regular intervals of 200 s. The sample displayed a quick response to NO₂ before reaching saturation. However, it was observed that desorption of NO₂ from the surface of the material was difficult. This could be ascribed to the strong interaction between NO₂ and NiO, which is a common problem for NiO based NO₂ sensors [31]. Moreover, it was found that the 5.0 at.% In–NiO performed much better compared to the other doped and pure samples. For the sake of comparison, sensors based on pure, 2.0 and 10.0 at.% In₂O₃-decorated mesoporous NiO were fabricated separately using the same synthesis procedure and their room temperature NO₂ sensing properties were studied. The results are shown in Fig. S1 as a Supporting Information.

Fig. 6

Responses of the sensor based on the 5.0 at.% In–NiO towards NO₂ at different concentrations at room temperature. (Color figure online)

Furthermore, the reproducibility in sensing performance of the 5.0 at.% composites was investigated and the results have been presented in Fig. 7. Despite the strong adsorption between the NiO and NO₂ which causes a considerably long recovery time in air, the sensor shows reproducible good performance under regular cycles of refreshment. All three cycles of response curves present very similar trends in the presence of a NO₂ atmosphere.

Fig. 7

Resistance changes of the 5.0 at.% In–NiO nanocomposites during three cycles of exposure to 15 ppm of NO₂ for 200 s at room temperature

Furthermore, good selectivity is a desirable characteristic of any sensing device. Hence, the gas response was examined in the presence of various target gases at room temperature, as shown in Fig. 8. The possible interferential gases include carbon monoxide (CO), ammonia(NH₃), ethanol(C₂H₅OH), formaldehyde(HCHO), methanol (CH₃OH), toluene (C₇H₈), and acetone (CH₃COCH₃). From the results, it was clear that the sensitivity towards NO₂ was the highest, despite maintaining a sixfold concentration of other interfering gases. Consequently, 5.0 at.% In₂O₃-decorated mesoporous NiO showed excellent selectivity towards NO₂ at room temperature.

Fig. 8

Responses of the 5.0 at.% In₂O₃-decorated ordered mesoporous NiO to NO₂ (15 ppm), CO (100 ppm), NH₃ (100 ppm), C₂H₅OH (100 ppm), CH₃OH (100 ppm), HCHO (100 ppm), C₆H₅CH₃ (100 ppm), CH₃COCH₃ (100 ppm) at room temperature

The sensor’s stability was examined at room temperature for a month, and the result is given in Fig. 9. It indicated that the long term stability was also excellent.

Fig. 9

The stability of the sensor based on 5.0 at.% In–NiO monitored 15 ppm NO₂ at room temperature for 30 days

3.3 Investigation of gas-sensing mechanism

At first, the sensing layer was composed of NiO and In₂O₃. The main body of sensing layer was NiO, and In₂O₃ in small amounts was present on the surface of NiO. Hence, the sensing layer as a whole showed the characteristics of NiO, the role of NiO was to form p-type semiconductor. And the role of In₂O₃ was to improve the surface activity and adjust the surface hole concentration of NiO, which was derived from the following specific mechanism analysis.

As a typical p-type semiconductor oxide, the widely accepted sensing model of NiO, is based on the modulation of the hole accumulation layer through absorption and desorption of gas molecules [9, 32, 33]. In air, the conduction electrons of NiO were captured via oxygen adsorption on the surface of NiO resulting in the formation of a hole accumulation layer. When the NiO is exposed to oxidizing gases (e.g. NO₂ in our case), the surface reaction can be expressed by the following Eqs. (1) and (2) [18, 34]:

$${\text{N}}{{\text{O}}_2}\left( {{\text{gas}}} \right)+{{\text{e}}^ - } \leftrightarrow {\text{N}}{{\text{O}}_2}^{ - }\left( {{\text{ads}}} \right)$$

(1)

$${\text{N}}{{\text{O}}_2}\left( {{\text{ads}}} \right)+{{\text{O}}_2}^{ - }+2{{\text{e}}^ - } \leftrightarrow {\text{N}}{{\text{O}}_2}^{ - }\left( {{\text{ads}}} \right)+2{{\text{O}}^ - }\left( {{\text{ads}}} \right)$$

(2)

Equation 1 shows NO₂ molecules capture electrons from the NiO surface, thereby increasing the hole concentration in the hole accumulation layer of NiO. Since hole is a major charge carrier in p-type semiconductor, their increasing concentration will lead to a decrease in the NiO’s resistance. Equation 2 shows adsorbed NO₂ molecules can react with the absorbed oxygen species on the surface of NiO, which consumes more electrons.

The enhanced NO₂ sensing can be attributed to the following aspects. Firstly, as previously mentioned, the specific surface area increased with raising In₂O₃ content when the modification amount of In₂O₃ was not too high. This means In doping improved the surface activity of the sensing material, which is favorable for the adsorption and penetration of NO₂. Thus, the surface reactions of Eqs. (1) and (2) are facilitated. Therefore, the enhanced NO₂ sensing can be partly explained by the increase of the specific surface area. Secondly, many previous studies have reported that the formation of p–n junctions can enhance the sensor response through changing the width of space-charge layer [35]. The effective electronic interaction of the p–n heterojunction actually occurred between the NiO matrix and In₂O₃ nanoparticles. As exhibited in Fig. 10, provided the Fermi level of In₂O₃ is higher than that of NiO [36], electrons will continue to flow from n-type In₂O₃ to p-type NiO, while the holes will flow in the opposite direction till an equilibrium is reached [37, 38]. Hence, the formation of nanoscale p-n junctions will lead to the creation of the hole depletion layer underneath the n-type In₂O₃, resulting in the narrowing of the hole accumulation layer for effective conduction. Because the In₂O₃ content in In-NiO is very low, the dominant NiO phase should determine the sensor resistance. Thus, the shrink of hole accumulation layer leads a higher sensor resistance in air. When the In–NiO sensors are exposed to NO₂, the hole will be generated by the process that NO₂ molecules capture the electrons on the surface of the composite, which will decrease height barrier of p–n junction, expand the hole accumulation layer and widen the conduction channel of NiO. Consequently, the conductivity of In–NiO is significantly increased and a high response is obtained. Further studies on deep understanding of In₂O₃-decorated NiO for gas sensing are still needed in view of the fact that there are few reports on this subject.

Fig. 10

Schematic representation of band configuration at the interface of the In₂O₃/NiO nanoheterostructure

4 Conclusions

The ordered mesoporous NiO was prepared by utilizing the of mesoporous KIT-6 silica as a hard template in the nano-casting process followed by In doping. Consequently, the In₂O₃-decorated ordered mesoporous NiO was synthetized successfully. The results of gas sensing measurement manifested the sensor based on the 5.0 at.% In₂O₃-decorated NiO possessing significantly enhanced gas response for 3–15 ppm NO₂ at room temperature. As compared to other In₂O₃-decorated and pure samples of NiO, this showed the best performance. In addition, it presented good selectivity and acceptable reproducibility towards NO₂ sensing. The enhanced NO₂ sensing properties of the mesoporous In₂O₃-decorated NiO sensor is attributed to its higher specific surface area and the formation of p–n junction which modifies the carrier concentration by In doping. This method can be used as an effective strategy to design metal-oxides-based gas sensors with enhanced gas sensing properties.