Introduction

Chemoresistive sensor based on metal oxides has been attracted for ammonia gas (NH3) [1]. NH3 can be indoor air pollutants which are widely used in various industries such as food industry, chemical plant, agriculture, chemical industry, and environmental monitoring [2,3,4]. The toxic gas (NH3) exposure can cause irritation to the skin, eyes, nose, mouth, and lungs [2, 5]. The acute exposure to high concentrations of ammonia causes headache, vomiting, dyspnea, pneumonedema, and even death. According to the occupational safety and health administration (OSHA) directed, the ammonia in the work place is limited to 50 ppm [6] .Consequently, the essential need of reliable and real ammonia gas sensor is urgent [1]. Therefore, it was very important to develop an ammonia sensor with fast response time, high selectivity, low power, and low detection limit, cost at room temperature for practical applications.

Due to its multifunctional characteristics, spinel-type materials, having a general formula of AB2O4, have shown great promise in supercapacitor applications [7], gas sensing [8], and lithium-ion batteries [9]. Recently, the spinel-structured materials have been used for the design of gas sensing material with improved sensing response and selectivity. CuCo2O4 [10], ZnCo2O4 [11], and MgCo2O4 [12] are some of the most widely investigated gas sensors for the detection of toxic gases. One such interesting spinel type is NiCo2O4, which offers more active sites suitable for several suitable applications. However, for the low sensitivity and poor response and recovery of individual NiCo2O4, nanostructure poses a serious challenge in using them for gas sensing applications [13]. Until date, some conductive materials have been used to improve the gas sensing properties of the NiCo2O4 nanostructured materials. Wu J et al. used rGO as a composite material with NiCo2O4 to get the improved gas sensing performances. In this perspective, graphene-based materials would be a better choice as composite, which would enhance the gas sensing properties of NiCo2O4, without affecting its primary properties [14].

Graphene, a unique two-dimensional material (2D), has attracted considerable attention of technologists in a variety of applications especially in the charge transport area, because of its excellent electronic conducting properties [15] with extremely high electron mobility [16]. Pristine graphene has several complications in preparing and the reduced graphene oxide (rGO) is its nearest analogue, which can be prepared by solution processing techniques. Recent investigations revealed that the use of reduced graphene oxide with mixed metal oxide nanoparticles can improve the gas sensing performance in terms of high sensitivity at room temperature [17].

In the present study, we have prepared the NiCo2O4 and NiCo2O4-rGO nanocomposite for the detection of NH3. After careful investigation of structural and morphological properties, the NiCo2O4 and NiCo2O4-rGO nanocomposite formations were confirmed. The as-prepared samples were than used to construct the gas sensors and tested against three different gases. The constructed sensors show high selectivity against NH3 gas. The NiCo2O4-rGO nanocomposite sensor shows better sensing performances along with fast response and recovery times. The stability of the constructed sensor is also studied and the obtained results were discussed in detail.

Chemicals

All the reagents were purchased commercially and used as without further purification. Nickel chloride hexahydrate (NiCl2·6H2O), cobalt chloride hexahydrate (CoCl2·6H2O), graphite powder, hydrogen peroxide, sulfuric acid, potassium permanganate, sodium nitrate, hydrochloric acid, cetyl trimethyl ammonium bromide (CTAB), and urea were purchased from Merck, India. All experiments were performed using deionized (DI) water with an electrical conductivity > 18 MΩ (S/m).

Preparation of reduced graphite oxide nanosheet

Graphene oxide (GO) was prepared by modified Hummer’s method and it is reduced by the following procedure to get reduced graphene oxide (rGO) [16]. One gram of as-synthesized GO powder was dispersed in 50-ml DI water using bath sonicator for 2 h and then 20 ml of ammonia solution was drop wise added into the solution. Then the solution was transferred in a 100-ml Teflon-lined autoclave and heated at 160 °C for 12 h in hot air oven. After that, the autoclave was allowed to cool naturally in the room temperature. The black precipitate was collected and washed with ethanol and DI water several times. Finally, the collected sample was dried in a hot air oven at 60 °C overnight [18].

Preparation of NiCo2O4 nanorod

Hydrothermal synthesis of nanorod NiCo2O4 samples was reported in our previous work [18]. In a typical synthesis, 30 (mmol) NiCl2·6H2O and 60 (mmol) CoCl2·6H2O (Ni and Co, molar ratio 1:2) were dissolved in 100 ml of deionized water and added 30 (mmol) CTAB dissolved in 100 ml of water. The mixture was stirred for 1 h to form homogeneous solution. Ninety (mmol) of solid urea was added to this solution and stirring was continued for three more hours to achieve complete homogeneity. The resulting solution was transferred into Teflon-lined stainless steel autoclaves of 200-ml capacity. The reactions in these autoclaves were carried out at120 °C, in an electrical oven for 48 h. The autoclaves were cooled to room temperature and the light pink–colored precipitates were separated by centrifuging at 4000 rpm. The product was repeatedly washed with deionized water, mixture of water and absolute ethanol, and finally with absolute ethanol for three times. The obtained products were dried in an oven at 60 °C for 12 h and calcined at 350 °C for 3 h.

Preparation of NiCo2O4-rGO composite

NiCo2O4-rGO composite was prepared by wet impregnation method [19]. 2:1 weight ratios of NiCo2O4 nanorod were added separately with rGO nanosheets in to 15 ml of ethanol. The resulting solution was continuously stirred and subsequently heated at 60 °C to evaporate the solvent. Then, the obtained NiCo2O4-rGO composite was dried for 6 h, collected, and stored for further processes.

Characterization techniques

The phase purity and crystal structure of the as-prepared samples were characterized by powder X-ray diffraction (XRD) using Rigaku D/Max Ultima III X-ray diffractometer with a Cu Kα radiation source (0.154 nm) operated at 40 kV. Fourier-transform infrared (FT-IR) spectra of KBr powder–pressed pellets were recorded by Bruker a NEXUS 470 in the range of 4000 to 400 cm−1. The surface morphology of obtained samples was visualized by field emission scanning electron microscopy (FESEM, Hitachi S-4800) and high-resolution transmission electron microscopy (HRTEM, JEM-2100). Nitrogen adsorption-desorption isotherm measurements were performed on AUTOSORB-1-MP to study the specific surface area.

Preparation and gas sensor fabrication

A 0.5 g of the NiCo2O4-rGO nanopowder was grounded well in an agate mortar for 10 min and then samples were mixed with 0.5 ml of the deionized water and subsequently, grounded for another 10 min to obtain a black slurry with good stability. The ceramic tube (diameter 1 mm and length 4.0 mm) was furnished with gold signal electrodes and homogeneous black paste with the sensor element was coated on a ceramic tube. Further heating was done to remove excess water molecules and improve contact between NiCo2O4-rGO with gold electrode. The sensor device was placed in a chamber. A constant voltage of + 2 V was applied to the sensor device. The gas sensing measurements were carried out using Keysight source meter. The source meter was coupled to the computer via the RS 232 interface to record the change in resistance with respect to time. The sensing measurement cycle is as follows: nitrogen (N2) 150 sccm was introduced as the carrier gas, The change in resistance R and the resistance of the sensor was calculated [20].

$$ S=\mathrm{Rg}/\mathrm{Ra} $$

where Rg and Ra were the resistance of the sensor in NH3 and air atmospheres, respectively

Results and discussion

The XRD patterns of rGO, NiCo2O4, and NiCo2O4-rGO nanocomposite are shown in Fig. 1. The successful reduction of GO into rGO is confirmed by the peak observed around 25.58°, which is attributed to the (022) plane [21]. The NiCo2O4 and NiCo2O4-rGO composite samples produced several diffraction peaks around 19.20°, 31.36°, 36.96°, 38.63°,44.91°, 55.71°, 65.29°, and 77.42°, which are indexed to (111), (220), (311), (222), (400), (422), (511), and (440) crystal plans, matching with the standard values (JCPDS no. 73-1702). Absence of additional peaks reveals that there were no impurities or other crystal phases present in the sample. The similarity in the diffraction peaks of NiCo2O4 and NiCo2O4-rGO composite samples indicate that the incorporation of rGO with NiCo2O4 did not affect the crystal structure properties of NiCo2O4. Also the absence of rGO diffraction peak in the NiCo2O4- rGO composite sample is attributed to the less amount of rGO when compared with the NiCo2O4. The crystallite size values calculated using the Debye Scherer formula [16] are 43 and 40 nm for NiCo2O4 and NiCo2O4-rGO samples, respectively. The variation in the crystalline size after addition of rGO is an indication of NiCo2O4-rGO composite formation. In order to investigate further the rGO, NiCo2O4 and NiCo2O4-rGO composite samples were subjected to FT-IR spectroscopy analysis.

Fig. 1
figure 1

Typical XRD patterns of all samples

Figure 2 shows FT-IR spectra of rGO, NiCo2O4, and NiCo2O4-rGO composite samples. The FT-IR spectrum of rGO sample displays two peaks in the range of 3500–4000 cm−1 corresponding hydroxyl and amine groups. The amine group could possibly be included from the ammonia, which is added for the reduction of GO into rGO in the hydrothermal reduction process. The peaks at 1196 and 1562 cm−1 are attributed to C–OH and C=O groups, respectively, thereby confirming the formation of functional groups rGO [22, 23]. The formation of NiCo2O4 is indicated by two strong peaks around 561 and 654 cm−1, which can be assigned to the stretching vibration of the Ni–O and Co–O bond, and is reported elsewhere [15]. The NiCo2O4-rGO composite sample contains the FT-IR fingerprints corresponding to both the NiCo2O4 with rGO, which further confirms the composite formation.

Fig. 2
figure 2

FT-IR spectra of rGO, NiCo2O4, and NiCo2O4-rGO

The as-prepared NiCo2O4 and NiCo2O4-rGO composite samples were then subjected to FESEM analysis in order to investigate the morphology. As presented in Fig. 3(a), (d), both the samples display aggregated nanorod structure. The observed morphology is highly porous, because the aggregation took place in particular centers. In others words, the nanorods are not aggregated one over the other, instead, they are attached to a particular center and formed a bunch-like structure. It is because of this interesting morphology, the samples are highly porous in nature, which would benefit the adsorption and desorption of gas. In addition, the change in morphology is observed in the case of NiCo2O4-rGO composite, in which the big bunch-like aggregated nanorod structure was split into smaller aggregations. This change in morphology is attributed to the incorporation of NiCo2O4 with rGO [13].

Fig. 3
figure 3

Low and high-resolution FESEM images of NiCo2O4 (a, c) and NiCo2O4-rGO (b, d) samples

After that, the detailed morphology of the as-prepared nanocomposite was investigated through high-resolution transmission electron microscopy (HRTEM) as represented in Fig. 4(a). The NiCo2O4 bunch-like nanorod can be seen uniformly distributed on rGO nanosheet. It further revealed the typical bunch-like nanorod structural characteristic of NiCo2O4 anchored on graphene oxide sheets in the composites. Figure 4(b) shows the HRTEM image of an individual NiCo2O4 nanorod with a lattice spacing of 0.23 nm well-matched the (311) lattice planes of cubic spinel NiCo2O4, respectively. These results further confirmed the existence of NiCo2O4, which was in agreement with the above-revealed XRD results.

Fig. 4
figure 4

(a, b) HRTEM image of NiCo2O4-rGO nanorod

This porous morphology is further investigated by nitrogen adsorption and desorption kinetics, in order to understand its role in the gas sensing performances. The specific surface area and the pore size distribution are the vital factors in understanding the gas sensing performances of the as-prepared sample. The nitrogen adsorption-desorption isotherms of NiCo2O4 and NiCo2O4-rGO composite samples were measured at 77 k using N2 adsorbent and presented in Fig. 5(a), (b) [17]. The BET isotherms of NiCo2O4 and NiCo2O4-rGO sample exhibit type IV characteristics, which indicate that the sample contains a certain amount of mesoporous structure. The calculated BET-specific surface area for NiCo2O4 and NiCo2O4-rGO samples are 5 and 9 m2g−1, and their corresponding pore volumes are 0.957 and 0.960 cm3 g−1, respectively [24].

Fig. 5
figure 5

(a, b) N2 adsorption-desorption isotherms of the NiCo2O4 and NiCo2O4-rGO nanocomposite

Gas sensing performance

The gas sensing performances of NiCo2O4 and NiCo2O4-rGO composite samples were evaluated against NH3 gas. Figure 6(a) shows the comparison of sensing response (S) between NiCo2O4 and NiCo2O4-rGO nanocomposite with 100 ppm NH3 at room temperature. It is notable that the response (1.068) of NiCo2O4-rGO nanocomposite to 100 ppm NH3 gas is greater than that of pure NiCo2O4 sensing response (1.029). The active sensing responses of NiCo2O4-rGO nanocomposite–based gas sensors by varying the NH3 concentration 10 to 100 ppm at room temperature are shown in Fig. 6(b). It is clear that the NiCo2O4-rGO composite sensor exhibits better response intensity with increasing gas concentrations and the sensor can completely recover to its initial state. Figure 6(c) shows the selectivity of the NiCo2O4-rGO nanocomposite investigated against by various toxic gases such as ammonia (NH3), ethanol (C2H6O), and acetone (C3H6O). In comparison, it is clear that the constructed NiCo2O4-rGO composite–based gas sensor is highly selective against NH3 gas and it is attributed to the lower molecular size (0.32 nm) and low dipole moment (1.4 nm) of the NH3, when compared with the other two gases (the molecular size and dipole moment of ethanol (C2H6O) and acetone (C3H6O) are 0.44 nm and 0.66 nm, and 1.69 nm and 2.88 nm, respectively) the stability of the NiCo2O4-rGO composite–based gas sensor is investigated by admitting 100 ppm NH3 gas over 10 cycles. In sensor, the stability and repeatability are important parameters for gas sensor optimizations. In Figs. 6(d) and 7, a slight variation in higher resistance point, for each cycle, may be due to the poor adhesion between sensing layer and subtract [25]. Figure 8 shows the response-recovery times of NiCo2O4-rGO samples (10 to 100 ppm) NH3. The NiCo2O4-rGO sample exhibits fast response (57 s) time and recovery (185 s) time at 20 ppm as compared with other contraction. This improved performance is attributed to the heterojunction formed between NiCo2O4 and rGO, in which the rGO acted as 1D conductive network for the electron transfer process.

Fig. 6
figure 6

(a) Comparison of response-recovery time curve of NiCo2O4-rGO and NiCo2O4 to 100 ppm NH3 gas. (b) Gas sensing response of the NiCo2O4-rGO to different concentration of NH3. (c) The same sensor with several other gases to NH3. (d) Stability of the NiCo2O4-rGO sensor to NH3

Fig. 7
figure 7

Response cycle stability of the NiCo2O4-rGO sensor to NH3

Fig. 8
figure 8

Response-recovery time curve of NiCo2O4-rGO

Figure 9 shows the sensor response versus different NH3 gas concentration in the range of 10–100 ppm. The correlation coefficient R2 of the fitting curve in the range is to be 0.9496, indicating good linearity in the sensing measurement. This linear behavior is suitable for approximation of NH3 in practical sensing application. Furthermore, from the above observation, it can be concluded that the availability of active sites for adsorption on the sensor material decreased with the increase in NH3 concentration [26].

Fig. 9
figure 9

Linear dependence of response of NiCo2O4-rGO sensor on different concentration of NH3 at room temperature

Gas sensing mechanism

Semiconductor metal oxide–based sensors work on the principle of the change in resistance corresponding to the reaction between gas molecules and the sensitive surface. This involves gas adsorption, surface reaction, and desorption processes [20]. Initially, when the NiCo2O4 surface is exposed to air atmosphere, the air molecules are adsorbed into the NiCo2O4 surface and capture electrons from the NiCo2O4 surface. This results in the formation of a hole accumulation layer (HAL) on the surface of NiCo2O4, which eventually results in the decrease of resistance. Upon exposure to the NH3 (strong reducing agents), the pre-adsorbed oxygen ion on the NiCo2O4 surface reacts with NH3 and the resistance of the sensing element increases, which obeys the p-type behavior of NiCo2O4.The sensing performance of NiCo2O4-rGO nanocomposites can also be explained using schematic diagram shown Fig. 10,except the fast response and recovery. It is caused by the hybridizing rGO, which triggers the fast electron transfer between the sensing element and the electrode [27]. From the above observations, it is clear that the room temperature NH3 sensing performance of NiCo2O4-rGO composite gas sensor can be attributed to two key factors: (i) the mesoporous structure of NiCo2O4-rGO nanocomposite which provided surface accessibility for the interaction with the analyte gas, and (ii) the fast charge transfer initiated by the rGO, which resulted in better response and recovery times.

Fig. 10
figure 10

Schematic diagram of the gas-sensing test

Conclusions

In conclusion, a cost-effective NiCo2O4-rGO nanocomposite–based gas sensor was developed and its room temperature NH3 gas sensing properties were demonstrated. Before the construction of gas sensor, the nanocomposite sample was characterized by several analytical techniques such as XRD, FT-IR, FESEM, HRTEM, and BET analysis. The better performance of NiCo2O4-rGO nanocomposite–based gas sensor is due to its mesoporous structure and fast charge transfer initiated by the rGO. With high selectivity, sensitivity, and stability, our sensor can be used in practical applications.