Introduction

With the acceleration of modern industrial processes, many toxic gases are being released directly into the air. Nitrogen dioxide (NO2), a common pollution emission released by fossil fuel combustion, power plants, and automotive engines, is harmful to the environment and human health [1, 2]. To effectively monitor air quality and protect human safety, it is of great significance to develop NO2 gas sensors with high sensitivity, low detection limit, and high selectivity.

Metal oxide semiconductors, such as SnO2, ZnO, TiO2, and CuO, are extensively used as gas sensors to detect hazardous gas leaks and monitor air pollutants. Among these gas sensing materials, ZnO has been widely studied for its stable physicochemical properties, controlled morphology and microstructure, optical activity, nontoxicity, and good gas-sensitive properties. It is an important direct wide band gap (~ 3.37 eV) semiconductor metal oxide after SnO2. However, the performance of most reported gas sensors based on ZnO is not applicable at room temperature. This kind of gas sensor requires sufficient reaction activation energy between the target gas molecules and the oxygen ions adsorbed on the surface. The working temperature is usually higher than 200 °C, which limits the application [3,4,5].

Graphene has attracted extensive interest due to its high surface area (2630 m2g−1), electron mobility (200,000 cm−1 V−1S−1), electrical conductivity (5000 W/mK), etc. [6] The high electron mobility and excellent electrical conductivity create extremely low detection limits in gas sensing [7, 8]. Tuning the semiconductor properties by adding graphene is a promising strategy for the fabrication of room temperature sensing materials. Although the sensors prepared from graphene can achieve the detection of NO2 gas at room temperature, the van der Waals force between graphene and gas is weak, which will directly lead to the drawbacks of gas sensors from pure graphene [9,10,11]. Therefore, reduced graphene oxide (rGO), a form of graphene produced by the reduction of graphene oxide, which contains many defects, is preferred for sensing studies [12,13,14]. rGO has more dangling bonds and defects than graphene, which is good for the adsorption of gas and the formation of heterojunction with metal oxides.

To obtain a room-temperature NO2 sensor based on metal oxide/graphene nanocomposites with low sensitivity, short response time, and response recovery time, many recent efforts have been devoted to the controllable synthesis of metal oxides with confined sizes and desired morphologies in nanocomposites [15,16,17]. For example, rGO-conjugated Cu2O nanowire monocrystals were hydrothermally fabricated. They exhibited high sensitivity toward NO2 at room temperature owing to the unique nanowire-assembled mesoporous structures. Wang et al. used a microwave-assisted hydrothermal method to uniformly load 3-nm SnO2 nanoparticles on rGO [18, 19]. The composite showed significant gas-sensitive properties, especially for NO2, with a high response, fast response, and good recovery. It has six times the response of SnO2 at 75 °C for 350 ppb NO2. The response to 50 ppb NO2 still has a response, so the composite material has many advantages over a single gas-sensitive material.

Herein, ZnO/rGO with different rGO ratios was synthesized by a hydrothermal method. The characterization results showed that rGO influenced the agglomeration of ZnO particles. ZnO and rGO were hybridized to construct heterojunctions and nanostructures. Among the ZnO/rGOs, ZnO NPs-1% rGO exhibited the best room-temperature NO2 gas-sensitive performance. The excellent gas-sensitive properties of the composites are derived from the strong synergistic effect between the smaller particles of ZnO and rGO. In addition, the mechanism of sensing NO2 gas by the ZnO/rGO heterostructure is discussed.

Experimental section: Materials and methods

Chemicals

Commercial grade ZnO dispersion was purchased from Zhejiang Zhongliang Chemical Reagent Co., Ltd. Graphite powder was purchased from Nanjing Xianfeng Nanomaterials Technology Co., Ltd. Potassium permanganate, hydrogen peroxide, and zinc nitrate were purchased from Shanghai Lingfeng Chemical Reagent Co., Ltd. Absolute ethanol and hydrochloric acid were purchased from Aladdin Reagent Co., Ltd. The corresponding solutions were prepared using distilled water. All required replacement parts were analytically pure and used directly without any further purification.

Preparation of graphene oxide

Graphene oxide (GO) nanosheets were synthesized via Hummer’s method from graphite powder. First, 3.0 g of graphite powder was dissolved in 69 mL of sulfuric acid and vigorously stirred for 10 min. Then, 1.5 g of sodium nitrate was added to the mixture to obtain a homogeneous solution, which was kept at 0 °C during this process. Then, 9.0 g of potassium permanganate was added to the mixture, vigorously stirred for 3 h, and continuously stirred for another 6 h at 35 °C. Then, 550 mL distilled water and 20 mL H2O2 were added and continuously stirred for 2 h. Finally, the samples were repeatedly washed with distilled water to pH = 7. To completely remove metal ions, the sample was dialyzed for 2 weeks. A graphene oxide/water dispersion (5 mg/mL) was made from freeze-dried GO powder for further use [20, 21].

Synthesis of ZnO/rGO

One gram of nano ZnO was added to 20 mL of deionized water with stirring; desired amounts of graphene oxide dispersion were added dropwise to the ZnO dispersion, forming a homogeneous ZnO/GO dispersion with 0 %, 0.5 %, 1 %, 2 %, and 3 % GO in mass ratio. The dispersion was treated with ultrasonication for 0.5 h. Then, it was put into a Teflon-lined stainless steel autoclave and heated at 180 ℃ for 6 h. After that, GO transformed to rGO, and the ZnO/rGO composite formed simultaneously. The ZnO/rGO was centrifuged, washed with distilled water and ethanol, and freeze dried. Samples were labeled ZnO NPs and ZnO NPs-x% rGO (x = 0.5, 1, 2, and 3) according to the GO mass ratio.

Material characterization

X-ray diffraction (XRD) patterns were recorded on a diffractometer (D8-ADVANCE, Bruker, Germany) with Cu Kα radiation (λ = 1.5418 Å) to analyze the phase and crystallinity of the composite materials. The morphology and microstructure of the as-prepared samples were characterized by field-emission scanning electron microscopy (FESEM, S-4800, Hitachi Ltd., Japan) and high-resolution transmission electron microscopy (TEM, JEM-2100, JEOL, Japan). X-ray photoelectron spectroscopy (XPS) was conducted on a spectrometer with an Al Kα X-ray source (Thermo ESCALAB 250 Xi, USA). The BET surface areas of the materials were determined using a nitrogen adsorption analyzer (Micromeritics ASAP 2020, Merrick, USA). Raman spectra (HORIBA, LabRAM HR, France) were recorded on a laser confocal microscopy Raman spectrometer. Ultraviolet visible (UV–Vis) diffuse reflectance spectroscopy (DRS) was performed using a UV–Vis spectrophotometer (Shimadzu UV-2600, Japan) and BaSO4 as the reflectance standard. Photoluminescence spectra (PL) were recorded on a fluorescence spectrometer (HORIBA, Quanta Master 8000, Canada).

Gas sensing measurements

The ZnO/rGOs were first mixed with a small amount of deionized water to form the paste in an agate mortar. Then, it was coated on an alumina substrate with two Ag-Pd electrodes to form a sensing film. The samples were aged in a muffle furnace at 200 ℃ for 6 h. The gas sensing property was measured on a CGS-4TPS system (Alite Technology Co. Beijing, China). Before the measurements, the chamber was purged with pure dry air until stable resistance was obtained, which was defined as the background gas. Then, in a measurement, the sensors were alternately exposed to different gases for the response process and to the pure dry air flow for the recovery process. The sensitivity (S) of the sensor is defined as Ra/Rg, where Ra is the initial resistance in dry air, and Rg is the resistance of the sensitive film after exposure to an NO2 atmosphere. For the p-type sensor, the resistance is decreased upon NO2 exposure (S > 1). The response time is defined as the time required until 90% of the whole response is reached, while the recovery time denotes the time needed until 90% of the baseline sign is recovered.

Results and discussion

Characterization

The structure and phase of the samples were examined by powder X-ray diffraction (XRD). As shown in Fig. 1, the XRD pattern of GO shows an intense peak (001) at 10.6° with a d-spacing of 0.8336 nm. After the reduction of GO to rGO by hydrothermal treatment, the peak in GO disappears, and a broad peak appears. The broad peak suggests a poorer ordering of the graphene sheets along their stacking direction than that of GO. The variation in diffraction peaks revealed successful transformation from GO to rGO via the same hydrothermal treatment as the synthesis of ZnO/rGO hybrids. The XRD patterns of ZnO and ZnO/rGO clearly demonstrate the stability of the ZnO nanoparticles toward hydrothermal treatment because the representative sharp peaks agree well with those from the standard JCPDS card (No. 79–206). In addition, the diffraction peaks of rGO were not observed among the ZnO/rGO samples, which may be related to the low doping concentration of rGO.

Fig. 1
figure 1

XRD patterns of GO, rGO, and ZnO/rGO

Full size image

The morphology and microstructure of the ZnO and ZnO/rGOs were characterized by SEM, TEM, and HRTEM. Figure 2 displays the SEM and TEM images of pure ZnO nanoparticles and ZnO NPs-1% rGO. The size (20–50 nm) and shape of pure ZnO are not perfectly uniform (Fig. 2a and b). The diameter of the ZnO nanoparticles remained unchanged after thin film–like rGO wrapped the ZnO nanoparticles through hydrothermal synthesis (Fig. 2c and d). However, the sample shows a kind of lamellar agglomeration morphology (Fig. 2c). This is due to the interaction of ZnO with hydrophilic GO sheets before hydrothermal reduction. The HRTEM images (Fig. 3) of ZnO NPs-1% rGO display more details of the materials. The lattice fringes with spacings of 0.26 nm and 0.19 nm (Fig. 3d) are attributed to the (002) and (102) crystal planes of hexagonal ZnO. This is consistent with the XRD results. In the HRTEM images, we also observed that ZnO is well combined with rGO (Fig. 3c).

Fig. 2
figure 2

SEM of ZnO NPs (a); TEM of ZnO NPs (b); SEM of ZnO NPs-1% rGO (c); TEM of ZnO NPs-1% rGO (d)

Full size image
Fig. 3
figure 3

a–d HRTEM image of ZnO NPs-1% rGO

Full size image

The XPS spectra of the samples (Fig. 4) show the presence of carbon, oxygen, and zinc elements. In the Zn2p region (Fig. 4b), Zn2p1/2 and Zn2p3/2 peaks at binding energies of approximately 1045.0 and 1022.0 eV were observed, which are ascribed to spin–orbit coupling of Zn2p with a ΔE (spin–orbit splitting) of approximately 23.1 eV. It is worth noting that the Zn2p peaks are shifted to higher energy positions with the doping of rGO. The formation of a higher energy for Zn is due to electrostatic repulsive interactions among O with a negative charge in the lattice. The O1 s signals (Fig. 4c) are quite asymmetric and can be decomposed into three Gaussian peaks at 530, 531, and 532.33 eV, corresponding to different oxygen species in the sample. The most prominent lowest binding energy peak is attributed to the lattice oxygen (OL) of ZnO, and the middle binding energy peak is linked to surface hydroxyls in the oxygen-deficient regions (OH), whereas the highest binding energy peak is associated with the chemisorbed or dissociated oxygen species on the surface of the sample. In the C1 s region (Fig. 4d and e), the spectra of GO, rGO, and ZnO/rGO can be subdivided into three peaks at 284, 286, and 289 eV corresponding to the C–C/C = C (sp2/sp3 carbon atom), C-O (epoxy/alkoy), and C = O/C–O–C (carbonyl) functional groups of rGO. After the hydrothermal treatment of ZnO nanoparticles with GO dispersion, the intensity of the C-O peak decreases, and the C–C peak increases in all samples.

Fig. 4
figure 4

XPS survey spectra of GO, rGO, ZnO, and ZnO/rGOs (a); magnified Zn 2p core level (b); deconvoluted high-resolution O1 s spectrum (c); deconvoluted high-resolution C1 s spectrum of (d) and (e)

Full size image

The physical nitrogen desorption curves of the five samples are shown in Fig. 5. According to the IUPAC classification, the adsorption isotherms of the samples in the figure are all type IV. With the increase in the rGO doping ratio, the specific surface area of the composites also increases, and the pore capacity and pore size are changed accordingly (Table 1). The main pore structure may be generated by the agglomerative accumulation of ZnO nanoparticles, and rGO acts as an isolation layer during its drying process, which can effectively prevent serious agglomeration during drying, leading to a high specific surface area. These hollow structures, also shown in SEM and TEM images, promote the diffusion of target gas molecules inside the material and improve the response and recovery ability of the gas sensing materials.

Fig. 5
figure 5

Nitrogen adsorption–desorption isotherms of ZnO and ZnO/rGO composites

Full size image
Table 1 Specific surface area and pore volume
Full size table

Raman spectroscopy is an effective and sensitive technique to characterize carbon-based materials. The ZnO/rGO composites exhibit two intensive peaks (Fig. 6), which correspond to the D and G bands. The D band at approximately 1340 cm−1 originates from the breathing mode of k-point phonons of the A1 g symmetry near the edges of graphitic structures, which increases with increasing defect density (local defects or disorder); the G band at approximately 1600 cm−1 belongs to the E2g mode of sp2-hybridized carbon atoms (stacking nature) of the rGO species. The intensity ratio of the D band to the G band (ID/IG) is usually used as a parameter to evaluate the disorder and defect degree of various carbon materials; that is, the larger ID/IG is, the higher the disordering in the carbon materials. The ID/IG of ZnO/rGO composites is 1.03, 1.01, and 1.02 for ZnO NPs-1% rGO, ZnO NPs-2% rGO, and ZnO NPs-3% rGO, respectively, showing the similar in-plane crystallite size of ZnO/rGO samples [22]. The second-order D(2D) bands at approximately 2667 cm−1 and D + G bands at approximately 2918 cm−1, belonging to rGO, were also observed in rGO and ZnO/rGO samples, further verifying the formation of rGO in the synthesis [23].

Fig. 6
figure 6

Raman spectra of ZnO and ZnO/rGOs

Full size image

Figure 7a shows the UV–Vis absorption of the materials. Pure ZnO shows a strong band-to-band transition absorption peak at ~ 380 nm. No apparent absorption is observed in the visible light region. As a zero-gap semiconductor, graphene is a good light-absorbing material, so introducing rGO into ZnO can greatly enhance the adsorption in the visible light region and generate an electronic energy level below the CB (conduction band) of ZnO. Meanwhile, the visible light absorption ability of ZnO/rGO can be effectively enhanced with the increase in the amount of rGO. The energy gap (Eg) is calculated with the formula for the near-edge optical absorption of semiconductors:

$${\left(\alpha \mathrm{hv}\right)}^{n}={c}_{0}\times \left(\mathrm{hv}-\mathrm{Eg}\right)$$
(1)

where C0 is a constant, hν is the photon energy, Eg is the energy gap, and n = 2 for the direct band gap semiconductor [24]. As shown in Fig. 7c–g, the Eg value changed from 3.179 eV (ZnO NPs) to 3.212 eV (ZnO NPs-3% rGO) as the doping amount of rGO increased.

Fig. 7
figure 7

UV–Vis absorption spectroscopy (a) and PL spectra (b). Plot of (αhν)2 vs. Hν (c–g) of ZnO and ZnO/rGOs

Full size image

The photoluminescence (PL) spectra provide the separation and recombination information of the photoinduced electrons and holes in the materials. Figure 7b shows the PL spectra of pure ZnO and ZnO/rGO with an excitation wavelength of 325 nm at room temperature. All five materials showed a strong UV emission at a wavelength of 380 nm and a broad green emission band at 580 nm. The peak emission at 380 nm is the recombination of the charge carriers of materials, and the green emissions centered at 580 nm are from the surface defects of ZnO. In addition, the intensities of all ZnO/rGOs are weaker than those of pure ZnO, indicating that the fluorescence of the ZnO/rGO composites is quenched more efficiently than that of pure ZnO NPs. This result also suggested a strong interaction between ZnO and rGO, which indicated more separation of photoinduced electrons and holes by the addition of GO.

Gas sensing properties

The sensors were made by depositing the material dispersion on the microelectrode surface. Figure 8 shows the resistance of the ZnO-, ZnO/rGO-, and rGO-based sensors at room temperature. The sensor based on ZnO shows the largest resistance. It behaves almost insulatively at room temperature. As an n-type sensor shows an increase in resistance in NO2 gas, it is difficult to test the sensing performance at room temperature. The best working temperature is usually above 150° [25]. rGO exhibits the lowest resistance because of its excellent electrical conductivity. As a p-type sensor showing a decrease in resistance in NO2 gas, the sensor based on rGO exhibits a weak response [26]. With the combination of ZnO and rGO, sensors based on ZnO/rGO show a medium resistance with p-type sensing. The higher the ratio of rGO in the hybrid materials, the lower the resistance was observed.

Fig. 8
figure 8

Resistance of ZnO and ZnO/rGO composites

Full size image

Gas sensors based on pure ZnO or rGO show no or a weak response in NO2, so only ZnO/rGO-based sensors were tested at room temperature. The sensors were first exposed to a range of NO2 gas concentrations (5 ~ 30 ppm). They exhibit a stable signal with a good p-type response (Fig. 9a). The maximum gas response is achieved at a reduced graphene oxide (~ 1%) content (ZnO NPs-1% rGO). This may be due to the synergistic interaction between ZnO and rGO in nanocomposites, where ZnO nanoparticles are the non-negligible sensing material involved in gas adsorption and interaction with NO2, while rGO with excellent transport capacity provides more electrical channels for the nanocomposites, directing current through the nanocomposites and electrodes to obtain a response and improved sensing. Figure 9b shows the dynamic response recovery curves of the ZnO NPs-1% rGO composites for different concentrations of NO2 gas. The gas response at room temperature is a relatively slow process, but all the samples have a good response, and the baseline can roughly recover to the initial value. As shown in Fig. 9d, the response time and recovery response time of ZnO NPs-1% rGO are 198 s and 144 s, respectively. Figure 9c shows the results of 3 consecutive cycles of ZnO NPs-1% rGO. ZnO/rGO can be adjusted and recover to its initial state without external heat or illumination, and there is no drift in the baseline after recovery.

Fig. 9
figure 9

Response of the sensors based on ZnO/rGOs (a); dynamic response curves of the sensor based on ZnO NPs-1% rGO toward NO2 at concentrations ranging from 0.5 to 30 ppm (b); reproducibility (c), recovery and response time (d) of the sensors based on ZnO NPs-1% rGO; the response curves toward 10 ppm NO2

Full size image

The signal-to-noise ratio method was used to quantify the detection limit (LOD) of NO2 [22]. The LOD can be calculated by Formula (2) below:

$${\text{LOD = 3}} \times \frac{{RMS_{nosie} }}{K}$$
(2)

where RMSnoise is the root mean square noise and “K” is the slope value in the linear fitting of the response (y = 0.585x + 2.022). The sensor noise can be calculated by the relative resistance variation from the average resistance measured. Thirty Ra values were taken to calculate the standard deviation before exposure to NO2 by Formula (3).

$${\text{RMS}}_{nosie} = \sqrt {\frac{{\sum\nolimits_{i = 1}^{N} ( R_{i} - \overline{R} )^{2} }}{N}}$$
(3)

where Ri is the randomly selected Ra value and R is the average resistance of the thirty randomly selected Ra values. In this work (Fig. 10), the RMSnoise of ZnO NPs-1% rGO is calculated to be 0.000869, and the LOD is 4.4 ppb.

Fig. 10
figure 10

The fitting curve of ZnO NPs-1% rGO to 0.5–4 ppm NO2 at room temperature

Full size image

To check the effect of humidity on sensing performance, the sensitivities of the ZnO NPs-1% rGO sensor were investigated at 10%, 25%, 50%, and 75% relative humidity (%RH) under 10 ppm NO2 gas (Fig. 11). The influence of the humidity obviously occurs from 10 to 75%, where there is a relatively obvious variation in the sensor response. Under 75% RH, the response drops to ~ 75% compared with the response under 10% RH.

Fig. 11
figure 11

The dynamic response curves of ZnO NPs-1% rGO to 10 ppm NO2 under various relative humidities at room temperature

Full size image

The cross-sensitivity of the ZnO NPs-1% rGO sensor toward other gases, such as toluene, ethanol, oxygen, ammonia, and n-hexane, at 10 ppm was tested at room temperature (Fig. 12). The ZnO NPs-1% rGO exhibits a high response toward NO2, a weak response to ammonia and oxygen, and almost no response to toluene, hexane, and ethanol gases, suggesting that the ZnO/rGOs have good selectivity to NO2. The remarkable selectivity to NO2 of the ZnO/rGOs could be related to the selective and high NO2 adsorption ability of the material at room temperature.

Fig. 12
figure 12

Responses of the ZnO NPs-1% rGO composites to 10 ppm NO2 and other gases (oxygen, ammonia, ethanol, toluene, n-hexane)

Full size image

In addition, the long-term stability of the ZnO NPs-1% rGO sensor was tested with 5 ppm NO2 for 20 days (Fig. 13). The response of the ZnO NPs-1% rGO gas sensor remained relatively stable (from 4.45 to 4.18).

Fig. 13
figure 13

Stability of the ZnO NPs-1% rGO composites when exposed to 5 ppm NO2 gas at room temperature (25 ℃)

Full size image

A comparison of NO2 sensing performance at room temperature between the present ZnO NPs-1% rGO composites and reported zinc oxide/graphene nanocomposite sensors in recent literature is summarized in Table 2. ZnO NPs-1% rGO composites have good room temperature gas sensing performance for NO2 gas and still show good repeatability and selectivity at low NO2 gas concentrations. It can be used as a gas-sensitive material for NO2 detection at room temperature to meet practical applications.

Table 2 The summary of ZnO/rGO composites
Full size table

Gas sensing mechanism

In this work, ZnO/rGOs showed P-type responses similar to those of rGO. rGO with an outstanding electric conductivity offers a conductive path. However, the gas-sensitive performance of ZnO/rGO composites does not improve monotonously with increasing rGO content. ZnO played a non-negligible role in contributing to the high responses of the materials. When the ZnO and rGO contact, a unique porous structure is formed, and the specific surface of ZnO/rGO became higher than that of ZnO (Table 1), which facilitated the diffusion of gas [32,33,34]. Furthermore, a p-n heterojunction forms at the interface, which can benefit the sensing performance [35, 36]. As shown in Scheme 1, the electron flows from n-type ZnO to p-type rGO due to the difference in work function (Scheme 1a) [37,38,39], which results in a decrease in the carrier concentration (electron/hole) and an increase in the resistance of ZnO/rGOs. In addition, an internal electrical field is formed. Under air conditions (Scheme 1b and d), oxygen molecules are adsorbed on the surface of the ZnO/rGOs and ionized into oxygen ions (Eq. (3)), capturing electrons from the conduction band of ZnO, which caused an increase in the ZnO electron depletion layer and upward band bending. Upon exposure to NO2 (Scheme 1c and e), the electron transfers from ZnO/rGOs to the adsorbed NO2 (Eq. (4)) [40,41,42,43]. Since the electronegativity of NO2− (ads) is higher than that of O2− (ads), they react to form NO3− (ads) (Eq. (5)). The reduction of free electrons leads to a larger difference between Ec and Ef in ZnO, weakening the internal electrical field. The electrons in rGO will be transferred to ZnO in the NO2 environment to increase the holes of rGO. Beside, electrons of rGO are also transferred to NO2, which further decreases Rg of ZnO/rGOs [44]. Therefore, they exhibit excellent NO2 sensing properties (S = Ra/Rg) [45,46,47]. In summary, the excellent gas sensing properties of the ZnO/rGOs are achieved by synergistic effect of the appropriate doping ratio ZnO/rGO composites, as well as their high specific surface area and unique porous structure.

$${\text{O}}_{{2}} ({\text{g}}) + {\text{2e}}^{ - } \to {\text{2O}}_{{2}}^{ - } ({\text{ads}})$$
(4)
$${\text{NO}}_{{2}} ({\text{g}}) + {\text{e}}^{ - } \to {\text{NO}}_{{2}}^{ - } ({\text{ads}})$$
(5)
$${\text{NO}}_{{2}}^{ - } ({\text{ads}}) + {\text{O}}_{{2}}^{ - } ({\text{ads}}) + {\text{e}}^{ - } \to {\text{2NO}}_{{3}}^{ - } ({\text{ads}})$$
(6)
Scheme 1
scheme 1

a–e Gas sensing mechanism of ZnO/rGO in air and in NO2 gas

Full size image

Conclusions

In summary, a hybrid material, ZnO/rGO, was successfully synthesized through a facile hydrothermal process. The ratio of rGO in ZnO/rGO had a great effect on the gas sensing behavior. We observed that the ZnO NPs-1% rGO sensor exhibits the highest gas response of 5.21 in 10 ppm NO2 at room temperature among all gas sensors. The enhanced gas sensing performance is attributed to the enlarged specific surface area of the smaller nanoparticles for more active adsorption sites and the formation of p-n heterojunctions. Our sensors are attractive due to their low energy consumption, simplified sensor manufacturing without heating elements, and low cost.