Ultra-sensitive and rapid response H₂S microsensor based on Co–ZnO with ppm level detection

2.1 Preparation of the sensing material Co _x Zn_1−x O (x = 0, 0.05, 0.1, 0.15, 0.2)

Every reagent employed in this research was of analytical grade quality. To prepare pure ZnO, a solution of oxalic acid dihydrate (COOH)₂·2H₂O was prepared by dissolving 3.78 g of dehydrated oxalic acid (COOH)₂·2H₂O in 50 mL of deionized water. Another solution was prepared by dissolving 3.38 g of zinc acetate dihydrate [Zn(CH₃COO)₂·2H₂O, ≈ 0.0154 mol]. The two solutions were mixed and stirred at 60 °C for 5 h. The precipitated zinc oxalate was filtered, washed, and dried at 100 °C for 12 h. Finally, the dried zinc oxalate was calcined at 450 °C for 2 h with a heating rate of 5 K min⁻¹, yielding ZnO powder. ¹⁴ Cobalt-doped zinc oxide was synthesized via a chemical co-precipitation method. Zinc acetate dihydrate [Zn(CH₃COO)₂·2H₂O] (0.0022 mol, 0.483 g) of was dissolved in 50 mL of deionized water, and 0.00022 mol (0.064 g) of cobalt nitrate hexahydrate [Co(NO₃)₂·6H₂O] was dissolved in 30 mL of deionized water. The two solutions were mixed under stirring, and a 2 m NaOH solution (∼25 mL) was slowly added dropwise until the pH reached 10. The mixture was stirred continuously for 2 h at room temperature. The resulting precipitate was filtered, washed several times with deionized water and ethanol, and then dried at 100 °C for 12 h. Finally, the dried powder was annealed at 450 °C for 2 h in air (heating rate: 5 K min⁻¹) to obtain Co_0.1Zn_0.9O. ¹⁴ The same procedure was used for the preparation of Co_0.05Zn_0.95O, Co_0.15Zn_0.85O and Co_0.2Zn_0.8O.

2.2 Characterization

A Bruker D8 ADVANCE X-ray diffractometer, running at 40 mA and 40 kV, was employed to examine the crystal structure of the sample. The wavelengths were adjusted to 1.5406 Å for the Cu target and 1.79026 Å for the Co target. The elemental composition of the powder was assessed, and the element content was validated through XPS analysis, carried out with the Thermo Scientific Escalab 250XI device. A monochromatic AlKα source with 1,486.6 eV photon energy was employed for the analysis, operating at 150 W. The spot size was adjusted to 650 μm, with settings of 14.8 kV for voltage and 1.6 A for current. Calibration of the charge was based on the C1s peak of contaminant carbon at 284.8 eV. Full spectra were measured with a 100 eV pass energy and 1 eV step size, while narrow spectra were gathered with a pass energy of 20 eV and a step size of 0.1 eV. To analyze the surface morphology of the samples, a HITACHI S-4800 field-emission scanning electron microscope (SEM) was employed. EDS analysis, conducted with a HORIBA EMAX mics2, employed accelerating voltages of 3 kV or 5 kV and reached a resolution of 50 nm. Measurements were taken under normal indoor temperature settings.

2.3 The gas sensing characteristics of the microsensors

The sensitive material and deionized water were blended in a ratio of 3 : 1 by weight and ground into a fine paste using an agate mortar. This paste was applied onto a micro-hotplate using a micropipette for drop-casting. After drying at room temperature, the material was aged for 48 h in a vacuum oven at 300 °C. The sensing elements then underwent a two-hour heat treatment at 450 °C, which evaporated any bound water and produced a dense oxide layer on the surface of the gas sensor. After cooling, the electrodes were powered using a CNC power supply to provide adequate reaction temperature for the micro-hotplate. An electrical evaluation of the sensor was performed using a gas analysis system (CHI760E, Shanghai Chenhua). For the experiment, the apparatus was situated within a fume hood, and a measured quantity of the target gas, blended with dry air, was introduced into the chamber to engage with the sensitive chip over a set time. Testing of the I-T curve commenced once the CHI760E software was launched.

As illustrated in Figure 1, the experimental setup of the hydrogen sulfide gas sensor designed in this paper comprises a numerical control power supply (DP 100), an electrochemical workstation (CH Instruments), a microsensor chip, a gas chamber, and the reactive gases (hydrogen sulfide gas and gas reaction gases), forming a micro-gas sensing system. One end of the chip is connected to the CNC power supply (DP 100), while the other end is linked to the CHI760E electrochemical workstation (CH Instruments) for real-time monitoring of electrical signals.

Figure 1:

Diagram of the experimental setup for H₂S monitoring using a microsensor.

The micro-gas sensor device is shown in Figure 2(a). The internal structure of the chip used in this experiment has a support layer, a sensitive film, two sets of electrodes, one set of heating electrodes and one set of test electrodes. The internal structure of the microsensor chip is shown in Figure 2(b). The size of the chip is 1 × 1 mm². The gas chamber of the gas sensor is made of 2 mm thick acrylic material, and the internal size is 10 × 5 × 5 cm³, as shown in Figure 2(c). The gas sensor is depicted in Figure 2(d).

Figure 2:

Microsensor of (a) internal structure of the chip, (b) the chip, (c) the chamber, and (d) the gas sensor device.

3.1 XRD characterization

Figure 3 presents the XRD results, obtained using CoKα radiation (λ = 1.79026 Å), where the peaks observed at 2θ diffraction angles of 37.1°, 40.2°, 42.3°, 55.8°, 66.9°, 74.6° and 79.1°, which belong to the (1 0 0), (0 0 2), (1 0 1), (1 0 2), (1 1 0), (1 0 3) and (2 0 0) faces of the hexagonal lattice (space group P63mc, #186; PDF#36-1,451). ¹⁵ The angles of the Co-doped ZnO diffraction peaks were consistent with those of undoped ZnO, confirming that the introduction of Co had no effect on altering the crystalline framework of the Co _x Zn_1−x O (x = 0, 0.05, 0.1, 0.15, 0.2) composition. Indeed a peak was observed in the XRD profile at 43.2° which can be attributed to a Co reflection. ¹⁶ No other impurity peaks, apart from those of ZnO and Co, were detected in the experiment, confirming that the Co _x Zn_1−x O (x = 0, 0.05, 0.1, 0.15, 0.2) sensitive composite material was successfully prepared with good crystallinity.

Figure 3:

XRD pattern of Co _x Zn_1−x O (x = 0, 0.05, 0.1, 0.15, 0.2) sensing material.

3.2 XPS characterization

The XPS spectrum for the Co _x Zn_1−x O (x = 0, 0.05, 0.1, 0.15, 0.2) samples is shown in Figure 4(a). The elements present in the samples are Co, Zn, and O. The C 1s ¹⁷ peak corresponds to either organic carbon contamination or a small amount of glycerol that was not completely decomposed in the composite.

Figure 4:

The XPS spectrum. (a) XPS full-scan image of Co _x Zn_1−x O (x = 0, 0.05, 0.1, 0.15, 0.2), (b) Zn 2p region, (c) Co 2p region, (d) O 1s.

The fitted Zn 2p peaks ¹⁸ in Figure 4(b) show binding energies of 1,044.3 eV for Zn 2p _1/2 and 1,022.2 eV for Zn 2p _3/2, respectively. Compared to pure ZnO (1,021.58 eV), the Zn peak in Co₃O₄/ZnO shifts to a higher binding energy, which suggests strong electronic interactions in the sample, likely due to the interaction between Co₃O₄ and ZnO. These findings further suggest an enhanced interaction between ZnO and Co₃O₄, indicating a unique surface architecture in the Co _x Zn_1−x O materials (x = 0, 0.05, 0.1, 0.15, 0.2). Figure 4(c) shows Co 2p ¹⁹ peaks at binding energies of 796.85 eV for Co 2p _1/2 and 779.21 eV for Co 2p _3/2, respectively, indicating that Co is present in both Co²⁺ and Co³⁺ states. Due to the low doping content, the signal intensity is relatively weak.

To determine the proportion of chemisorbed oxygen, the O 1s ²⁰ peak was analyzed via XPS testing. As shown in Figure 4(d), deconvolution of the O 1s peak reveals three components. The binding energies measured at 530.84 eV and 531.25 eV are associated with oxygen in the adsorbed state and lattice-bound oxygen, respectively. The binding energy at the adsorbed oxygen site (Oc) is typically attributed to chemisorption on the surface of the sensitive film. As the Co content increases, the binding energy of adsorbed oxygen shifts to higher values. The interaction between oxygen and the material’s surface through chemisorption plays a key role in influencing the response behavior of the gas sensor. ²¹

3.3 SEM characterization

The prepared sensitive materials were mechanically ground, then dispersed in deionized water, and finally tested at room temperature. The SEM images presented in Figure 5 illustrate the structural morphology of the Co _x Zn_1−x O (x = 0, 0.05, 0.1, 0.15, 0.2) sensitive materials at magnifications of 5×, 10×, and 20×,. The Co-doped sensitive material has a relatively regular and loose porous spherical material on the surface. Under 50× magnification, the surface of the sensitive materials is clearly discernible. The images reveal unevenly shaped droplets of various sizes, characterized by a porous, loosely structured arrangement. ²² The smallest irregular spherical grains are primarily distributed in the 20–25 μm range, with clear particle boundaries and relatively smooth surfaces. Gaps between the sheets enhance the ability to adsorb and desorb gases.

Figure 5:

The FE-SEM spectrum of the Co _x Zn_1−x O (x = 0, 0.05, 0.1, 0.15, 0.2) sensitive material (a)–(c) ZnO, (d)–(f) Co_0.05Zn_0.95O, (g)–(i) Co_0.1Zn_0.9O, (j)–(l) Co_0.15Zn_0.85O, (m)–(o) Co_0.2Zn_0.8O.

In Figure 6(a)–(d), the elemental distribution maps of the sensitive material Co_0.1Zn_0.9O are presented. The prepared Co_0.1Zn_0.9O contains the corresponding elements, with no other impurities. Figure 6(e) presents the HAADF spectrum of Co_0.1Zn_0.9O.

Figure 6:

The Co_0.1Zn_0.9O HRTEM images. (a) The HADDF image of Co_0.1Zn_0.9O, the mapping image of (b)O, (c)Zn, and (d)Co, and (e) the EDS spectrum of Co _x Zn_1−x O.

3.4 Gas sensing

The prepared sensitive materials were placed in a 100 ppm hydrogen sulfide atmosphere. As can be seen from Figure 7, the response was strongest when x = 0.1. Therefore, the sensitive material selected for the gas experiment is Co_0.1Zn_0.9O. When functioning at 200 °C, Co_0.1Zn_0.9O materials were subjected to hydrogen sulfide gas concentrations ranging between 0 and 100 ppm, with measurements taken at 0, 20, 50, 80, and 100 ppm. Figure 8(a) presents the sensor’s response curves for varying concentrations of hydrogen sulfide gas. With higher concentrations of the test gas, the sensor covered with the reactive material displays a heightened response to hydrogen sulfide. The microsensor showed the best sensitivity to H₂S at 0.02565 mA ppm⁻¹ (95 % confidence intervals, CI: 0.0255 to 0.0258), with a linearity of 0.99506 (95 % CI: 0.9986 to 0.9977) (Figure 8(b)).

Figure 7:

The gas sensor testing of different materials under 100 ppm hydrogen sulfide conditions.

Figure 8:

The response test diagrams. (a) Gas sensitivity response characteristic curve of the Co_0.1Zn_0.9O microsensor to different concentrations of H₂S, (b) the linear fitting of the sensor.

The experimental data was linearized using Origin, and the standard deviation (δ) was obtained. The gas sensor’s limit of detection (LOD) was determined through calculation, with K representing the slope of the linear fit. The sensor’s LOD is: ²³

The calculated LOD for the gas sensor can reach 0.03521 ppm.

The ability to selectively detect specific gases is a vital performance criterion in determining the application potential of sensors. For evaluating how the coated gas sensor responds to different substances, including air, Cl₂, CO₂, CO, ethanol, SF₆ and H₂S, all at equal 100 ppm concentrations, were prepared and introduced into the gas chamber for testing.

Figure 9 shows the gas selectivity of the sensor, exhibiting low sensitivity to air, Cl₂, CO₂, CO, ethanol, SF₆ and H₂S gases at the same concentration, indicating that Co_0.1Zn_0.9O has specific selectivity for H₂S.

Figure 9:

The selectivity of the 0.1Co–ZnO microsensor for air, Cl₂, CO₂, CO, ethanol, SF₆ and H₂S.

Figure 10 illustrates the sensor’s dynamic behavior during its response and recovery phases. Measurements were recorded every second, with response time assessed at randomly chosen moments within the response interval. To mitigate inaccuracies caused by manual computation, the response and recovery durations must be repeatedly evaluated. With a response time near 7.9 s (95 % CI: 7.757 to 8.043) and a recovery time close to 6.5 s (95 % CI: 6.357 to 6.643), the sensor demonstrates fast performance. Table 1 ^{24 , 25 , 26 , 27 , 28 , 29 , 30} outlines a comparative analysis of the Co_0.1Zn_0.9O sensor’s response and recovery efficiency relative to various other gas sensors.

Figure 10:

The response and recover time of the Co_0.1Zn_0.9O micro-hotplate sensor.

As seen in Figure 11(a), the sensor demonstrates good reproducibility in 50 ppm H₂S gas. For the same sensor, the reproducibility error, or the relative standard deviation of sensitivity, should not be greater than ±5 %. Achieving the reproducibility error within the defined limits indicates that the Co_0.1Zn_0.9O microsensor maintains reliable reproducibility. To assess long-term stability, the sensor was tested for 30 days, with response measurements conducted at 5-day intervals in a 50 ppm hydrogen sulfide atmosphere. As shown in Figure 11(b), the response exhibited minimal fluctuation, remaining largely stable. This strongly suggests that the Co_0.1Zn_0.9O gas sensor maintains remarkable consistency over both brief and prolonged durations, making it a trustworthy choice for practical use in the market.

Figure 11:

The reproducibility and stability testing. (a) Reproducibility of the Co_0.1Zn_0.9O sensor for H₂S at 50 ppm concentration, (b) time stability of the sensor.

4.1 Sensing principle

The basic structure of the gas sensor consists of a microheater and a Co _x Zn_1−x O composite material. The sensitive material is uniformly spin-coated onto the microheater and then placed in the gas chamber for gas experiments. When the detected gas enters the sensor, it adsorbs or reacts with the Co _x Zn_1−x O sensitive film, altering the concentration and mobility of carriers (electrons or holes) on its surface or in its volume, which in turn changes its resistance, enabling gas concentration detection by measuring the current.

Co _x Zn_1−x O ^{31 , 32} is an n-type semiconductor with a wide bandgap, and its bandgap characteristics are suitable for illustrating the traditional electron depletion model. The porous structure in the layered microspheres provides more reaction sites for H₂S, greatly enhancing the efficiency of gas-sensitive reactions.

The gas sensing mechanism is shown in Figure 12. When air comes into contact with the microheater sensor, oxygen molecules adsorb onto the defect sites of the thin sensitive coating. At this interface, oxygen molecules are reduced by removing electrons from the conduction band of the material, forming oxygen species such as O⁻, O₂ ⁻, and O²⁻. ^{33 , 34} This removal of electrons creates a charge depletion region on the surface of the Co _x Zn_1−x O material, which increases its resistance (Equations (1) and (2)). This keeps the sensor in a high resistance (Ra) state when exposed to air. The formation of Co(III) is closely related to the reduction process of oxygen. During the adsorption of oxygen and electron removal, Co(III) plays an important role as an intermediate species, facilitating the reduction of oxygen molecules. Co(III) accepts electrons removed from the conduction band and is converted into Co(II), a process that is crucial for the response of the gas sensor. ³⁵ When the reducing gas H₂S is introduced into the test chamber, it reacts with the oxygen ions adsorbed on the surface, releasing the electrons captured by the chemisorbed oxygen. These free electrons then return to the conduction band of the material (Equations (3) and (4)). ^{33 , 34} This causes a significant decrease in the material’s resistance (Rg) and greatly enhances the sensing response.

Figure 12:

Gas sensing mechanism of Co _x Zn_1−x O.

With the reduction of the boundary barrier between neighboring particles, the electron depletion zone shallows, resulting in a swift drop in the resistance of the sensor. Adjusting the heterojunction barrier height in the presence of adsorbed oxygen is critical for improving the sensing performance. When the reducing gas is present, the resistance (Rg) is significantly lowered, thereby greatly enhancing the sensing response.