Synthesis of ZnO Hierarchical Structures and Their Gas Sensing Properties

Firecracker-like ZnO hierarchical structures (ZnO HS1) were synthesized by combining electrospinning with hydrothermal methods. Flower-like ZnO hierarchical structures (ZnO HS2) were prepared by a hydrothermal method using ultrasound-treated ZnO nanofibers (ZnO NFs) as raw material which has rarely been reported in previous papers. Scanning electron microscope (SEM) and transmission electron microscope’s (TEM) images clearly indicated the existence of nanoparticles on the ZnO HS2 material. Both gas sensors exhibited high selectivity toward H2S gas over various other gases at 180 °C. The ZnO HS2 gas sensor exhibited higher H2S sensitivity response (50 ppm H2S, 42.298) at 180 °C than ZnO NFs (50 ppm H2S, 9.223) and ZnO HS1 (50 ppm H2S, 17.506) gas sensors. Besides, the ZnO HS2 sensor showed a shorter response time (14 s) compared with the ZnO NFs (25 s) and ZnO HS1 (19 s) gas sensors. The formation diagram of ZnO hierarchical structures and the gas sensing mechanism were evaluated. Apart from the synergistic effect of nanoparticles and nanoflowers, more point–point contacts between flower-like ZnO nanorods were advantageous for the excellent H2S sensing properties of ZnO HS2 material.


Introduction
Hydrogen sulfide (H 2 S) is a flammable, explosive and harmful gas [1]. Long-time exposure to low concentrations of H 2 S atmosphere could cause human irritation in the nasal cavity and eyes, and in high concentrations, an H 2 S atmosphere, even for short-time could cause humans to die [2,3]. Therefore, exploring gas sensors for precise detection of H 2 S gas is becoming an urgent research project.
Recently, 3D hierarchical structures have been extremely appealing due to their unique structures and better gas sensing performance than 1D and 2D materials [4][5][6]. Meanwhile, metal oxide (MO) materials gas sensors, such as ZnO [7][8][9], CuO [10,11], In 2 O 3 [12,13], WO 3 [14,15], CeO 2 [16], SnO 2 [17,18], etc., have attracted extensive attention. Besides, MO materials exhibit outstanding sensing properties in the detection of H 2 S gas [19][20][21][22]. Kaur et al. [23] prepared In 2 O 3 whiskers by the carbothermal method and the In 2 O 3 whiskers exhibited excellent sensing properties toward H 2 S gas. Li et al. [24] reported hierarchical flower-like CuO synthesized by a hydrothermal method, which exhibited high sensitivity and selectivity toward H 2 S. Nakla et al. [25] synthesized flower-like SnO 2 nanowires through a high current heating method and the SnO 2 nanowire sensor showed good H 2 S sensing properties. Among these MO materials, zinc oxide (ZnO) has been developed for the detection of toxic gases, owing to its wind band gap (3.37 eV) [26], big exciton binding energy (60 meV) [27] and excellent chemical stability [28]. Various methods have been developed and applied to prepare ZnO nanomaterials, including thermal evaporation electrodeposition [29], physical vapor deposition [30], the hydrothermal method [31] and the solvothermal method [32]. Among these methods, the hydrothermal method has been widely used in the preparation of nanostructures, owing to its advantages of low costs, a low reaction temperature and easy process control. At the current stage, ZnO hierarchical structures have been created and applied for the detection of the harmful gases. Zhu et al. discovered that ZnO hierarchical nanoflower assembly by nanosheets exhibited superior ethanol gas-sensing performances than those of nanoparticles and nanoplates [33]. ZnO hierarchical structure was also considered to be a potential material for detecting low-concentrations of harmful gases [34,35]. ZnO hierarchically structured materials are expected to have excellent H 2 S gas sensing properties. Liu et al. reported the ZnO hierarchical porous structure exhibited excellent sensing performance toward H 2 S, owing to its higher porosity and surface area [36]. However, flower-like ZnO was generally fabricated by the assistance of surfactant [37,38], it is still a challenge to synthesize hierarchical ZnO structures in low-temperatures without using surfactants and other additives.
As far as we know, the preparation of flower-like ZnO hierarchical structures by the hydrothermal method using ultrasound-treated ZnO nanofibers (NFs) as raw material has been rarely reported. Although serval researchers have synthesized ZnO materials with similar morphologies [39][40][41], the gas sensing performances of firecracker-like ZnO hierarchical structures have also been rarely studied. In this work, ZnO nanofibers and ZnO hierarchical structures with firecracker and flower morphologies were prepared and their gas sensing performances were examined in detail. The gas sensing tests showed that all fabricated ZnO materials have a high selectivity toward H 2 S gas while flower-like ZnO hierarchically structured material achieved the best gas sensing performance toward H 2 S at a 180 • C working temperature. Herein, the growth process and gas sensing mechanism of flower-like and firecracker-like ZnO hierarchical structures were also investigated and proposed.

Synthesis of ZnO NFs
In the typical preparation, 0.2 M Zn (CH 3 COO) 2 ·2H 2 O was dissolved in a solution with an equal volume (10 mL each) of N, N-Dimethylformamide (DMF) and absolute ethanol (C 2 H 5 OH). Next, 2.4 g of polyvinylpyrrolidone (PVP) was added to the solution and vigorously stirred for 10 h. In the electrospinning experiment, the applied electrical voltage was set at 12 kV and the distance between the cylinder collector and the syringe needle loaded with precursor solution was maintained at 15 cm. The dosage rate of solution was kept at 0.56 mL/h. After that, the obtained NFs were calcined at 500 • C for 4 h in a furnace with a 1 • C/min heating rate.

Synthesis of Firecracker and Flower Morphological ZnO Hierarchical Structures
Firecracker and flower morphological ZnO hierarchical structures were prepared via electrospinning and hydrothermal methods. In the experiment, 0.280 g HMTA and 0.595 g Zn (NO 3 ) 2 ·6H 2 O were added to two beakers with 40 mL deionized water (DI water), respectively. Afterward, the aqueous solutions were mixed together and stirred for 5 min. Finally, 0.03 g of the prepared ZnO NFs were put into the solution and agitated for 10 min. The reaction solution was transferred to a 100 mL autoclave and heated at 80 • C for 12 h in the oven. After the temperature of the oven cooled to room temperature, the autoclave was removed. The acquired precipitates were centrifuged and rinsed in the absolute ethanol and DI water three times, respectively. Then the precipitates were dried at 80 • C for 12 h in the oven. At last, ZnO hierarchical structures with firecracker morphology were collected. The difference when preparing flower morphological ZnO hierarchical structures was that the reaction solution with ZnO NFs was sonicated under 100 W power for 2 min before the hydrothermal experiment. The obtained firecracker-like ZnO and flower-like ZnO hierarchical structures are denoted ZnO HS1 and ZnO HS2, respectively.

Characterizations
The phases and compositions of materials were recorded on X-ray diffraction (XRD, Bruker D8 ADVANCE, Karlsruhe, Germany) with Cu Kα radiation (λ = 1.5418Å). The morphology and microstructure of the materials were determined by a scanning electron microscope (SEM, FEI Sirion 200F, USA). High-resolution transmission electron microscopy (HRTEM), transmission electron microscopy (TEM), scanning transmission electron microscopy (STEM) mapping and energy dispersive X-ray (EDX) analysis patterns were obtained by JEOL JEM-2200FS (Tokyo, Japan). Figure 1 displayed the gas sensing system used in the gas sensing tests. It can be seen that the gas sensing system is composed of a PC (personal computer), control station and test station. The working temperature of gas sensors can be adjusted from 20 • C to 500 • C by the control station. In this study, the target gases were injected to the inner chamber through injection holes on the test station. Gas sensors were tested on the test station using two silver probes to form a current loop. The data of gas sensing tests were collected on the PC. Additionally, the gas sensor consisted of Ag-Pd electrodes, a heating layer, a ceramics wafer, sensing materials and a thermocouple. In the gas sensors' manufacturing processes, the prepared materials were mixed together with absolute ethanol and ground by a gridding rod to form a uniform paste. Then the ceramics substrate with electrodes was covered with the paste using a brush. The sensors were annealed at 500 • C for 2 days to improve the stability of sensors. The volume of target gas was calculated and a certain amount of target gas was injected into the test chamber (18 L in volume) using a microsyringe to configure a certain concentration of target gas. The concentration of target gas was controlled by injecting various volumes of gas into the chamber. The gas sensors were examined by a CGS-4TPS intelligent gas sensing analysis system (Beijing Elite Tech Co. Ltd., Beijing, China). The response signal of the gas sensor was defined as Rs = Ra/Rg, where the Ra refers to the resistance of the sensors after it is stabilized in the air and Rg is the resistance of the sensors upon exposure of the target gases. Furthermore, the response time and recovery time were determined as the time to acquire the 90% total resistance change after the target gases were injected, and the time required to reach the 10% total resistance change after the target gases were removed, respectively.

Materials Characterizations
The phase composition of acquired materials was analyzed by XRD. The XRD patterns of ZnO NFs, ZnO HS1 and ZnO HS2 samples were illustrated in Figure 2. All of the diffraction peaks correspond to the hexagonal wurtzite structure ZnO (JCPDS number . Moreover, it is worth noting that no impurity peaks are detected and no peaks shift, indicating that the samples were highly pure. The main diffraction peaks at 2θ = 31.9°, 34.5° and 36.3° were indexed to the 100, 002 and 101 planes of hexagonal ZnO, respectively. The shrill peaks revealed the high crystallinity of all prepared samples. 20  ZnO JCPDS 36-1451 Figure 1. The schematic diagram of the gas sensing analysis system and the composition of a gas sensor.

Materials Characterizations
The phase composition of acquired materials was analyzed by XRD. The XRD patterns of ZnO NFs, ZnO HS1 and ZnO HS2 samples were illustrated in Figure 2. All of the diffraction peaks correspond to the hexagonal wurtzite structure ZnO (JCPDS number . Moreover, it is worth noting that no impurity peaks are detected and no peaks shift, indicating that the samples were highly pure. The main diffraction peaks at 2θ = 31.9 • , 34.5 • and 36.3 • were indexed to the 100, 002 and 101 planes of hexagonal ZnO, respectively. The shrill peaks revealed the high crystallinity of all prepared samples.

Materials Characterizations
The phase composition of acquired materials was analyzed by XRD. The XRD patterns of ZnO NFs, ZnO HS1 and ZnO HS2 samples were illustrated in Figure 2. All of the diffraction peaks correspond to the hexagonal wurtzite structure ZnO (JCPDS number . Moreover, it is worth noting that no impurity peaks are detected and no peaks shift, indicating that the samples were highly pure. The main diffraction peaks at 2θ = 31.9°, 34.5° and 36.3° were indexed to the 100, 002 and 101 planes of hexagonal ZnO, respectively. The shrill peaks revealed the high crystallinity of all prepared samples. 20   The SEM micrographs of precursor and calcined ZnO NFs are displayed in Figure 3, respectively. The synthesized ZnO NFs display white powder morphology, as shown in the inset of Figure 3b. The as-prepared precursor ZnO NFs are smooth and continuous. The ZnO NFs were prepared by annealing the precursor ZnO NFs. The diameter of the precursor ZnO NFs and ZnO NFs were assessed at 430 and 210 nm, respectively. The decrease in diameter is owed to the removal of PVP in precursor ZnO NFs. It can be clearly seen that the ZnO NFs were composed of ZnO nanoparticles, which were tightly bonded. The SEM micrographs of precursor and calcined ZnO NFs are displayed in Figure 3, respectively. The synthesized ZnO NFs a display white powder morphology, as shown in the inset of Figure 3b. The as-prepared precursor ZnO NFs are smooth and continuous. The ZnO NFs were prepared by annealing the precursor ZnO NFs. The diameter of the precursor ZnO NFs and ZnO NFs were assessed at 430 and 210 nm, respectively. The decrease in diameter is owed to the removal of PVP in precursor ZnO NFs. It can be clearly seen that the ZnO NFs were composed of ZnO nanoparticles, which were tightly bonded.      The SEM micrographs of precursor and calcined ZnO NFs are displayed in Figure 3, respectively. The synthesized ZnO NFs a display white powder morphology, as shown in the inset of Figure 3b. The as-prepared precursor ZnO NFs are smooth and continuous. The ZnO NFs were prepared by annealing the precursor ZnO NFs. The diameter of the precursor ZnO NFs and ZnO NFs were assessed at 430 and 210 nm, respectively. The decrease in diameter is owed to the removal of PVP in precursor ZnO NFs. It can be clearly seen that the ZnO NFs were composed of ZnO nanoparticles, which were tightly bonded.      Figure 5 represents the TEM, HRTEM, scanning transmission electron microscopy (STEM) mapping and EDX analysis data of the as-prepared ZnO HS1 and ZnO HS2. As shown in Figure 5a, the firecracker-like ZnO structure material was revealed by the low-magnification TEM images. ZnO Nanomaterials 2019, 9, 1277 6 of 15 nanorods were clearly observed from the image. Figure 5b displays a high magnification TEM image of the firecracker-like ZnO. By measurement, the lattice spacings were determined to be 0.28 nm and 0.247 nm, which were indexed to 100 and 101 planes of hexagonal wurtzite structure of ZnO, respectively. Figure 5d displays the low magnification TEM image of ZnO HS2. Many ZnO nanoparticles were distributed on the ZnO HS2. The lattice spacing of nanoparticles and nanorods were measured to be 0.247 nm in both the cases, as shown in Figure 5c,f. Figure 5g shows the STEM mapping image of ZnO HS2 material. The image indicates that the existence of Zn and O elements of ZnO HS2 material. EDX analysis results of ZnO HS2 were shown in Figure 5h. The results reveal that the ZnO HS2 material had high purity and had no impurity phases. In addition, these results were consistent with the findings of XRD. Nanomaterials 2019, 9, x FOR PEER REVIEW 6 of 15 the firecracker-like ZnO structure material was revealed by the low-magnification TEM images. ZnO nanorods were clearly observed from the image. Figure 5b displays a high magnification TEM image of the firecracker-like ZnO. By measurement, the lattice spacings were determined to be 0.28 nm and 0.247 nm, which were indexed to 100 and 101 planes of hexagonal wurtzite structure of ZnO, respectively. Figure 5d displays the low magnification TEM image of ZnO HS2. Many ZnO nanoparticles were distributed on the ZnO HS2. The lattice spacing of nanoparticles and nanorods were measured to be 0.247 nm in both the cases, as shown in Figure 5c, f. Figure 5g shows the STEM mapping image of ZnO HS2 material. The image indicates that the existence of Zn and O elements of ZnO HS2 material. EDX analysis results of ZnO HS2 were shown in Figure 5h. The results reveal that the ZnO HS2 material had high purity and had no impurity phases. In addition, these results were consistent with the findings of XRD.

Growth Mechanism
On the basis of the experiment results, a possible growth mechanism of the ZnO HS1 and ZnO HS2 is illustrated in Figure 6. The specific chemical reactions are as follows [42]:

Gas Sensing Tests
The gas sensing performances of ZnO NFs, ZnO HS1 and ZnO HS2 sensors were investigated and the results are displayed in Figure 7. Figure 7a shows the responses of gas sensors toward 20 ppm H2S ranging from 120 °C to 270 °C. As the working temperature increased, the responses of gas sensors toward H2S gradually improved. At a 180 °C working temperature, all gas sensors obtained the highest response value, but the gas sensors' responses toward H2S decreased above the 180 °C working temperature. Based on these results, further gas sensing tests were carried out at 180 °C.  Generally, ZnO nanorods tended to achieve the crystal growth on the seed layer and existing nucleus [43]. Meanwhile, the ZnO nanoparticles of ZnO NFs could be the seed layer and template to grow ZnO nanorods [39,40]. In the hydrothermal process, with the temperature of the hydrothermal solution increasing, HMTA hydrolyzed to HCHO and NH 3 . Then, NH 3  these nanoparticles could be the nucleus for the growth of ZnO. Finally, the ZnO HS1s were obtained from the hydrothermal reactions. Besides, the ZnO nanoparticles ( Figure S1) generated from the ultra-sounded ZnO nanofibers would become the nucleus for the formation of ZnO HS2. Ultimately, ZnO nanorods would grow around the hybrid nanoparticles, which could lead to the formation of nanoflowers. Next, ZnO HS1 and ZnO HS2 materials were used as the sensing materials of the gas sensors for further gas sensing tests.

Gas Sensing Tests
The gas sensing performances of ZnO NFs, ZnO HS1 and ZnO HS2 sensors were investigated and the results are displayed in Figure 7. Figure 7a shows the responses of gas sensors toward 20 ppm H 2 S ranging from 120 • C to 270 • C. As the working temperature increased, the responses of gas sensors toward H 2 S gradually improved. At a 180 • C working temperature, all gas sensors obtained the highest response value, but the gas sensors' responses toward H 2 S decreased above the 180 • C working temperature. Based on these results, further gas sensing tests were carried out at 180 • C. Figure Figure 7c presents the dynamic response and recovery curves of gas sensors toward different concentrations of H 2 S. The ZnO HS2 sensor exhibited excellent reproducibility under various concentrations of H 2 S and presented a larger change of response compared with ZnO NFs and ZnO HS1 sensors. Furthermore, the selectivity of gas sensors was also measured. The sensing performances of sensors to different target gases, including 100 ppm ethanol, acetone, methanol, formaldehyde and ammonia, and 50 ppm H 2 S, were investigated. Although the results in Figure 7d reveal that all the gas sensors based on ZnO NFs, ZnO HS1 and ZnO HS2 exhibited good selectivity toward H 2 S gas at 180 • C, the ZnO HS2 sensor showed stronger selectivity toward H 2 S. Overall, the ZnO HS2 sensor exhibited a better selectivity and response performance toward H 2 S. Therefore, it has more potential to be applied in the detection of H 2 S gas.
The long-term stability is also a vital parameter of gas sensors. Figure 8 shows the stability results of three gas sensors toward 50 ppm H 2 S over a term of 20 days. The sensing performance of gas sensors was carried out every two days for 20 days. During 20-days stability tests, the ZnO NFs gas sensors exhibited the best stability, with only a 3% drop in response. However, both ZnO HS1 and HS2 sensors still exhibited a higher response than the ZnO NFs sensor, with 7% decreases in response. Therefore, the ZnO NFs sensor is better in terms of stability, and ZnO HS1 and HS2 sensors are better in terms of response performance.
Overall, the ZnO HS2 sensor exhibited a better selectivity and response performance toward H2S. Therefore, it has more potential to be applied in the detection of H2S gas. The long-term stability is also a vital parameter of gas sensors. Figure 8 shows the stability results of three gas sensors toward 50 ppm H2S over a term of 20 days. The sensing performance of gas sensors was carried out every two days for 20 days. During 20-days stability tests, the ZnO NFs gas sensors exhibited the best stability, with only a 3% drop in response. However, both ZnO HS1 and HS2 sensors still exhibited a higher response than the ZnO NFs sensor, with 7% decreases in response. Therefore, the ZnO NFs sensor is better in terms of stability, and ZnO HS1 and HS2 sensors are better in terms of response performance.  Table 1. The ZnO gas sensors in this work displayed a shorter recovery time than those other reported ZnO gas sensors. ZnO HS2 sensor exhibited shorter recovery time and higher response value compared with other ZnO gas sensors. Therefore, ZnO HS2 material exhibited better H2S sensing properties, which has the potential to be applied in the detection of H2S gas. The comparison results of the sensing characteristics of ZnO gas sensors in the present work with other previous ZnO nanostructures gas sensors are shown in Table 1. The ZnO gas sensors in this work displayed a shorter recovery time than those other reported ZnO gas sensors. ZnO HS2 sensor exhibited shorter recovery time and higher response value compared with other ZnO gas sensors. Therefore, ZnO HS2 material exhibited better H 2 S sensing properties, which has the potential to be applied in the detection of H 2 S gas.

Gas Sensing Mechanism
The change of resistance because of the charge transfers between detected gas and the surface of ZnO samples is generally the sensing mechanism of ZnO based gas sensors [48]. When ZnO sensors were exposed to air at 180 • C, oxygen molecules were absorbed on the surface of ZnO [49]. Meanwhile, the electrons were extracted from the conduction band of ZnO, resulting in the formation of chemisorbed oxygen species O − (100-300 • C) [50,51]. The depletion layer and potential barrier would have formed during the above process, which increased the resistance of sensors. That mechanism could be expressed by the equations [35]: When the ZnO sensors are exposed to H 2 S, the electrons are returned to the conduction band after the O − reacts with H 2 S. At the same time, the H 2 S molecules react with ZnO and forms zinc sulfide (ZnS). The resistance of ZnO sensors would have an obvious and rapid decrease after the introduction of H 2 S. Owing to the bond energy of H-SH in H 2 S being 381 kJ/mol, lower than ethanol (436 kJ/mol), methanol (436.8 kJ/mol), ammonia (391 kJ/mol) and acetone (393 kJ/mol), breaking the bond H-SH at 180 • C was easier than those other gases. Although formaldehyde (364 kJ/mol) and acetone (393 kJ/mol) and ammonia (391 kJ/mol) also have small bond energies, the reactions between the sensor and target gases were also important. The reaction between ZnO and H 2 S is a spontaneous reaction, as shown in Equation (10), which can increase the sensitivity of ZnO to H 2 S. However, formaldehyde, methanol and ethanol, etc., more easily react with active oxygen species (O − ), as shown in Equation (11), where the R refers to these gases. The related reactions are as described [20,36,46,52]: ZnO(s) + H 2 S(ads) → ZnS(s) + H 2 O(g) (10) Finally, the above process would result in the resistance of ZnO sensors and a decrease in the potential barrier as shown in Figure 9. At the desorption process, the ZnS is converted to ZnO by reacting with O 2 ; the reaction is shown as follows [53,54]: 2ZnS(s) + 3O 2 (g) → 2ZnO(s) + 2SO 2 (g) (12) The reasons for the better gas sensing performance of the ZnO HS2 sensor are described as follows. It can be observed from the Figure 4e,f (SEM) and Figure 5d (TEM) that the ZnO nanoparticles without growing into ZnO nanoflowers were distributed on the surface of the ZnO HS2 and the diameters of nanoparticles were measured to range from about 50 to 150 nm. In previous literature [55], the synergistic effect has been used to improve the sensing response, because more active sites could be generated. Therefore, the ZnO HS2 structures were composed of both nanoparticles and nanoflowers, which could synergistically improve gas sensing properties [56][57][58]. Moreover, owing to the large size of nanorods constituting the ZnO HS2, more point-point contacts existed in the ZnO HS2 sensor. It can be obviously seen from Figure 9 that many point-point contacts were in the ZnO HS2, which could increase the number of potential barriers. To the best of our knowledge, electrons are required to overcome the potential barriers which could result in the higher resistance of the sensor [59,60]. In contrast, the nanorods grown on the surface of ZnO HS1 were smaller in length as shown in Figure 4a-c. The number of point-point contacts was smaller than that of ZnO HS2. The high selectivity of ZnO HS2 may be attributed to the morphology of nanorod bundles of ZnO HS2 and lower H-SH bond energy [61,62].
H2S(g) + 2 O (ads) → H2(g) + SO2(g) + 2e (9) ZnO(s) + H2S(ads) → ZnS(s) + H2O(g) (10) Finally, the above process would result in the resistance of ZnO sensors and a decrease in the potential barrier as shown in Figure 9. At the desorption process, the ZnS is converted to ZnO by reacting with O2; the reaction is shown as follows [53,54]: 2ZnS(s) + 3O2(g) → 2ZnO(s) + 2SO2(g) (12) The reasons for the better gas sensing performance of the ZnO HS2 sensor are described as follows. It can be observed from the Figure 4e, f (SEM) and Figure 5d (TEM) that the ZnO nanoparticles without growing into ZnO nanoflowers were distributed on the surface of the ZnO HS2 and the diameters of nanoparticles were measured to range from about 50 to 150 nm. In previous literature [55], the synergistic effect has been used to improve the sensing response, because more active sites could be generated. Therefore, the ZnO HS2 structures were composed of both nanoparticles and nanoflowers, which could synergistically improve gas sensing properties [56][57][58]. Moreover, owing to the large size of nanorods constituting the ZnO HS2, more point-point contacts existed in the ZnO HS2 sensor. It can be obviously seen from Figure 9 that many point-point contacts were in the ZnO HS2, which could increase the number of potential barriers. To the best of our knowledge, electrons are required to overcome the potential barriers which could result in the higher resistance of the sensor [59,60]. In contrast, the nanorods grown on the surface of ZnO HS1 were smaller in length as shown in Figure 4a-c. The number of point-point contacts was smaller than that of ZnO HS2. The high selectivity of ZnO HS2 may be attributed to the morphology of nanorod bundles of ZnO HS2 and lower H-SH bond energy [61,62].

Conclusions
In summary, both firecracker-like and flower-like ZnO hierarchical structures have been successfully fabricated without using surfactants. ZnO HS2 displayed the morphology of nanoflowers modified by nanoparticles. The growth process diagram of firecracker-like and flower-like ZnO hierarchical structures was described. The gas sensing results revealed that gas sensors exhibited high selectively and good long-term stability toward H 2 S. The gas sensing studies demonstrated that the ZnO HS2 gas sensor exhibited the highest sensing properties toward H 2 S. The ZnO HS2 gas sensor has a much higher response toward H 2 S gas (50 ppm H 2 S, 42.298) than the ZnO NFs gas sensor (50 ppm H 2 S, 9.223) and ZnO HS1 gas sensor (50 ppm H 2 S, 17.506). Systematic sensing tests of gas sensors toward H 2 S, including different temperatures and different concentrations were carried out, and the gas sensing mechanism of the sensors was also proposed. The synergistic effect of nanoparticles and nanoflowers, and more point-point contacts between nanorods of ZnO HS2 are advantageous for the high H 2 S sensing properties of flower-like ZnO hierarchically structured material. Our study indicated that the morphology of ZnO materials affect the gas sensing properties of materials, and that hierarchically structured ZnO materials could be developed as inexpensive and excellent H 2 S gas sensors.