1. Introduction
Recently, humidity sensors have been developed rapidly and have attracted much attention in many fields such as agriculture, food safety, industrial production, medical treatment and so on [
1,
2,
3]. Among the many sensor types, resistive sensors are the most common sensitive components. The applicability is greatly expanded due to its simple manufacturing, high detection sensitivity, and low production cost. A resistive humidity sensor is obtained by covering a substrate with a film made of a humidity-sensitive material. The resistance of the component will change when a water molecule is adsorbed onto the sensitive film under different relative humidity environments, so as to achieve the purpose of relative humidity detection [
4,
5]. In terms of the structure of the resistance humidity sensor, the selection of sensitive materials is still the key to improving the performance of the relative humidity sensor. Recently, metal oxides, polymers, and carbon materials are often used in the design and manufacture of humidity sensors. Among many materials, SnO
2 [
6,
7], TiO
2 [
8,
9], WO
3 [
10,
11], ZnO [
12,
13] and other metal oxide semiconductor materials [
14,
15,
16] are widely used in the preparation of sensors because of their simple preparation process, low price, wide source of raw materials and good biological compatibility. Among them, ZnO nanostructures have become a potential sensing material due to their high activity and large specific surface area [
17,
18], which has attracted widespread attention. Furthermore, ZnO nanostructures exhibit morphologically dependent sensing characteristics [
19,
20], so the morphological structure also plays a key role in the performance of ZnO sensors.
ZnO as a sensitive material shows poor linearity and low sensitivity for relative humidity detection, which limits its application as a humidity sensor. With the in-depth research requiring the sensitivity of the humidity sensor, the linearity stability needs to be further improved. To improve the responsiveness of ZnO humidity sensors, noble metal doping (Au, Ag, Pd and Pt) is a frequently used method [
21,
22]. Silver (Ag) is one of the most conductive materials, and its low cost and good catalytic performance have been widely used in sensors. The introduction of modified silver nanoparticles into ZnO can control the surface morphology and crystal structure of ZnO, which is expected to improve the performance of humidity sensors. Moreover, after Ag is introduced into ZnO, the adsorption sites on the surface of the material will increase, and the number of surface defects will increase. On the one hand, large adsorption sites and high surface defects can make water molecules adsorbed on surface of the material decompose quickly, and hence improves the response speed of the sensor. On the other hand, silver particles have good electrical conductivity, which is beneficial to improve linearity of the sensor, so Ag/ZnO composite material can provide a new idea for the preparation of high-performance humidity sensors.
In this work, we successfully prepared an Ag/ZnO humidity sensor using the hydrothermal method, and response of the sensor under different light conditions was studied. Experimental results show that an Ag/ZnO humidity sensor exhibits high sensitivity, low hysteresis and short response/recovery time when a molar ratio of Ag+ to Zn2+ is 1:100. This is mainly because in this case there are a large number of oxygen vacancies and active sites on the surface of ZnO. Under the action of oxygen vacancies and active sites, the water molecules on the surface of the Ag/ZnO-2 humidity sensor are rapidly decomposed, thereby increasing response of the humidity sensor. Compared with pure ZnO, such a relative humidity sensor based on modified silver has better linearity.
3. Results and Discussion
Figure 2 shows XRD patterns of ZnO, Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3, which can be used to analyze crystal structures of all the samples. The diffraction peaks of all samples at 2θ of 31.8°, 34.2°, 36.3°, 47.5°, 56.7°, 62.7°, 66.4°, 68.1° and 69.1° are similar to the ones of typical hexagonal wurtzite structures consistent (JCPDS No. 36–1451). Compared with ZnO, the weaker diffraction peaks corresponding to (111), (200), and (311) crystal planes in Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3 belong to Ag crystals of face centered cubic (fcc) structure. The appearance of (111), (200), and (311) crystal plane diffraction peaks indicate the successful recombination of Ag and ZnO. Compared with Ag/ZnO-1, the intensity of diffraction peak of Ag in Ag/ZnO-2 and Ag/ZnO-3 gradually becomes weaker, which is mainly due to the gradually decreasing Ag content in the sample.
Figure 3 shows the scanning electron microscope images of ZnO and Ag/ZnO. In
Figure 3a,b, ZnO and Ag/ZnO-1 exhibit irregular microparticles with a diameter of about 50–70 nm and 130–160 nm, and these irregular microparticles are agglomerated. Ag/ZnO-2 are irregular microparticles with a diameter of about 70–90 nm, and the particles are evenly distributed and dispersed without agglomeration (
Figure 3c). The structure of Ag/ZnO-2 is evenly distributed and more dispersed. Compared with the other three structures, it can provide more adsorption sites for water molecules, making more water molecules adsorb to the surface of the material. This structure plays a vital role in the humidity sensor. In
Figure 3d, Ag/ZnO-3 is also of irregular particles with a diameter of about 60–70 nm, and the particles are slightly agglomerated. In general, compared with the other three structures, Ag/ZnO-2 may show better humidity sensing performance.
In
Figure 4, the optical absorption characteristics of all the samples are given by UV-vis absorption spectrum. In the ultraviolet region as shown in
Figure 4a, the strong absorptions for all of ZnO, Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3 indicate that all the samples are wide bandgap direct semiconductors. The strong absorption peak located at 350 nm belongs to intrinsic absorption of ZnO. Because ZnO has strong absorption in ultraviolet light and weak absorption in visible light, there is a sharp decrease at 390 nm. Compared with ZnO, there is blue shift for all the absorption peaks of Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3, which indicates that the band gap of the samples gradually becomes smaller. The absorption peaks of Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3 at 480 nm are gradually weakened with the decreasing of Ag content.
Figure 4b shows the band gaps of ZnO, Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3 calculated by the Kubelka–Munk formula (αhν)
2 = A (hν − E
g), where α is the absorption coefficient, hν is the photon energy, E
g is the band gap energy, and A is a constant. The band gaps of ZnO, Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3 are 3.00 eV, 1.94 eV, 2.80 eV and 2.92 eV, respectively. With the increase of the doped Ag content, the forbidden band width of the sample gradually decreases, which is consistent with the analysis result as in
Figure 4a. The lower the band gap energy of the material, the lower the energy it should generate, which is beneficial to improving the conductivity of the humidity sensor.
In order to further determine the state of oxygen on the material surface, the O 1s peak XPS spectra of ZnO and Ag/ZnO are fitted in
Figure 5. The O 1s of all samples are fitted to three peaks, namely O
1, O
2 and O
3. The O
1 peak at 529.3 eV is the O
2− bonded to Zn
2+, the O
2 peak at 530.2 eV is the defect oxygen on the zinc oxide surface, and the O
3 peak at 531.2 eV is some adsorbed oxygen on the zinc oxide surface. After calculation, it was found that the defect oxygen area ratios of the four samples are 27.8%, 24.1%, 38.9% and 25.6%, respectively. It can be seen that the number of oxygen vacancies on the Ag/ZnO-2 surface is the largest. It is well known that oxygen vacancies can accelerate the decomposition of water molecules into conductive ions [
23,
24]. For Ag/ZnO-2, recombination of Ag leads to the most oxygen vacancies to dissociate water molecules adsorbed on the surface of zinc oxide to form H
3O
+ conductive ions, hence improving humidity sensitivity.
Figure 6a shows the response curves of ZnO, Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3. We found that the performance of humidity sensor became worse with the increase of Ag
+ concentration. The humidity sensor shows a good response and the resistance changes more than three orders of magnitude when the molar ratio of Ag
+: Zn
2+ is 0.01 (Ag/ZnO-2). It is reported that
is defined as the response of the sensor [
14,
25], so it can be calculated that the responses of ZnO, Ag/ZnO-1, Ag/ZnO-2 and Ag/ZnO-3 are 64,300%, 1510%, 151,700% and 10,200%, respectively. Compared with the rGO/ZnO nanorods/Cu humidity sensor (97.79%) reported by Kuntal et al. and the Er:ZnO humidity sensor (impedance change of about three orders of magnitude) reported by Zhang et al., the Ag/ZnO-2 humidity sensor showed a better response [
26,
27]. The response of Ag/ZnO-2 is probably due to the large amount of oxygen vacancy on surface of ZnO when molar ratio of Ag
+: Zn
2+ is 0.01. As an active site, oxygen vacancy can accelerate the decomposition of water molecules adsorbed on surface of ZnO, which makes more water molecules decompose into conductive ions, thus improving the response of Ag/ZnO-2 humidity sensor. In addition, the uniformly dispersed sheet structure of Ag/ZnO-2 can also be used for more water molecules to be adsorbed, which increases the amount of water absorbed on surface of the material and can also enhance the performance of the humidity sensor. Compared with Ag/ZnO-2, ZnO, Ag/ZnO-1 and Ag/ZnO-3 have poorer response. On one hand, there are not enough oxygen vacancies on the surface, which limits the decomposition of water molecules. On the other hand, ZnO, Ag/ZnO-1 and Ag/ZnO-3 particles are agglomerated, resulting in relatively few adsorbed water molecules, which in turn affects humidity sensor response.
In order to determine the optimal test frequency of the Ag/ZnO-3 humidity sensor, we tested the response of Ag/ZnO-1 at 40 Hz, 100 Hz, 1 kHz, 10 kHz and 100 kHz under different relative humidity, and the test results are shown in
Figure 6b. Although the response of Ag/ZnO-2 humidity sensor is high at 40 Hz, the overall linearity of the sensor is poor, so 40 Hz cannot be selected as the best test frequency. At 1 kHz, 10 kHz and 100 kHz, the response and linearity of the sensor are poor because the water molecules cannot be polarized at high frequencies, and the polarization of water molecules adsorbed by the sensor cannot keep up with the direction of electric field change in the high-frequency region. Only when the test frequency is 100 Hz, Ag/ZnO-3 humidity sensor shows high response and good linearity. Therefore, we choose 100 Hz as the best test frequency, and all subsequent tests are conducted at this frequency.
The hysteresis, response/recovery time, and repeatability of the humidity sensor are also the main factors that determine the performance of the sensor.
Figure 7a shows the hysteresis test of Ag/ZnO-2 humidity sensor in the range of 11% RH to 95% RH. When the environmental RH changes from 11% to 95% RH, the process of the sensor continuously adsorbing water molecules is an adsorption process. Conversely, when the RH is changed from 95% to 11% RH, the process by which the sensor continuously separates from water molecules is a desorption process. It can be observed that the resistance value of adsorption process is almost higher than that of desorption process in whole detection range, which is mainly caused by the endothermic desorption process of water molecules on Ag/ZnO-2 surface is slower than the exothermic adsorption process. The hysteresis error can be calculated according to γH = ±ΔH
max/2F
FS, where ΔH
max is the maximum hysteresis value and F
FS is the full range output of the sensor [
28]. It can be calculated that the maximum hysteresis error of the Ag/ZnO humidity sensor is 3%. Compared with previously reported ZnO humidity sensor, the Ag/ZnO-2 humidity sensor also has a low hysteresis error. Ag/ZnO-2 humidity sensor response/recovery time and repeatability test is shown in
Figure 7b. It is well known that the time required for the sensor response or recovery process resistance to reach 90% is defined as response or recovery [
29]. Response time and recovery time of Ag/ZnO-2 humidity sensor are 36 s and 6 s, respectively. The faster response and recovery speed of Ag/ZnO-2 is mainly due to the abundant oxygen vacancies on its surface, which causes a large number of water molecules to be decomposed quickly. In addition, the distribution of Ag/ZnO-2 is uniform, and the absence of agglomeration will also cause water molecules to quickly detach from the surface of the material. We have continuously tested the adsorption and analysis process of Ag/ZnO-2 humidity sensor in two cycles, and we can see that adsorption and desorption process of the sensor in the two cycles are almost the same, which shows that Ag/ZnO-2 humidity sensor has good repeatability. All the above properties show that Ag/ZnO-2 humidity sensor has the potential to develop a high-performance humidity sensor, and also provides a new idea for the preparation of a new humidity sensor.
Figure 8 shows the resistance response curve under different lights. This section studies the effects of light with different wavelength on the resistance of Ag/ZnO-2 and judges the sensing characteristics of the humidity sensor from the situation of the resistance response curve, taking into account that the use of different illuminations can make the surface of the sample obtain a certain amount of energy, and that a small amount of photoelectrons may be generated on the surface of the sample, thereby increasing the conductivity of ZnO. In the experiments, visible (5 W), red (5 W, 546 nm), blue (5 W, 465 nm), and ultraviolet light (5 W, 365 nm) were used to illuminate Ag/ZnO-2 sensors at different relative humidity. We found that in the absence of light, the sensors showed the best response and the best linearity, reflecting that the light affects the performance of the Ag/ZnO-2 humidity sensor, and in the absence of lighting, the synthesis cost of the sensor can undoubtedly be reduced.
The humidity sensing mechanism of Ag/ZnO-2 was studied by complex impedance spectra, as show in
Figure 9. At low humidity (11%, 33%, 54% RH), complex impedance spectrum approaches a semicircular shape. In this process, a small amount of water molecules is adsorbed on Ag/ZnO surface in forms of physical and chemical adsorption. The rich oxygen vacancies on the Ag/ZnO-2 surface can accelerate the dissociation of adsorbed water molecules into OH
− and H
+. At this time, a small amount of H
3O
+ will be formed, and the protons will conduct jump on the Ag/ZnO surface. When the humidity reaches 75%, 85% and 95% RH, complex impedance spectrum becomes smaller in the semicircle in low frequency region and a straight line gradually appears at the tail. At this time, when a large amount of water molecules is adsorbed on Ag/ZnO-2 surface in the form of chemisorption, a continuous water film is formed. With the increasing number of water molecules and the formation of an ion transport mechanism, H
3O
+ continues to migrate onto the surface of the material, enhancing conductivity of the material, thereby improving the performance of the sensor.