*3.1. Structure and Morphology Characterizations*

The XRD pattern of the MAPbI3 powder was shown in the Figure 4a. The peaks at 14.1◦, 28.4◦, 31.7◦, 40.5◦, and 43.0◦ are assigned to (110), (220), (310), (224) and (314) planes, respectively. These peaks agree perfectly with MAPbI3 perovskite tetragonal structure (The COD ID of MAPbI3 is 2107954) [16]. The average grain size of the MAPbI3 powder was clarified to be ~1.39 μm from the SEM image shown in the inset of Figure 4a. Figure 4b shows the *P-E* loops of the MAPbI3 measured at room temperature with a frequency of 10 Hz under various voltages of 15, 30, and 45 V. The thickness of the film is 450 mm. The ellipse-shaped *P-E* loops indicate that the MAPbI3 exhibits no ferroelectric behavior macroscopically at room temperature. This result is consistent with that reported in Refs. [3,17].

**Figure 4.** (**a**) XRD pattern and SEM image of MAPbI3 powder; and (**b**) *P-E* loops of MAPbI3 film measured at the frequency of 10 Hz under the voltage of 15, 30, and 45 V.

### *3.2. Humidity Sensitive Properties*

The humidity sensitive performance of the MAPbI3-based resistive sensor was studied by testing the impedance upon exposing the sensor to various RH levels with the a duration time of 5 min in each level. Figure 5a presents the impedance as a function of RH level of the MAPbI3-based sensor at different frequencies. The impedance decreases with increasing measurement frequency. The curve measured with 100 Hz shows the largest impedance variation. Hence, 100 Hz is chosen as the optimum measuring frequency and will be used in the following part.

**Figure 5.** (**a**) The impedance as a function of RH level of the MAPbI3-based sensor at different frequencies; (**b**) the contrast between the MAPbI3-based sensor and other impedance-type humidity sensors in published literature.

Based on the data in Figure 5a, the humidity sensitive response (S) of the sensor can be calculated according to the relation [18,19]:

$$\mathbf{S} = Z\_{\mathbf{d}} / Z\_{\mathbf{h}} \tag{1}$$

where *Z*<sup>d</sup> and *Z*<sup>h</sup> are the impedance values measured at 11%RH and at a specific RH level, respectively. The MAPbI3-based sensor showed a superior sensitivity of S = 5808. Figure 5b displays a contrast of sensitive response between the MAPbI3-based sensor and other impedance-type sensors reported in the literature [12,20–24]. The comparison highlights that the MAPbI3 shows the largest sensitivity.

For the purpose of fully characterizing the performance of the MAPbI3-based sensor, the hysteresis and recovery/response curves of the MAPbI3-based sensor were tested at 100 Hz. The results were given in Figure 6a,b, respectively. The hysteresis curve was acquired by switching the sensor between the containers with the different RH levels of 11, 33, 54, 75, 85 and 94% in turn, and then shifting back. After an exposure duration of 5 min in each of the RH levels, the impedance was recorded under the optimum frequency of 100 Hz. Based on the measured impedance values, the humidity hysteresis values can be reckoned by the following formula [25]:

$$
\left[\frac{\log(Z\_{\text{ads}}) - \log(Z\_{\text{des}})}{\log(Z\_{\text{ads}})}\right] \times 100\% \tag{2}
$$

where *Z*des and *Z*ads represent the impedance value of the desorption and the adsorption processes, respectively. A hysteresis value of 6.76% is obtained at 11%RH for the MAPbI3 based sensor. The physical adsorption, which is toilless to be desorbed in MAPbI3 material, is much larger than the chemical adsorption that can be weakly influenced by environment humidity at low humidity level, which results in good hysteresis behavior of the sensor at low RH levels. Studies have shown that there is a steep increase in the water uptake in the RH level beyond 80%RH [26]. Therefore, the sensor still exhibits notable hysteresis under high humidity levels because the physical adsorption is greatly increased. Figure 6b displays the response and recovery time for the MABbI3-based sensor recorded between 11% and 94%RH. The result shows that the response and recovery times are 31 s and 148 s, respectively. Response/recovery time under other relative humidity levels are listed in Table 1.

**Figure 6.** (**a**) Hysteresis behavior and (**b**) recovery/response time of the MAPbI3-based sensor (*Z*/*Z*<sup>0</sup> is the normalized impedance to the initial impedance *Z*0).

**Table 1.** Hysteresis and response/recovery time under different RH levels of the MAPbI3-based sensor.


The repeatability of the sensor was measured by switching the sensor between 11 and 94%RH levels for 5 cycles. The stability test was conducted over a period of 135 days. The results of repeatability and stability were shown in Figure 7a,b, respectively. After 5 cycles, the impedance under 11%RH remains 88.13% of the initial value, while under 11%RH, the impedance hardly changes. This result indicates that the sensor shows satisfactory repeatability. The impedance curves shown in Figure 7b remain almost constant, revealing that the sensor exhibits good long-term stability.
