2.1. PENGs preparation
Two set of devices include single layer PENGs and multilayer PENGs based on ZnO NRs and NSs were prepared as schematically shown in Fig. 1 (b,c). For fabricating the single layer ZnO NRs based PENG, PET/ITO was used as top electrode. This sample with the structure of PET/ITO-ZnO NRs-ITO/PET, was named as S-PENG1. In the case of single layer PENG based on ZnO NSs, the PVDF/Au was used as top electrode. This nanogenerator with the structure of PET/ITO-ZnO NSs-Au/PVDF, was named as S-PENG2.
The double-layer PENGs based on ZnO NRs and NSs were prepared as described below. For multilayer layer ZnO NRs based PENG fabrication, a Ni foam was sandwiched between two PET/ITO-ZnO NRs layers. The volumetric porosity of Ni foam is 90–98% with the density of 0.15–0.45 g/cm3.The double-layer device with the structure of PET/ITO-ZnO NRs-Ni foam-ZnO NRs-ITO/PET named as D-PENG1. For the double-layer PENG based on ZnO NSs design, the ZnO NRs have grown on Au in same way as described for PET/ITO. A Ni foam sandwiched between Al-ZnO NSs and PVDF/Au-ZnO NRs. This double-layer sample with the structure of Al-ZnO NSs-Ni foam-ZnO NSs-Au/PVDF, was named as D-PENG2. The dimensions of all devices were 1.5 × 2 cm2.
To complete the devices, Cu wires were connected to the electrodes. An image of prepared D-PENG2 is shown in Fig. 1 (d), illustrating the flexibility of fabricated nanogenerator devices, making them to be an appropriate choice for wearable electronic devices. For the single layer PENGs, the Cu wires were connected to the top and bottom electrodes, which is shown as an example for the S-PENG1 in Fig. 1(e). For the double-layer PENGs, the Cu wires were connected to the top, middle and bottom electrodes as shown in Fig. 1 (f) for the D-PENG1. The output voltage of all prepared devices were measured as depicted in Fig. 1 (g).
2.4. Piezoelectric output
The voltage output of all single and double layer prepared nanogenerators has been measured under cyclic impacts. To evaluate the effect of working frequency on the piezoelectric output of PENGS, an impact experimental setup has been used to trigger PENGs at different frequencies. The employed setup is consisting of an impactor, a load cell and an oscilloscope. The samples were placed on impact stage and the vertical cyclic forces at desired frequencies were applied. The frequencies of the impacts were selected in the range of 1–5 Hz which is the normal frequency range for the human motions. To ensure repeatability of results, all the experiments were repeated three times and the median was considered as the response of the device.
Figure 4(a,b) show output voltages measured from S-PENG1 and S-PENG2, respectively, under the force of 4N at the frequencies of 1, 3 and 5 Hz. For both of the devices, the positive and negative peaks have been observed under compression and release owing to charging and discharging, respectively. As the frequency increased from 1 to 5 Hz, the average peak-to-peak voltage (Vpp) of 1 and 1.8 V were increased up to 7.2 and 10.3 V for S-PENG1 and S-PENG2, respectively. As results show, the voltage from both single layer nanogenerator devices increased with increasing frequency. This enhancement in the output voltage with applied frequency originates from the effect of initial impact speed[30].
To investigate the effect of sandwich design on the performance of nanogerator devices, the output of upper part, lower part and total of double layer PENGs were measured and compared. The upper part refer to when the output measurement is made between electrodes 1 and 2, depicted in Fig. 1 (f). As the same way, the lower part means when the output is measured between the electrodes of 2 and 3. For total double layer measurement, the electrodes of 1 and 3 are connected and the output has been measured between 1 + 3 and 2. Figure 5 shows the output performance of PENG1 and PENG2 at the three state of up, down and total under the forces of 4N at the frequency of about 1.5 Hz. From Fig. 5 (a), the output voltage of PENG1-up and PENG1-down were observed to be 1.7 and 1.5 V, respectively. For PENG1-total, the output voltages from the up and down parts were successfully collected and a total voltage of 3.1 V was obtained. The output voltage of PENG2-total has been enhanced almost 2-fold compared to those related to the up and down parts, owing to the integration of the piezoelectric output from both ZnO nanorods and nanosheets.
Figure 5 (b) illustrate that the output voltage of the upper and lower parts of PENG2 were 1.8 and 2.2 V, respectively. The PENG2-total showed a total output voltage of 4 V. It is clear that the output of nanogenerators in the state of total is almost two times higher than those for up and down parts. The results imply that each nanorod acts as a charging pump when is compressed. As a result, if the polarization of many nanorods under periodic vertical impacts could be synchronized, the output voltage could be constructively integrated.
From the Fig. 5, the effect of the nanostructure of ZnO on the output performance of nanogenerators can also be evaluated. As we know, for PENG1, the structure of the up and down parts are same and are based on ZnO NRs, leading to almost the same output. Whereas, the PENG2 is a hybrid system composed of the upper part based on ZnO nanorods and the lower part based on ZnO nanosheets. As evidence from Fig. 5 (d) and (e), the output of ZnO nanosheets is higher than ZnO nanorods. This is originated from the network structure of nanosheets (Fig. 2 (c)) along with the existence of vacant spaces results in the stress applied on a small area being transferred to the entire network, leading to deformation in relatively large area. Hence, in comparison to 1-D structure, higher surface polarization occurs in a 2-D networked structure which results in greater voltage output owing to enhancement of electromechanical coupling efficiency[31].
Another consequence that can be concluded from the Fig. 4 and Fig. 5 and is related to the effect of structure and material of the electrodes on the output performance. In this study there are three types of PENGs based on ZnO NRs with different electrodes including ITO/ZnO/ITO, ITO/ZnO/Ni and Au/ZnO/Ni corresponding to S-PENG1, D-PENG1-up (or down) and D-PENG2-up, respectively. In ITO/ZnO/ITO and ITO/ZnO/Ni, the top electrodes are different. From Fig. 4 and Fig. 5 it can be concluded that the output performance of ITO/ZnO/Ni is higher than ITO/ZnO/ITO which may be owing to porous surface of Ni foam that can efficiently increase the output performance[32]. From comparing the output voltage of ITO/ZnO/Ni and Au/ZnO/Ni, which are different in bottom electrode, it was observed that the output are not too much different.
To confirm the practical applicability of the prepared PENGs for converting the mechanical energy generated from human motions to electricity, the output of all prepared nanogenerator devices were recorded under different human impacts including finger tapping, hand slapping and hand hammering. It should be noted that the force and frequency selected in Fig. 5 are also similar to human walking conditions as explained in following. Since the diameter of the impactor is 9 mm, the applied pressure on the PENg devices is obtained 62.87 kPa which is similar to the applied pressure of a human foot during walking which is considered to be in the range of 60 to 80 kPa, as explained in our previous work[33]. The frequency of walking is also about 1–2 Hz.
Figure 6 (a, b) illustrate the Vpp of single and double layer PENG1 and PENG2 under finger tapping, hand slapping and hand hammering. As evidence from this figure, for all devices, the output voltage generated by hand hammering is obviously greater than those for finger tapping and hand slapping. These results imply that the prepared PENGs can be utilized as a body motion sensor, detecting human body activity. Such lightweight flexible nanogenerator have also potential application to be utilized in self-powered wearable electronic devices as the multifunctional power sources.
The overall effect of various parameters mentioned above include sandwich design, electrode structure and nanostructure of ZnO on the piezoelectric output of PENGs can be clearly seen from Fig. 6 (a, b). As evidence, porous structure of the Ni electrode can enhance the output about 1.8 fold. Also, the ZnO nanosheets show the output 1.4 times higher than ZnO nanorod. Finally, by applying a sandwich design, the output increases up to two times rather than single layer. Overall, the piezoelectric output has been enhanced about five times for all applied forces.
The working mechanism of single and double-layer PENGs can be explained through the schematic diagram shown in Fig. 7. For single layer structure, Fig. 7 (a), as an external compressing stress is applied on a piezoelectric ZnO nanorod, relative separation of O2− anions and Zn2+ cations results in generation of a piezoelectric field along the ZnO NR driving the electrons flow from the top electrode to the bottom one through the external circuits. In double layer structures, Fig. 7 (b), the generated electric field enforce the electrons to flow from the both sides of PENGs. These electrons gather together, resulting in the increment of output voltages of the double layer nanogenerator compared to the single layer one.