PVDF Based Static Charge Induced Flexoelectric Microphone

Polyvinylidenedifluoride (PVDF)‐based microphones are either categorized as piezoelectric (PE) microphones or flexoelectric (FE)‐induced PE microphones. Based on the production of electrostatic charge and higher sensitivity, here, a static charge induced flexoelectric microphone (SCIFEM) is introduced. In this research, two SCIFEM transducers are fabricated, one by sandwiching P(VDF‐TrFE) laminate between pair of two aluminum electrodes wrapped in stretched paper sheets (called Device 1) and the other by sandwiching P(VDF‐TrFE) laminate between a pair of two Al electrodes wrapped in stretched rubber sheets (called Device 2). Devices 1 and 2 are compared with a simple PVDF‐based FE microphone (called Device 0). In this study, Devices 0, 1, and 2 are characterized for flexoelectric (FE) and charge induction (CI) properties. It is observed that despite its poor FE behavior, higher sensitivity of Device 1 is due to its electrostatic charge induced beta phase present at 1170 cm−1 of Fourier transform infrared spectra. The X‐Ray diffraction spectroscopy is also performed in accordance to the polar beta phase. It is reported that due to the enhancement in beta phase and the unique charge induction property represented at 1170 cm−1 of the FTIR spectra, the performance of SCIFEM microphone is much improved at a lower frequency range.


Introduction
Due to the flexibility, light weight, and high dielectric properties of ferroelectric polymers, scientists are still interested to use them in future technologies. [1] The use of ferroelectrric polymers in practical device applications is challenging due to their relatively low polarization (P), low piezoelectric coefficient (d 33 ), and high electric field (E) requirement. [2,3] To enhance the properties of ferroelectric polymers, researchers either add extrinsic www.advelectronicmat.de structure and associated dielectric loss at high temperature. [26] While the literature reports on many properties of flexible ferroelectric polymers, the FE behavior of these materials is not fully understood. There remain open questions about FE coefficient, dielectric loss, design consideration, and stress conditions that need to be further studied.
In this study, the charge density per unit degree of sonic vibration was measured for Devices 0, 1, and 2. It was reported that due to its stretched paper substrates, the Device 1 developed more electrostatic charges on its surface than Device 2. Although the Device 2 used the rubbers as stretched substrates, it showed much lesser FE and CI properties than Device 1. The acute polarization of these devices was also observed due to activation of domain boundaries as a result of continuous wave front propagation. It was observed that the FE and CI property of SC I F E M were influenced by the stretched substrates and for the paper substrate, its efficiency was much improved due to its electrostatic charge induced beta phase present at 1170 cm −1 of fourier transform infrared (FTIR) spectra. The induced electric charge could also be harvested to charge a battery. The schematics of FE and CI mechanism in response to the sonic vibration in SC I F E M transducer is illustrated in Figure 1.

Acoustic, Dielectric, and Spectroscopic Properties of SC I F E M
Here, we reported P(VDF-TrFE) based SC I F E M transducers presented in Figure 2. The properties of Devices 1 and 2 were compared with Device 0 transducer for the clear insight about its FE and CI behavior toward the sonic vibrations. Figure 2b shows a schematic layout of Devices 0, 1, and 2. The PVDF based devices (i.e., Devices 0, 1, and 2) were studied for the FE and CI behavior initially chosen to perform vibration test through sonic vibration because of its known piezoelectric properties. In each laminate, a pair of aluminium electrodes attached to the P(VDF-TrFE) layer served to collect generated charges. The microphone devices were characterized by driving them with an audio frequency vibrational excitation (0.1 Hz to 20 kHz). A loudspeaker driven by a function generator integrated with amplifier was placed directly below the laminate at the distance of 10 cm from target such that sound waves impinged orthogonally on the surface. The usual polarization hysteresis loop and permittivity constant of P(VDF-TrFE) is shown in Figure 2c,d. It is observed that the sensible sonic vibration for microphones was less than 1000 Hz and it may vary with the toughness of the target surfaces. The output voltage of the devices in response to 100 dB SPL sonic vibration is shown in Figure 2a. Device 1 has a higher peak-to-peak output voltage than either Device 0 or Device 2 across the entire frequency band. The output voltage of Device 1 is higher because of its higher ability to absorb the vibration energy (shown in polar graph in Figure 2g) due in part to film thickness and toughness (shown in Figure 2e [film toughness] and Figure 2f [film thickness]). It has better mechanical collection efficiency-impedance match and lower loss. The overall mechanical impedance characteristics are a function of layer thickness and mechanical loss. The mechanical impedance of the laminates also shows an effect in the shape of the frequency responses of the devices, where none of the devices show much response above 300 Hz. It is also observed in literature that the researchers often use dielectrics such as papers to enhance the piezoelectric properties of transducers. [27] Figure 2h shows the sensitivity of microphone laminates in which it is observed that the sensitivity of Device 1 is much higher than Devices 0 and 2. In this study, it is reported that the sensitivity of Device 0, 1, and 2 laminates is 0.015, 5.54, and 5.05 V pa −1 , respectively at lower frequency 10 −1 ≥ ω ≥10 1 Hz. The enhancement in the performance and fidelity of the Device 1 over the lower frequency range of sonic vibration is correlated with the phase present at 1170 cm −1 of FTIR spectra presented in Figure 2i. The static charge induced beta phase is also confirmed by the X-Ray diffraction (XRD) peak at 2θ = 20° for pure P(VDF-TrFE) is presented in Figure 2j. Therefore, we have reported that the performance of Devices 1 and 2 is better than that of Device 0 due to its static charge induced beta phase of P(VDF-TrFE) which enhanced the performance of SC I F E M transducers.  Figure 3 shows the polarization (µC cm −2 ) versus frequency (ω) hysteresis loop for Devices 0, 1, and 2 from 0.1 Hz to 1 KHz. Here, ω is target sound frequency, ω 0 is reference sound intensity, and Pa is the sound pressure. After applying the frequency from 0.1 Hz to 1 KHz, the devices behave in different but in predictive manner due to its unique FE behavior. A comparison of Pω hysteresis loop of Devices 0, 1, and 2 in Figure 3a-d shows that the polarization of Device 1 is higher than that of either Device 0 and 2. It is due to the paper substrate that helps to enhance the polarization of Device 1. On the other hand, the polarization of Device 2 is less due to the ability to store the static charge for rubber surfaces being less than the papers. [28] The pure P(VDF-TrFE) polymer is a very good ferroelectric but it turns to the weak FE. This is because of its weak flexo-electro-mechanical behavior under high vibration frequency. [29,30] The area under the curve of Pω hysteresis loop is shown in Figure 3e derived in Equation (2) as, [31] ω

Pω Hysteresis Loop of SC I F E M Microphones
where U E is the stored energy density, P is the polarization, ω is the frequency, ω 0 is the reference frequency, and Pa is the sound pressure level in Pa. The energy density of Pω loops shows that the Device 1 is showing higher (U E = 39 µC N −1 ) response than the Device 0 (U E = 10 µC N −1 ) and 2 (U E = 25 µC N −1 ), which shows that using paper substrate improves the performance of SC I F E M transducers. The Figure 3f,g shows the PFM hysteresis loop of Device 0. The actual image of Devices 0, 1, www.advelectronicmat.de and 2 is shown in Figure 3h. The PE coefficient (d 33 ) is reported as 10 pC m −1 for pristine P(VDF-TrFE) laminate.

CI Properties of SC I F E M Transducers
The CI properties of SC I F E M transducers were measured and confirmed by the mathematical model presented in the Supporting Information. The comparison of vibration intensity constant (K) and charge intensity constant (M) in response to vibration angle is shown in Figure 4. Charge density per unit angle of vibration is termed as M (C deg −1 ) and product of M and vibration angle (θ 0 ) is equal to the static charge on the surface of vibrating devices. The quantitative sum of all the static charges produced on the surfaces during high vibration is calculated by taking integral of M × θ 0 from 0.I Hz to 1 KHz. Intensity (K) of the wave-front which is geometrically impinged at the surface of devices is plotted against the change in vibrating frequency at various SPL. Each SPL produces variable pressure in air and influences lamb waves on the surface of devices. [38] Increasing the SPL reduces the slope of K constant. At 100 SPL, slope of K constant is maximum which helps us to rationally maintain its trend. Increasing the SPL increases the sound pressure in air and reduces the slope of K constant. It is similar to the behavior of vibration at high frequency. Therefore, the period of oscillation (T = ω/ω 0 ) is divided by the pressure in air (Pa) to simplify the experiments in order to measure the charge density shown in Figure 4a. Vibration angle in response to the oscillating frequency is plotted in Figure 4b which shows that the Device 2 is having maximum vibration angle but the vibration angle measured for Device 0 and 1 is very less. It is because the Device 2 is made with the stretched rubber substrates and shows high flexibility even at low frequency. Contrarily, the M (C deg −1 ) for Device 1 is higher than Device 0 and 2 shown in Figure 4c because of the other factors such as ability to absorb the mechanical energy and its dielectric behavior in which it makes the Device 1 a perfect capacitor to store the electrical charges. [31] The Figure 4d gives the information of the static charges at different frequency. The maximum static charge is produced from 100 to 250 Hz. The static charge production at the surface of Device 1 is also greater than the Device 0 and 2. It is observed that 10 mC of static charge is produced during the high frequency vibration on Device 1. The quantitative sum of the static charges of
Whirlpool of the polarization is a phenomenon in which the extrinsic stimulus induces the inner ferroelectric domains. [32] The high level of disturbance in structures creates a static charge on the surface of microphone that is eventually translated into acute polarization. [33] The polarization response of these devices is very limited even under high sonic vibrations. [34] It is therefore difficult to visualize these polarization fields on the surfaces of ferroelectrics. The intensity of the scattered electric charges gives rise to electrostatic polarization. [35] The CI results in a change in ambient temperature can be sensed using thermographic imaging techniques. Figure 4h shows the thermographs of Devices 0, 1, and 2 as the excitation frequency changes from 1 Hz to 1 KHz, from left to right. Figure 4g shows the thermograph temperature profile of Devices 0, 1, and 2. It shows the change in temperature versus frequency graphs of Devices 0, 1, and 2, respectively. The change in temperature for Device 0 increases from 1 to 10 Hz and decrease with increasing frequency above 10 Hz. It is due to the scattering or sonication of electric charges in air as the result of higher oscillations. The Device 0 is not using any stretched substrate unlike Device 1 and 2. Therefore, the ability to capture the excessive electrons at the surface of Device 0 is less than those of the Devices 1 and 2. Therefore, Device 0 is showing less polarization in response to the higher sonic oscillations. On the other hand, the temperature profile of Devices 1 and 2 keeps on increasing. This is due to the encapsulation of electric charges during the sonic vibrations and does not allow to scatter in air. This is perhaps one of the reasons that the Devices 1 and 2 show the higher polarization behavior than Device 0 during sonic vibration. [36]

FE Properties of SC I F E M Transducers
The FE property of SC I F E M transducers is reported in Figure 5. It is reported that we can create change in acute polarization in SC I F E M transducers by sound excitation. The change in the acute polarization is the result of electrostatic charge produced at the surface of SC I F E M laminates. Resultantly, the change in the polarization in response to the mechanical strain (cm −1 ) caused by the sound waves in the stretched laminates results in its FE properties. Figure 5a reports the FE effect of Devices 0, 1, and 2, respectively which shows the change in acute polarization in the result of mechanical strain (cm −1 ) caused by the sound waves at different frequency at 100 dB-SPL. It is observed that the FE property of Device 1 is higher than the Devices 0 and 2 due to nature of its stretched substrates in fabrication of SC I F E M transducers. In Device 1, the paper is used as stretched substrates and due to its enhanced CI properties, the FE property is also improved. The FE coefficient (µC m −1 )

www.advelectronicmat.de
is calculated by measuring the slope of change in polarization in response to the mechanical strain (cm −1 ). It is reported that the FE coefficients of Devices 0, 1, and 2 are 0.6, 1, and 0.5 µC m −1 respectively, shown in Figure 5b. The SC I F E M transducer is also integrated in facemask and intensity of signals is measured, presented in Figure 5c. The facemask acts as a SC I F E M microphone for the speech recognition. After wearing the SC I F E M facemask, it senses the human voice and changes the ac voltage as a result of sonic vibration. The change in the vibration intensity induces the electrostatic charges at its surface and results in the change in polarization. The electrostatic potential in response to the sonic vibration is measured by the oscilloscope. Figure 5e shows the schematic illustration of FE and CI properties of SC I F E M facemask in which the change in the vibration induces the electrostatic charges and then electrostatic polarization translates in its FE properties. The real image of facemask SC I F E M transducer is shown in Figure 5d. In this research, it is reported that despite having lower FE properties of SC I F E M devices, its sensitivity and fidelity toward sound recognition is much improved due to its static charge induced beta phase presented at 1170 cm −1 of FTIR spectra.

Discussion
In this study, we reported that by vibrating paper and rubber substrates through voice, we can create the change in polarization of SC I F E M transducers. This change is triggered by the higher frequency vibration induced by sound. The electric charge density per unit degree of vibration is measured and it is found that the Device 1 has an ability to generate the static charge on its surface higher than Devices 0 and 2. It is reported that we can produce 0.34, 10, and 3.86 mC static charge on Device 0, 1 and 2 laminates, respectively at (10 −1 Hz ≤ ω ≤ 10 3 Hz). The FE effect of Device 0, 1, and 2 was measured and compared at different frequency. The increasing trend of FE behavior of P(VDF-TrFE) was observed and it was reported that we can get µ 33 = 0.6, 1, and 0.5 µC m −1 for Device 0, 1 and 2 respectively. It was reported that despite its poor FE behavior, we got higher fidelity of SC I F E M transducers due to its higher static charge density. Last, the enhanced performance of Device 1 was reported for real time speech recognition with higher sensitivity S = 5.54 V Pa −1 at 100 dB-SPL due to the static charge induced beta phase presented at 1170 cm −1 and FTIR spectra. In this research, we concluded that the paper as a stretched substrate in Device 1 was proved to be good candidate in developing the static charge induced microphones due to its static charge induced beta phase presented at 1170 cm −1 of FTIR and 2θ = 20° of XRD spectra.

Experimental Section
Materials Preparation: In order to make a microphone device, PVDF based polymers were needed. Therefore a 70/30 PVDF-TrFE was purchased from Piezotech Inc. The PVDF powder was dissolved in acetone and stirred at 500 rpm for 1 h. The solution was kept overnight to remove the trapped gases and filtered. The solution was then dispensed to make freestanding films. The thin films were sandwiched between top and bottom aluminum electrodes to maintain the electrical connectivity in order to fabricate as piezoelectric microphone.

www.advelectronicmat.de
Fabrication of Microphone: Three microphone devices named Devices 0, 1, and 2 were fabricated. The Device 0 was fabricated by cutting the thin film PVDF-TrFE in circular pattern of 6.5 cm 2 radius. Top and bottom electrodes were pasted and the whole device was mounted on the top of equipment to FE coefficients. For the Device 1, a simple sheet of paper was taken and a piezoelectric microphone was pasted between top and bottom paper substrates maintaining its electrical connectivity. This arrangement was fixed on a circular disc of 6.5 cm 2 radius to place on the top of sonic vibrator (Multy Net). For the Device 2, a balloon was taken to make a stretched surface on 6.5 cm 2 radius disc, and a piezoelectric microphone was pasted between the two stretched rubber balloons maintaining its electrical connections. A proper ground connection was made to avoid any stray charges developing on the surface.
Measurement of Vibration Angle: A tiny laser was placed on the top of devices and the sonic oscillations performed. The laser projection was far enough to see the tiny vibrations. Then, the laser pointer was spotted in camera and the amplitude of laser scan on the target was measured. Vibration angle (θ 0 ) was calculated by simple trigonometric expressions.
Average Area of Vibrations: The average area of vibrations was measured by sprinkling the table salt on the circular devices. During test, the oscillating area of well ground table salt changed and clearly showed the average area designated to each frequency of sonic vibrations.
Thermographic Analysis: The thermographic analysis was done by using the handheld thermographic equipment (FLUKE Thermographic Imager). The SC I F E M device was excited with sound and the thermographs were taken in vertical position by placing the thermograph camera near to the resonant surface. Full infrared (IR) spectrum thermographs were taken in order to get the better contrast of radiating electric field.
PK Hysteresis Loops and k 33 , M 33 : The polarization versus K hysteresis loops were measured by developing an equipment with the help of oscilloscope (UNI-T UTD2025CL), function generator (UNI-T UTG1005A), multimeter (UNI-T UT52), and data acquisition card (NI USB-6009 779026-01). The LABVIEW program was created to control the I/Os and generate the continuous data. Then, the continuous data was plotted against P and K to create the PK hysteresis loops. The first loop was measured by increasing the period of oscillations and for the second loop, the device was poled at high frequency (i,e. 1000 Hz) for 30 s and then, the second loop measured. While reaching the 0.1 Hz frequency, the harvested energy of PE device was drained to return it to zero at 0 Hz to complete one loop. Due to the acute mechanical poling, the time-lapse of 30 s was suggested to impart stability of polarization during Pω hysteresis loops. In this research, it was observed that the delay of 30 s at 10 3 Hz for mechanical poling of flexoelectric device helped to stabilize the polarization. A researcher in 2022 reported the mechanical nanoscale poling of P(VDF-TrFE) films. He suggested that the in-plane polarization could be mechanically poled along any chosen direction after vertical electric poling. [37] Measurement of Static Charge: First, the static charge (q) was measured by measuring the electrostatic charge meter (Digital static charge meter Z205). Second, the calculations were confirmed by performing similar test with the help of lab made Versorium scope on the principle governed by William Gilbert. The electroscope was built in analogy of static charge detection. The deflection in aluminum foils was indicator of static charge production during vibration. Deflection in aluminum foil was measured and plotted to get the clear insight of static charge density. In an analogy, it was supposed that higher the deflection, greater would be the density of charge. Third, the SCIFEM was integrated in the facemask and threw a voice. The change in the frequency of sound created the static charge on its surface and resulted in the sonic oscillation sounds such as undetected alien voice. That's how the static charge production was confirmed by using simple paper-mask.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.