1 Introduction

Energy harvesters have become significantly popular and crucial due to accelerated increase in demand of self-sustaining devices. Self-powered micro scale energy harvesting devices is a promising solution to replace conventional batteries that have limited lifetime, bulky and contain hazardous chemicals. Several publications covering wide range of energy harvesting applications have been published recently (Paradiso and Starner 2005; Kok et al. 2008; Liu et al. 2008; Choi et al. 2006).

Energy can be scavenged from various environment sources such as solar, wind, light and vibration. Solar energy is frequently used but it is not an attractive source for indoor environments (Kim et al. 2009). Among these methods, mechanical vibration and acoustic sources provide the most potential solution for energy harvesting. There are three transduction mechanisms for vibration based energy harvesting, piezoelectric, electromagnetic and electrostatic transduction (Roundy et al. 2005). Of the three mechanisms, piezoelectric transducers have been reported to produce the highest power output for a given size (Roundy et al. 2005). A piezoelectric energy micro-generator typically harvests mechanical energy or vibrations and converts it to electrical energy through piezoelectric effect. The micro-cantilever structure is typically used due to its high power output and compatibility with microelectromechanical system (MEMS) fabrication processes (Kim et al. 2009). Fabricated MEMS cantilever generators have been reported to produce output in the μW when harvesting mechanical vibrations in the kHz range. Piezoelectric micro-generators are the perfect candidate to replace batteries as they are easily fabricated, are silicon compatible and demonstrate high efficiencies for mechanical to electrical energy conversion (Horowitz et al. 2005).

One of the niche applications is aero acoustics, where wasted, high dB, high frequency sound generated by aircrafts can be transformed into useful energy. The sound generated by the unsteady aerodynamic forces on propellers is categorized in the domain of aero acoustic. Noise generated during take-off and approaches to landing are the main noise contributions that determining the airport noise levels (Sheplak et al. 1999). Generally, the aircraft engines have noise generated from its engine outlet, slat noise, landing gear, nose wheel, fan exhaust, turbine, fan inlet and jet exhaust. The Aircraft Aerodynamics and Design Group at Stanford University (1999) has the details of aircraft noise is in the range of high dB noise level of 60–110 dB as shown in Fig. 1.

Fig. 1
figure 1

Graph noise sources versus maximum noise levels in dB

Full size image

Noise pollution is of increasing community concern and is especially dominant during aircraft landing and take-off. Exposure to such high dB noise can negatively impact human health since it increases the risk of cardiovascular diseases; causes sleep disturbance and psychological effects. There is a need for miniature acoustic sensors for noise detection system which monitors these hazardous levels of noise.

Aircraft noise monitoring systems deployed on wireless sensor networks can be utilized to trigger alarms when the noise quality level is above the safety standards. The self-sustained wireless sensor network is powered by acoustic energy harvester which is driven by aircraft noise vibration. These acoustic energy harvesters that convert acoustic vibrations into electrical power can be fabricated using microelectromechanical system (MEMS) technology. Such implementation techniques shrink the size of the wireless sensor nodes, leading to a reduction in unit cost for large volumes.

The range of frequency applied from the ambient acoustic as shown in Fig. 2. Horowitz 2005 designed a 50 kHz micro machined piezoelectric diaphragm microphone with a circular piezoelectric ring. The diaphragm transduces acoustical pressure into mechanical deformation, while the piezoelectric material transduces that mechanical deformation into electrical signal (charge or voltage).

Fig. 2
figure 2

Frequency range for acoustic devices

Full size image

This paper details the design of a piezoelectric cantilever beam micro generator for acoustic energy harvesting application. The rest of the paper is organized as follows. Section 2 explains the theoretical design of the piezoelectric working principle and piezoelectric material selection. Section 3 presents the simulation modelling of piezoelectric micro generator. Sample preparation, characterization and preliminary experimental setup and results are presented in Sect. 4. Finally the conclusion is given in Sect. 5.

2 Theoretical design of the piezoelectric micro generator

2.1 Piezoelectric working principle

Piezoelectricity is an electromechanical phenomenon exhibited by a certain solid state materials that converts mechanical energy into electrical energy and vice versa. Electric polarization is developed when there is an applied mechanical stress or strain. Such phenomenon is known as direct piezoelectric effect. Conversely, the reverse piezoelectric effect describes the production of stress and/or strain when an electric field is applied (Horowitz et al. 2005).

A piezoelectric micro generator demonstrates the direct piezoelectric effect where acoustic energy or mechanical vibrations are converted into electrical energy. It has been reported that the cantilever structure with proof mass at the end is the most promising structure for the energy conversion (Roundy et al. 2005). The piezoelectric energy conversion can be described using the equivalent linear spring mass system as shown in Fig. 3 where z = x − y is the net displacement of the mass, M is the lumped mass, K is the spring constant and C is the damping coefficient (Kim et al. 2009).

Fig. 3
figure 3

a Schematic representation of the cantilever beam at one fix end b first order model of a resonant system

Full size image
$$ M\ddot{z} + C\dot{z} + Kz = - M\ddot{y} $$
(1)

A simplified model of piezoelectricity involves the concept of anions and cations moving in opposite direction under the influence of an electric field or mechanical force (Johnson 2005). The ion charges are neutralized and balanced in the unstrained piezoelectric crystal. When the cantilever is excited due to ambient vibration, the piezoelectric layer deforms and mechanical stress is applied to the cantilever beam. The stress induced in the beam will strain the piezoelectric crystal and generate surface charges due to unbalanced ions. The charges generated on the surface of the piezoelectric crystal are proportional to the potential difference which is a function of the changing stress (Johnson 2005). The charge is measured through the electrodes on top and bottom of the piezoelectric layer which is then converted to AC voltage (Pan et al. 2010)

If we assume that the micro generator operates at resonance frequency, the resonant frequency of spring mass system, ωn is shown in Eq. 2 where M eq is the equivalent mass and K eq is the equivalent stiffness. At resonance frequency, the cantilever vibrates with maximum deflection. Consequently, maximum charges and potential difference will be produced at the resonance frequency.

$$ \omega_{\text{n}} = \sqrt {{\frac{{K_{\text{eq}} }}{{M_{\text{eq}} }}}} $$
(2)

The equivalent stiffness for the cantilever beam can be calculated as shown in Eq. 3 where E is Young modulus (GPa), I is the moment of inertia, and L is the length of the beam.

$$ K_{\text{eq}} = {\frac{3EI}{{L^{3} }}} $$
(3)

The moment of inertia of the rectangular cantilever beam can be written as shown in Eq. 4 where B beam is the width of the beam and H beam is the height of the cantilever beam.

$$ I = \frac{1}{12}B_{\text{beam}} *H_{\text{beam}}^{3} $$
(4)

The equivalent mass, m eq for the cantilever beam without any proof mass and with proof mass are given in Eqs. 5, 6 and 7 where ρ is density of the material, W beam is the width of the cantilever beam, H beam is the height of the cantilever beam, L beam is the length of the cantilever beam, M m is the mass of the proof mass at the end of the cantilever beam and M beam is the mass of the cantilever beam.

$$ m_{\text{eq}} = \frac{1}{3}\rho *W_{\text{beam}} *H_{\text{beam}} *L_{\text{beam}} $$
(5)
$$ m_{\text{eq}} = M_{\text{m}} + \frac{33}{140}M_{\text{beam}} $$
(6)
$$ M_{\text{beam}} = \rho *W_{\text{beam}} *H_{\text{beam}} *L_{\text{beam}} $$
(7)

At resonance frequency, the output power is shown in Eq. 8 where P is output power, m is equivalent mass, k is the equivalent stiffness, R is the load resistance, C b is the capacitance of piezoelectric bimorph, k 31 is the piezoelectric coupling coefficient, t c is the thickness of the piezoelectric material, C p is the elastic constant of the piezoelectric, ε is the dielectric constant of the piezoelectric material, A in is the acceleration and ζ is the dimensionless damping ratio which represents the viscous loss from the system. Only the first mode of resonance frequency is considered since it provides the maximum displacement and consequently the maximum output power (Roundy et al. 2005).

$$ P = {\frac{{m^{2} }}{2k}}*{\frac{{RC_{b}^{2} \left[ {{\frac{{2k_{31} t_{c} }}{{k_{2} }}}} \right]^{2} {\frac{{C_{P} }}{\varepsilon }}A^{2}_{\text{in}} }}{{\left( {4\zeta^{2} + k_{31}^{4} } \right)\left( {RC_{b} } \right)^{2} k + 4\zeta^{2} k_{31}^{2} RC_{b} \sqrt {km} + 4\zeta^{2} m}}} $$
(8)

2.2 Piezoelectric thin film as the material selection

Suitable piezoelectric material is characterized by the product of piezoelectric energy constant (g) and piezoelectric strain coefficient (d) given as | d.g | (Kim et al. 2009). Piezoelectric material can be divided into five types, polycrystalline ceramics, single crystal materials, polymers and thin films (Hyunuk Kim et al. 2009). Piezoelectric thin films offer significant advantages such as generation of large motions, low hysteresis, high sensitivity with wide dynamic ranges and low power requirements (Trolier-Mckinstry and Muralt 2004). The parameters of interest for piezoelectric thin films are piezoelectric strain coefficient, coupling coefficient and relative dielectric constant. The most important parameter is coupling coefficient which describes the effectiveness of energy transformation from electrical to mechanical energy.

The choice of the piezoelectric thin film material also depends on deposition methods, process complexity and CMOS compatibility. Common piezoelectric thin films are zinc oxide (ZnO), aluminium nitride (AlN), and lead zicronate titanate (PZT) (Horowitz 2005). Table 1 shows that PZT provides highest coupling and strain coefficient compared to ZnO and AlN thin films. However, PZT deposition is hazardous due to lead contamination. AlN and ZnO requires relatively low deposition temperature, has high piezoelectric coupling coefficient and excellent bonding to substrate materials such as silicon (Nordin 2008).

Table 1 Comparison of piezoelectric thin film materials for energy harvesting application
Full size table

ZnO has become a material choice for piezoelectric energy harvesting due to highly resistive c-axis oriented ZnO films. Doping was found to give pronounced effect on both microstructure and piezoelectric properties of ZnO by increasing the longitudinal piezoelectric coefficient (d 33) (Bdikin et al. 2010). Many researchers have worked on doping ZnO n-type using group III elements like aluminium (Al), gallium (Ga) and indium (In) (Hamid 2009). This process was found to produce ZnO films with more n-type conductivity, improving its transparency and conductivity. Doping of ZnO had improved not only the electrical properties of the film but also its thermal stability. Aluminium is found to be the best dopant as it produces films with the highest conductivity and transparency as compared to any other dopant (Hamid 2009). Therefore, aluminium doped zinc oxide (AZO) was chosen as the piezoelectric material for this work.

3 Simulation modeling of piezoelectric micro generator

The prototype cantilever structure consists of three layers namely: stainless Steel, AZO, and copper electrode as shown in Fig. 4. The substrate is stainless steel and the chosen piezoelectric layer is AZO. The copper was deposited on top of the piezoelectric layer. The cantilever beam is fixed at one end. The device dimensions are shown in Table 2.

Fig. 4
figure 4

Piezoelectric cantilever beam design

Full size image
Table 2 Device dimension
Full size table

3.1 Material properties

The three material properties matrix that are crucial for analysis of piezoelectric thin films are stress coefficient (d)t, dielectric constant (ɛ), and stiffness matrix (C), are shown in Eqs. 9, 10 and 11, respectively. The material properties for each layer of the piezoelectric micro generator are summarized in Table 3.

Table 3 Material properties
Full size table
$$ \left[ d \right]^{t} = \left[ {\begin{array}{*{20}c} 0 & 0 & {d_{31} } \\ 0 & 0 & {d_{31} } \\ 0 & 0 & {d_{33} } \\ 0 & 0 & 0 \\ {d_{15} } & 0 & 0 \\ 0 & { - d_{15} } & 0 \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} 0 & 0 & { - 5.43e - 06} \\ 0 & 0 & { - 5.43e - 06} \\ 0 & 0 & {1.167e - 05} \\ 0 & 0 & 0 \\ {1.1344e - 05} & 0 & 0 \\ 0 & { - 1.1344e - 05} & 0 \\ \end{array} } \right]{\text{C/m}}^{2} $$
(9)

While, the piezoelectric zinc oxide dielectric constant in matrix form is in Eq. 10.

$$ [\varepsilon ] = \left[ {\begin{array}{*{20}c} {8.55} & 0 & 0 \\ 0 & {8.55} & 0 \\ 0 & 0 & {10.2} \\ \end{array} } \right] $$
(10)

Piezoelectric zinc oxide stiffness matrix (Young modulus) is shown in Eq. 11.

$$ [C] = \left[ {\begin{array}{*{20}c} {20.97} & {12.11} & {10.51} & 0 & 0 & 0 \\ {12.11} & {20.97} & {10.51} & {} & 0 & 0 \\ {10.51} & {10.51} & {21.09} & 0 & 0 & 0 \\ 0 & 0 & 0 & {4.43} & 0 & 0 \\ 0 & 0 & 0 & 0 & {4.24} & 0 \\ 0 & 0 & 0 & 0 & 0 & {4.24} \\ \end{array} } \right]*10^{10} N/m^{2} $$
(11)

The 3D view and fabrication steps are shown in Fig. 5 and Table 4, respectively. Stainless steel acts as the substrate to control the bow of the structure. AZO was chosen as the piezoelectric layer with thickness of 36.4 μm. Copper electrode was deposited on top of the piezoelectric layer. The size of the cantilever is 20,000 μm × 20,000 μm with thickness of 58.6 μm.

Fig. 5
figure 5

3D view of piezoelectric cantilever beam design

Full size image
Table 4 Fabrication steps in simulation
Full size table

3.2 Finite element analysis

The device was modelled and simulated using a finite element (FEM) simulator, CoventorWare© which has the capabilities of producing a 3D model based on the layout and the fabrication process. Finite element modelling analysis was used to predict the resonance frequency of the maximum displacement of vibration which therefore provides maximum output power for piezoelectric micro generator (Ralib et al. 2010). The applied boundary condition is at the fixed end of the cantilever beam and pressure is applied at the bottom of the piezoelectric layer as shown in Table 5. Once the model specifications have been completed, it was meshed using Manhattan Bricks, which are preferred for rectangular structures. The mesh details are summarized in Table 6.

Table 5 Surface boundary condition
Full size table
Table 6 Meshing properties
Full size table

3.3 Simulation result and discussion

MemMech piezoelectric modal analysis was done to compute the resonance modes of the structure. Simulation result shows that the highest displacement is at the z direction as shown in Fig. 6. Fifty modes were simulated, where the mode with the highest displacement was identified as the mode of interest. Figure 7 displays various vibration mode of piezoelectric energy harvester. Mode 48 of 32.8 kHz is of the most interest. The following harmonic analysis will be based on this frequency.

Fig. 6
figure 6

MemMech piezoelectric analysis

Full size image
Fig. 7
figure 7

MemMech modal analysis

Full size image

Harmonic analysis computes a displacement solution based on a user range of input harmonic frequencies to find the best resonance frequency which has maximum displacement (Ralib et al. 2010). A total of 50 frequencies at the range of 32.3111–34.4667 kHz were simulated. Simulation results shows that maximum z axis deflection of 1.735 μm was found at the resonance frequency of 32.92 kHz. Figure 8 shows the graph of displacement versus frequency generated. The graph shows expected sharp change in the displacement at z direction as the frequency reached 32.92 kHz.

Fig. 8
figure 8

Piezoelectric harmonic analysis displacement versus frequency

Full size image

4 Fabrication of piezoelectric micro generator

Figure 9 shows the process flow for sample preparation steps of piezoelectric micro generator. The stainless steel was chosen as the substrate. AZO was sputtered using ULVAC sputtering machine. Copper material was pasted on top of the AZO as the electrode using carbon paste. The sample was characterized by X-ray diffraction to evaluate the orientation and crystallization qualities. The scanning electron microscopy (SEM) was used to measure the thickness of the sample and evaluate its surface morphology.

Fig. 9
figure 9

Fabrication steps for sample preparation

Full size image

4.1 Aluminium doped zinc oxide (AZO) RF magnetron sputtering

Piezoelectric deposition techniques consist of various method and techniques such as sputtering methods, chemical vapor deposition (CVD) techniques, and spin on sol–gel processing. Highly oriented AZO were prepared by RF Magnetron sputtering on stainless steel substrate. The optimized parameters for RF magnetron sputtering are summarized in Table 7. Six samples with same parameters were prepared as shown in Fig. 10.

Table 7 RF magnetron sputtering parameters
Full size table
Fig. 10
figure 10

Sample of aluminium doped zinc oxide sputtered on stainless steel

Full size image

4.2 Sample characterization result

4.2.1 Crystalline structure

Good piezoelectric ZnO thin films are expected to show c-axis orientation in 002 direction, band gap of 3.37 eV, refractive index 2.008 and high transparency in the visible region (Gokhale et al. 2008). X Ray diffraction was performed on the samples and its intensity versus 2 theta. X Ray diffraction confirms that the thin films has a preferred orientation with c-axis perpendicular to substrate (Trolier-Mckinstry and Muralt 2004). The c-axis orientation at 002 peak was clearly shown at 2 theta values of 34.4° which is very close with pure ZnO crystal orientation (34.45°) as shown in Fig. 11. The diffraction angle of 002 plane of AZO thin film was shifted towards lower angle and smaller to standard ZnO crystal due to ionization of ZnO rather than Al doping. (Jeong et al. 2003). The intensity of AZO at 2 theta of 34.4° is 2,108.

Fig. 11
figure 11

X-Ray diffraction result intensity versus 2theta

Full size image

4.2.2 Surface morphology

The surface morphology of the AZO thin film structure was examined using scanning electron microscopy at 500× magnification and is shown in Fig. 12a. The thickness was measured through the cross section of the sample as shown in Fig. 12b. The thickness of AZO thin film is 38.4 μm and the thickness of stainless steel is 16 μm.

Fig. 12
figure 12

a Surface morphology at ×500 magnifications. b Cross-section thickness of the sample

Full size image

4.3 Experimental setup and results

The measurement system was schematically shown in Fig. 13. The sample was placed on the ultrasonic cleaner that act as the vibration shaker. The frequency of the vibration from the ultrasonic cleaner is 30 kHz and the power is 80 W. The generated voltage was measured using digital multimeter UT803.

Fig. 13
figure 13

a Schematic illustration of voltage generation of piezoelectric micro generator. b Preliminary experiment setup for piezoelectric micro generator

Full size image

4.3.1 Experimental result and discussion

Figure 14 shows the open circuit voltage versus time, RMS output voltage versus time and calculated RMS output power at load resistance of 1 kΩ versus time, respectively. The result in Table 8 shows that the maximum output of open circuit voltage produced is approximately 345.4 mV. The closed circuit voltage (CCV) is measured as a function of load resistance and the input frequency of 30 kHz. The output power is calculated by assuming the load resistance is 1 kΩ. The calculated peak power is calculated with load resistance of 1 kΩ is approximately 59.63 μW. This measured power level is satisfactory to power a miniature wireless acoustic sensor nodes for high dB noise detection monitoring system generated by aircrafts.

Fig. 14
figure 14

a Open circuit voltage versus time, b RMS Output Voltage versus time, c calculated output power at load resistance of 1 kΩ versus time

Full size image
Table 8 Peak voltage and peak power for experiment
Full size table

5 Conclusion

Piezoelectric vibration energy harvester is a very attractive method to power wireless sensor networks due to their higher power density, robustness and simplicity. This paper presents the simulation and experimental study of a piezoelectric micro generator for aero acoustic application. Aluminium doped zinc oxide was chosen as the piezoelectric material due to its excellence bonding properties to substrate material and high piezoelectric coupling coefficient. Finite element simulation analysis was done to simulate the resonance frequency which provides maximum displacement of vibration and maximum output power. For sample preparation, aluminium doped zinc oxide piezoelectric thin film was RF-sputtered on stainless steel substrate thin film. Characterization was performed using X-ray diffraction (XRD) and scanning electron microscopy (SEM) to evaluate the piezoelectric qualities and surface morphology, respectively. The c-axis orientation at 002 peak was clearly shown at 2 theta values of 34.4° which is very close with pure ZnO crystal orientation. Experimental measurements indicate that approximately 345.4 mV AC output voltage (open circuit voltage) is produced at vibration frequencies of 30 kHz. This is in accordance with the Coventorware™ simulation result that shows resonance frequency of 32.92 kHz. This measured power level is sufficient to power a miniature wireless acoustic sensor nodes for high dB aircraft noise detection monitoring.