Energy Harvesting Aspects of Irregular Vibrating Forces Acting on Piezoelectric Devices

Introduction

Vibrating energy harvesting systems are commonly implemented by means of piezoelectric materials. Many piezoelectric materials have been studied by researchers during the past years. A piezoelectric material can convert energy between the mechanical and electrical fields through relationships between the stress T, the strain S, the electric E, and the electric displacement field D coupled with elasticity matrix, coupling matrix, and relative permittivity matrix which are characteristics of the piezoelectric material. Piezoelectric materials can be composed by different materials (composed/nano-composed materials) or can be constituted by thin films. Thin piezoelectric films can be etched by chemical wet methods and processed by photolithography in order to fabricate Micro Electro-Mechanical Systems (MEMS). The common piezoelectric thin films studied in literature are Lead Zirconate Titanate (PZT), Zinc Oxide (ZnO) and Aluminium Nitride (AlN). Considering d₃₁ PZT cantilevers at low frequencies, Glynne-Jones et al. (2001), have found a power density of P = 0.017 μW/mm³ at a working frequency f = 80 Hz for an acceleration A = 4g, besides at f = 70 Hz for A = 1gIsarakorn et al. (2008) have found P = 2.5 μW/mm³. At higher frequencies, Fang et al. (2006) have estimated P = 10.84 μW/mm³ at f = 608 Hz for A = 1g, Renaud et al. (2007) have evaluated P = 21.68 μW/mm³ at f = 1.8 KHz for A = 1.9g, Shen et al. (2008) have found P = 3.27 μW/mm³ at f = 462.5 Hz for A = 2g, and, finally, Kim et al. (2008) have appraised P = 0.3 μW/mm³ at f = 870 Hz for A = 8g. Considering d₃₁ Aluminium Nitride cantilever beam, Marzencki et al. (2005) have found P = 0.01 μW/mm³ at f = 204 Hz for A = 8g, and Renaud et al. (2008) have measured P = 0.22 μW/mm³ at f = 320 Hz for A = 0.02g. At lower frequencies of f = 64 Hz, Massaro et al. (2011a) have measured a higher power density of P = 30.20 μW/mm³ at a resonant frequency f = 64 Hz and for A= 2g by considering curved layouts in ring MEMS configuration able to resonate at low frequencies. Generally, in the real cases, the maximum density of power does not correspond to the vibrational force frequencies which is very low. Moreover, the vibrational sources are irregular and are totally different if compared with the controlled vibrational sources used during the experiments. Consecutively, the traditional reading electronic circuits are not so efficient for irregular signals. In this direction we propose to study in this chapter the behavior of freestanding piezoelectric structures vibrating by irregular forces in order to predict the output current signals. This prediction is useful for the design of correct reading electronic circuits. The structures proposed in this chapter takes into account Molybdenum/Aluminium Nitride/Molybdenum (Mo/AlN/Mo) structures, where the metallic Mo layers and the piezoelectric layers can be grown by means a radio frequency (RF) sputtering machine. The sputtering machine is composed by rotating holders containing the target materials to deposit. By means of a Mo target, the first metallic layer is grown on SiO₂ substrate which is centered (before to active the vacuum condition) in order to provide an uniform metallic deposition. By acting on the DC bias parameter it is possible also to change mechanical stress properties which are function of the used substrate material, and, provide surface adhesion properties. The Mo thickness is a function of sputtering deposition time. In order to deposit the AlN layer a pre-sputtering (pre-hating) process is necessary. After this stage it is important to control N₂ and Ar gases during the growth. The last metallic layer deposition is performed by furthermore rotating the sputtering holder again on Mo target, and by waiting before the sample cooling. The layer thicknesses will define the curvature of the final ring/helical freestanding structure as experimentally proved in Massaro et al. (2011b). After the deposition of the three layers a wet chemical process is required. Working in clean room, on the sample is applied a positive photoresist layer (coating) and, successively, the sample is processed by UV radiation and by chemical etching of the Mo (by H₂O₂), AlN (by H₃PO₄) and SiO₂ (by HF) layers as described in Massaro et al. (2011-b). A good approach is to check the surface color during the chemical etching process by means of a microscope: as described in Massaro et al. (2011b), the different observed colors will represent the effect of the etching process at each stage. Finally, an analysis by a profilometer is useful to check the layer thicknesses and an analysis by means of a Scanning Electron Microscope (SEM) is important in order to check the freestanding layout. The piezoelectric coefficient of AlN can be measured using piezoresponse force microscopy as discussed by Shin and Song (2010) or by means (d₃₃ coefficient) of acoustic wave propagation (see Ho et al. (2004)). The vibrating piezoelectric efficiency will be a function of the geometrical and physical properties of the used materials: the induced mechanical properties will define the vibration flexibility and the mechanical force resistance, moreover, the layouts will improve mainly the frequency resonance properties. However, as described in Massaro and Cingolani (2011c), the use of the proof mass allows to operate at lower frequencies if compared with the natural resonant frequencies of the freestanding cantilever beam. This aspect is significant especially when wind/ocean vibrating forces are considered. Usually it is no possible to find in natural sources regular vibrating forces coinciding with the natural resonant frequencies of the structure. For this reason it is very important to first study accurately the typology of source signals. The whole energy harvesting system can be stimulated by different vibrating source signals as discussed in Inmann (2000). The source vibrating signals (represented by carriers) can be modulated by discontinuous functions such as square/step waves representing the noise acting on the vibrating forces present in nature usually characterized by irregular noise signals due to a possible turbulence regime, and, by different low frequency waves (harmonics) acting simultaneously (we think for example to the irregular force of the wind, or of the ocean wave or of the human body motion). The total effect of external vibrating forces (including particular noises) applied to the piezoelectric beam, can be represented by a random noise function. The various directions of the vector forces can define the vibrating turbulence of the whole systems. Moreover, the magnitude of currents generated by the piezoelectric devices decreases in the time domain by means of an irregular damping effect which follows the signal modulation trend. In this chapter the theoretical effect of irregular vibration forces acting on AlN piezoelectric material is mainly discussed. Furthermore, by means of the elastic property together with the torsional degree of freedom, AlN curved layouts could exhibits very high power density at low frequency and for low intensities of vibration forces. The proposed study will be completed by discussing various experimental aspects of measurement.