Simulation, Fabrication and Validation of Surface Acoustic Wave Layered Sensor Based on ZnO/IDT/128° YX LiNbO3

In this study, we report a fully described modeling and simulation of a piezoelectric layered structure for surface acoustic wave (SAW) applications. SAW propagation characteristics were numerically investigated in ZnO/IDT/128° YX LiNbO3. The numerical analysis was conducted using Finite Element Method (FEM) in COMSOL Multiphysics 4.3b platform. The numerical results show that the electromechanical coupling coefficient (K) was maximized to 11.7 % with 1.5 m ZnO and more sufficient acoustic displacement was obtained. To validate the simulation results, a layered SAW device was fabricated. Using conventional photolithography, the LiNbO3 substrate was equipped with two interdigital transducers (IDTs). RF magnetron sputtering technique was used to coat the structure with 1.5 m ZnO layer. The SAW device was characterized electrically using a vector network analyzer and a good correlation was obtained between the experiment and numerical results. Furthermore, the structural and morphological properties of the ZnO film were studied using X-ray diffraction (XRD) and atomic force microscope (AFM) methods. Keyword: FEM; SAW sensor; Coupling coefficient; LiNbO3; ZnO. INTRODUCTION Surface acoustic wave devices (SAWs) are one of the widely used technologies in micro and nano electronics. It was investigated as gas sensors [1], biosensors [2], pressure sensors [3] and many other applications like humidity and magnetic field sensing [4], [5]. Various studies have been conducted to develop layered SAW device with high sensitivity and fast response [6]. A two port layered SAW device is consist of a piezoelectric substrate, transmitter and receiver IDTs, intermediate layer and sensing layer [7] as illustrated in Figure 1. Moreover, much work has been reported using a non-piezoelectric substrate with piezoelectric thin film to generate the surface acoustic wave [8]. To develop a high performance layered SAW device, many factors must be considered in the piezoelectric material to be employed [9]. Piezoelectric material with high phase velocity (p) and high electromechanical coupling coefficient (K) are crucial [10]. In a study reported by Armstrong et al. [11] in ZnO/YZ LiNbO3, more efficient surface acoustic waves can be obtained when using piezoelectric layer on piezoelectric substrate. Furthermore, a study done by Choi et al. [12] shows that adding a dielectric intermediate layer to the SAW device structure can influence the propagating SAW mode(s) characteristics such as velocity and temperature coefficient also can change the permittivity and K of the structure. Therefore, the dielectric layer material properties can highly affect the performance of the SAW device. ZnO is widely used piezoelectric material in different technological applications due to its electrical, optical and structural properties [13] ZnO thin films were investigated for the excitation of acoustic waves in non-piezoelectric substrates such as silicone, diamond and sapphire [14]-[16]. A study conducted by Zadeh et al. [17] shows that using ZnO on an ST-quarts substrate increases the gas sensor sensitivity compared with SiO2. From the above we conclude that a prefabrication modeling and simulation is very important to investigate the layered SAW device performance and gain a better understanding about the propagation characteristics of the structure. Figure 1: Schematic of a layered SAW device. Finite Element Method (FEM) provides a very powerful platform and can be employed to examine the SAW propagation characteristics. In the last recent years, FEM was able to successfully predict the behavioral parameters of different Micro-electro mechanical systems (MEMS) [18][20] and it was employed to investigate the SAW device performance in many applications [21]-[23]. Maouhoub et al. [24] reported a 2D FEM model to investigate Rayleigh waves for SAW devices based on IDT/ZnO/AlN/Si. Tigli et al. [25] reported a 3D modeling and analyzes of surface acoustic waves in COMS technology using ZnO as the piezoelectric material. Ippolito et al. [26] reported a 2D FEM model of layered SAW device using ZnO/XY LiNbO3 with a good agreement between the simulation and the experimental International Journal of Applied Engineering Research ISSN 0973-4562 Volume 11, Number 15 (2016) pp 8785-8790 © Research India Publications. http://www.ripublication.com 8786 results. In this paper, we report a high accuracy 3D modeling and simulation of SAW propagation characteristics in ZnO/IDT/128° YX LiNbO3 using COMSOL Multiphysics 4.3b. The reported model can be investigated in various SAW device applications. By applying an appropriate sensing layer to the structure, this model can be investigative in a multiple applications. The rest of this paper is organized as follows: Section 2 presents the FEM 3D model description for a two port delay line layered SAW device. The experimental procedure and characterization methods are provided in Section 3. The theoretical and experimental results of the SAW device electrical properties are presented in section 4. Finally, the conclusion is provided, along with a number of directions for future work. MODELING AND SIMULATION A 3D work space was selected to build the layered SAW device model in COMSOL Multiphysics version 4.3b. LiNbO3 was used as the SAW device piezoelectric substrate for the generation and propagation of SAW along the Y-axis. For a sufficient analysis, an appropriate number of acoustic waves reflected from the surface boundaries are required [27]. Thus, the substrate length and width was calculated to 804 m and 744 m respectfully to satisfy the nodal density needed to generate 9 SAW cycles. In three-dimensional analyzes, the structure is created on xz-plane and data will be generated in the x, y and z directions. This will increase the number of elements in the structure and extend the solving time and operating system requirements. Therefore, the substrate width was reduced to 2 wavelengths (48 m) rather than 500 m which are the depth of typical substrate. This will not affect the simulation accuracy since SAWs travels with most its energy near the surface and decays after approximately two wavelengths [28]. The operational frequency (fo) for the SAW devices is given by fo=p / where p is the phase velocity, and is the device wavelength correlated to the distance between the electrode fingers (periodicity) [29], [30]. In this model, each IDT was fixed with four finger pairs with periodicity of 24 m and finger width of 6 m enabling the device to work in the VFH range. Figure 2 presents the SAW device model geometry used. Figure 2: SAW device model geometry. The selection of a specific crystal orientation is governed by a set of three Euler angles (α, β and γ), which defines the rotation around the piezoelectric crystal axes (X', Y', and Z'). Rotated system of Euler angles (0°, 38°, 0°) was used for 128° YX-LiNbO3 [31]. Figure 3 shows the rotations specified by the Euler angles sets used in this model. To investigate the ZnO layer thickness influence to the structure phase velocity (p) and electromechanical coupling coefficient (K), a parametric sweep was used in a time dependent study starting at 0.5 m to 5 m with step of 0.5 m. The time step was calculated to 0.24024 ns according to the Courant Friedrichs Lewy (CFL) condition given by [32]: 0.2 = vp . ts h (1) where 0.2 presents the Courant Friedrichs Lewy (CFL) condition, p is the substrate phase velocity, h is the maximum mesh element size and ts is the time step taken by the solver. The total simulation time for 9 SAW cycles was approximately 55 ns. Figure 3: Piezoelectric crystal orientation governed by Euler angles (0°, 38°, 0°). The SAW device frequency response was analyzed by using the Fast Fourier Transaction (FFT) function in MATLAB 2015b. From the SAW device frequency response results, the free and metalized surface velocities (f and m), were calculated by multiplying the operational frequency by the wavelength. The results were applied in Equation (2) to calculate the electromechanical coupling coefficient of the structure [33].