Principle of operation of SAW sensors:
The Piezoelectric effect: Piezoelectricity is the phenomenon exhibited by certain materials allowing them to transduce between their electric and mechanical energies. As shown in Figure 1.9 (b), when a voltage is applied across it, the inverse piezoelectric effect causes the material to exhibit a mechanical strain along with an electric polarization. Conversely, a mechanical stress causes the material to exhibit the direct piezoelectric effect (Figure 1.9 (a)), leading to the generation of an electric voltage across the ends, along with a mechanical strain. This transduction effect is generally a linear interaction, in materials that do not have inverse symmetry. In crystalline materials, the external stress (and thus strain) causes a relative displacement of electric charges in the ions, thus leading to a rise of polarization in the material. Without the external stress, the charges nullify each other and hence no polarization is observed. Piezoelectricity may be exhibited by crystals, ceramics, polymers and composites. Single and polycrystalline piezoelectric materials may be produced through sintering. Their dielectric, mechanical and piezoelectric properties are often controlled and optimized as per requirement. Piezoelectric polymers have found numerous applications in the electronic industry owing to their flexibility, lightweight nature as well as their ability to exhibit large strains without damage. Piezo-composites had been developed to overcome problems inherent to ceramics (such as fragility and stiffness). They allow great freedom to have better properties, including electromechanical coupling factor and suitable acoustic impedance. In this study, we shall consider piezoelectric crystals such as lithium niobate or quartz that enable very low propagation losses.
Magnetic sensors based on SAW devices and their applications:
Magnetic SAW sensors are designed either in the delay line configuration with a full film of magnetostrictive layer as the sensitive part [Yam’80, Gan’76, Elh’16] or in the resonator configuration with the IDTs made from the magnetostrictive material [Kad’11]. As discussed previously, one of the primary attractions for using SAW sensors pertains to the ease of wireless functioning while at the same time offering a passive and batteryless operation. This attraction is fulfilled by the resonator geometry, which is the most suitable for wireless interrogation. Several advancements have been made in the field of magnetic SAW sensors starting from as early as 1975 when Ganguly et.al. [Gan’76] proposed a magnetically tuned SAW phase shifter. Over the years several pioneering research has been made to understand, describe as well as optimize magnetic SAW sensors. Yamaguchi et.al. [Yam’80] designed a SAW delay line based on TbFe2 and studied the variation of the SAW velocity with applied magnetic field. Kadota et.al. [Kad’11] pioneered to develop and study in detail SAW sensors with magnetostrictive IDTs (Ni electrodes). Recent developments in the area of 5G technology has rendered SAW devices a huge technological momentum and is thereby promoting more research into sensors based on SAW technology. Thus exploration of SAW devices for magnetic field detection has again taken a leap and is now at the forefront.
Applications of a magnetic SAW sensor:
The previous sections have enumerated the different types of magnetic sensors available in the market today. Thus it becomes imperative that the applications of such a varied number of devices are also numerous. In the current section we aim to enlist the applications of magnetic SAW (MSAW) sensors. Owing to their wireless and battery-less configuration, SAW sensors find tremendous applications, especially in areas of hostile conditions and thus their magnetic counterparts are no exception. SAW based magnetic sensors are primarily of two types: (i) Magnetostrictive SAW sensors [Gan’76, Yam’80, Kad’11]; (ii) Giant magneto-impedance (GMI)– SAW sensors [Hau’00, Hau’06, Ste’00, Kon’18].
From magnetic switches for implantable biomedical devices to current, torque or rpm sensors in the automobile industry the applications of SAW sensors are indeed varied. Of late, a new use of these sensors for detection of vehicular traffic along marked tracks has also developed. Not only does this kind of sensing allow the monitoring of the number of vehicles but also its speed. With giant leaps in the field of electric vehicles, it is not long before magnetic SAW sensors find use in the segment either as current, temperature or speed sensors etc. Long term monitoring of structural health using the localised perturbation of the magnetic field in ferromagnetic pipes for example, is another important avenue for these sensors and thus provides enormous potential to find possible uses in the construction as well as pipeline industries.
Simulation of the magneto-acoustic response:
In order to understand the various aspects of the geometry of a SAW device, we first started off our simulation based study to comprehend the effect on the magneto-acoustic response of the device to a magnetic field. The study is based on the numerical model described earlier that is based on the Equations 2.9 -2.12. The aim of this study is to optimize the structure of the device to obtain the best configuration for a magnetic sensor. The 128o-YX crystal configuration of Lithium Niobate (LiNbO3) is a widely used substrate for SAW devices, optimizing between a high Rayleigh wave SAW velocity and a high electromechanical coupling coefficient (K2) among piezoelectric crystals [Kan’06, Nau’06]. Thus, we consider a substrate of 128o Y-cut LiNbO3 with propagation along both the X and X+90o directions in this research work (Chapter 3). However, since the idea of this chapter was to get a basic idea of the effect of the device geometry on the magneto-acoustic response, we have chosen to work with a propagation direction along the (X+90o) direction, the results of which, we believe, will be similar if the propagation was along the X-direction. As we will describe in subsequent chapters, this is also the substrate we will use for our experimental analysis. Further, to keep the study simple and to ensure that we completely understand the nature of the magnetic material in it, we consider simulation model (described in a previous section), encompasses the magnetic behaviour coupled with the mechanical and electrical responses of the device. Thus, we considered a simplified 2-D geometry of the entire device, using antiperiodic boundary conditions along the X-axis and perfectly matched layer (PML) at the bottom of the LiNbO3 layer to mimic a semi-infinite substrate. As mentioned, the magnetisation of the sensitive layer is considered to follow the ideal Stoner-Wohlfart model. This ideal nature of the magnetisation allows us to focus entirely on the effects of the geometry of the device. Here we also consider that the anisotropy of the layer is directed along the wave propagation direction in all the cases of study with an anisotropy field value of 100 Oe. The material properties used during the simulation are also kept constant for all the cases. The chosen elastic properties and constants corresponded to Ni because of their easy availability for the study.
2.3.1 Variation of electrode thickness: The first set of simulations concern the impact of the thickness of the IDT electrodes on the magnetic response. Here, we have considered the electrodes themselves to be the active magnetostrictive layer. Thus, we consider a device that has a wavelength (λ) of 6.5 µm and a metallization ratio of 60%. Keeping these values constant in the geometry, we consider electrode thicknesses ranging from 50 nm to 1 µm. The resonance frequency observed varied for each thickness of the IDT, but was around the 500 MHz range.
Fabrication of the device:
The substrate: In this study we have used lithium niobite (LiNbO3) as the piezoelectric substrate. There are different crystal cuts of LiNbO3 each dictating separate surface, optical and piezoelectric properties. For this study, we consider the 128o rotated Y-cut LiNbO3 crystal. This indicates that the crystal is cut along the plane 128o rotated from the Y-plane about the X-axis. The Z-axis of the crystal is therefore tilted by -38o from the normal of the surface. This particular cut of the substrate is often used to produce SAW filters, thanks to its high electromechanical coupling (K2=5.4%). In our case we consider it because it provides a good optimum between the (K2) and the quality factor
(Q). This LiNbO3 cut contains only the X-axis (and not Y or Z) and thus the in-plane direction perpendicular to this X-direction is referred to as the X+90o-axis. The substrate used are polished on one side and have a thickness of 500 µm. Subsequent fabrication steps are carried out on the polished side of the LN.
Deposition of the magnetic film: The deposition of the magnetostrictive layers on the concerned substrate is the first important step in the fabrication process. In the case of this study, the deposition step is of special interest as it is in this step that we induce the preferential magnetic anisotropy direction. The LNO wafer is first diced into rectangular pieces of 24 mm × 20 mm. This is followed by a thorough cleaning of the substrate. A 5 stage cleaning process is used wherein the substrate is first subject to ultrasound for 10 minutes while being submerged in a 2% solution of NaOH. This is followed by 10 minutes of ultrasound each in de-ionised water, acetone and isopropyl alcohol. Finally, at the end of the ultrasonic bath process, we subject the substrates to oxygen plasma under vacuum for 10 minutes. This final step ensures that no organic debris remains on the surface of the substrate (in case it was missed out by the ultrasonic bath).
Temperature Coefficient of Frequency:
The temperature coefficient of resonant frequency or TCF as it is known, is a measure of the thermal stability of the resonator. It is indicated by a gradual change of the resonant frequency with variation in ambient temperature. Since the resonators find extensive applications in the communication industry, temperature stability is an important factor and needs to be as close to zero as possible.
The origin of the TCF can be traced to the linear expansion coefficient (αi). An increase in the expansion coefficient directly affects the resonator’s dimensions and thereby its wavelength. The TCF is also dependent on the change of the elastic constants with temperature. Mathematically it is represented as: 1 = ×0.
where, f represents the frequency and T the temperature. The TCF is usually expressed in parts per million per degree Celsius (ppm/oC). Although the actual curve of the variation of the resonance frequency is parabolic, a linear behaviour is generally considered for smaller ranges, such as the one in this current study. The increase of the ambient temperature can either cause the resonant frequency to drift to higher values in which case the TCF for the said resonator is positive or drift to lower values in which case the measured TCF is a negative value.
Use of [dielectric / piezoelectric] multi-layered structures: In certain cases, a thin amorphous dielectric film may be deposited on the piezoelectric wafer to provide the function of surface passivation, reduction of pyroelectric properties and smoothening to reduce propagation losses. It may also be used to modify the coupling factor, reduce the temperature coefficient of frequency and enable the use of higher order SAW in the wafers. They can also play the crucial role of isolating the electrodes from shorting by metal particles and electrode deterioration due to electric and acoustic fields as well as environmental and other harmful effects. A uniform dielectric film on the surface of a SAW device can also be engineered in a way to alter the propagation characteristics of the surface wave. Additionally, the presence of a dielectric film presents the possibility to produce Love waves. [Hic’00].
Use of ZnO on Quartz for TCF compensation: One of the primary advancements for a ZnO / Quartz multi-layered structure has been for the development of devices with a high degree of temperature stability. Such devices have special applications as vapour or liquid sensors in extreme environments as well [Cal’19]. ST-cut quartz has been used widely in frequency control due to its temperature stability and low cost. Moreover, it has been shown that the positive TCF of the ST-cut quartz can be very well compensated for by the negative TCF of a ZnO layer [Tsa’15, Kad’03, Tom’05, Tal’04]. Thus, the role of the ZnO layer is of a double benefit for us; it not only compensates the TCF but also aids in the excitation of Love waves. Moreover, thanks to its large electromechanical coupling coefficient (K2) compared to those of Quartz, the ZnO layer enhances the K2 of the final structure (effective K2 = 0.25%) [Kad’04].
Fabrication of the multi-layered structure:
The substrate: In this study we have used quartz as the piezoelectric substrate. There are different crystal cuts of quartz each dictating separate surface, optical and piezoelectric properties. For this study, we consider the 42.75o Y rotated ST-cut quartz crystal. This indicates that the crystal is cut along the plane 42.75o rotated from the Y-plane about the X axis. Since this cut contains only the X-axis, the in-plane direction perpendicular to the X-axis is referred to as the (X+90o) axis. Thus we obtain two propagation directions in this particular cut of quartz i.e. X and (X+90o) which respectively promote the excitation of Rayleigh waves and Shear Horizontal waves. The substrate considered for this study are polished on only one side and are 500µm in thickness. The devices fabricated subsequently for the study are on this polished side of the ST-cut quartz.
Previous research in the field by Kadota et.al. [Kad’02] has shown that along the (X+90o) propagation direction, the formation of SH waves with a higher velocity is preferred while along the X direction there exists only a Rayleigh wave of a lower velocity. Since we are primarily interested in Love waves, which are basically Shear-Horizontal waves in a confined structure, the direction of propagation in the devices is considered along the (X+90o) direction.
Table of contents :
Chapter 1: Surface Acoustic Wave Devices for Magnetic Field Detection
Part 1: Magnetic Sensors
1.1.2 Types of Magnetic Sensors
Part 2: SAW devices: State-of-the-Art
1.2.1 SAW Devices
1.2.2 Principle of operation of SAW sensors
1.2.3 Applications of SAW sensors
Part 3: Magnetic SAW Sensors
1.3.2 Magnetic sensors based on SAW devices and their applications
1.3.3 Physics behind the working of a magnetic SAW sensor
Chapter 2: Optimization of the SAW structure
2.1 The Numerical Model
2.2 Ideal magnetisation curves
2.3 Simulation of the magneto-acoustic response
2.3.1 Variation of electrode thickness
2.3.2 Variation of device wavelength
2.3.3 Variation of metallization ratio of electrodes
2.3.4 Magnetic overlayer structure..
Chapter 3: Studies on magnetic SAW sensors based on [TbCo2/FeCo] multilayered electrodes
3.1 Introduction to shape effects in magnetism
3.2 Fabrication of the device
3.2.1 The substrate
3.2.2 Deposition of the magnetic film
3.2.3 Fabrication of the electrodes
3.2.4 Topography of the fabricated device
3.3 Magnetometry measurements
3.4 RF Characterization
3.5 Numerical Simulation
3.6 MSAW Measurements
3.7 Applications – A Prototype
3.7.1 The concept
3.7.2 The experimental set-up
3.7.3 The measurements
Chapter 4: Temperature compensated magnetic field sensor
4.1 Temperature Coefficient of Frequency
4.2 Fabrication of the multi-layered structure
4.2.1 The substrate
4.2.2 Fabrication of the electrodes
4.2.3 Deposition of ZnO and CoFeB
4.3 Magnetometry measurements
4.3.1 Choosing the right thickness of CoFeB
4.3.2 Magnetization measurements
4.4 RF Characterization and TCF measurement
4.5 MSAW Measurements
Chapter 5: Multi-sensory SAW device
5.1 Temperature Coefficient of Magnetic Anisotropy
5.2 Fabrication of the device
5.2.1 Fabrication of the electrodes
5.2.2 Deposition and micro-structuration of ZnO and CoFeB
5.2.3 Topography of fabricated device
5.3 Magnetometry measurements
5.3.1 Magnetisation measurements on multilayers – effect of temperature
5.4 RF Characterization and TCF Measurement
5.5 MSAW Measurements