Piezoelectric MEMS Resonators

Piezoelectric thin films are of interest for micro-electromechanical systems (MEMS) since the earliest developments in MEMS technology. This is quite natural or logic because the piezoelectric effect is an electromechanical effect. Resonators and ultrasound wave generators were among the first demonstrated MEMS devices [1–3]. In the 1970s and 1980s, the investigated thin film materials were mainly ZnO and AlN. In the 1990s, PZT was added to the list for having a stronger piezoelectric material for actuators (see, e.g., [4]). For higher-frequency applications, as, e.g., pass band filters for telecommunication in the GHz frequency range, the two wurtzite structures AlN and ZnO remained the champions, simply because they exhibit much higher mechanical quality factors than PZT, and in comparison to LiNbO₃, they are much more easily grown in thin film form. Moreover, integration and process compatibility with the rest of the device are less difficult using the relatively simple wurtzite materials. The strong polarity of their crystalline structure allows for a polar growth and a stable piezoelectric response with time, whereas ferroelectrics always risk depoling.

This section concentrates on the leading ferroelectric material used in thin-film piezoelectric MEMS: lead zirconate titanate (PbZr _x Ti_1−x O₃) or PZT. PZT-based MEMS technology has been explored extensively for a variety of actuator applications [1] but has received less attention for RF MEMS resonator and filter applications. Despite the long and successful history of bulk PZT resonators [2], the process complexity/compatibility issues and high mechanical losses have discouraged the exploitation of this strong piezoelectric for these applications. However, for select MEMS resonator and filter applications, thin-film PZT offers many unique advantages due to the high electromechanical coupling factors, permittivity, piezoelectric stress constants, and the DC-bias electric field dependence of these properties. Significant progress has also been made in developing materials deposition and processing technologies that address the fabrication challenges with PZT thin films. The goals of this section include an analysis of the deposition of these materials, patterning techniques, identification of device design and processing concerns, and finally a detailed subsection covering examples of how PZT thin films have been incorporated into resonant-based devices.

The holy grail of seamless monolithic integration of MEMS with supporting circuitry has driven the development of electromechanical devices in compound semiconductors. Piezoelectricity manifests itself in the majority of compound semiconductors due to the inherent crystal asymmetry of the two or more atomic species comprising the material. The interaction of acoustic waves with charge carriers in these piezoelectric semiconductors was investigated as early as 1953 [1] and revealed interesting phenomena such as reduced electron effective mass due to phonon drag [2] and wave amplification and velocity shift in the presence of free carriers [3–5]. The unique properties of piezoelectric semiconductors and their implications for resonant devices are explored in this chapter.

A piezoelectric-based resonator is an electromechanical device in which electrical and mechanical energies are reciprocally converted to each other at a resonance frequency through the two-way coupling between stress and electric field in a piezoelectric material. Moreover, the mechanical energy in the resonator body and the applied electrical energy through the metallic electrodes convert from potential to kinetic and back in every vibration half cycle (Fig. 5.1). Therefore, the overall performance of a resonant system is determined by both of the energy conversion mechanisms mentioned above. To quantify the efficiency of these energy conversions in a resonator, there are two specific parameters defined as electromechanical coupling factor and quality factor (Q).

Flexural piezoelectric resonators can be broadly classified as being one-dimensional devices, such as bars [1,2, 3, 4], and tuning forks [5], or two-dimensional devices, such as round or square diaphragms [6–8]. Models for the static deflection of piezoelectrically actuated one-dimensional beams are presented in [9, 10], while a lumped-element model for flexural plate-wave resonators is presented in [11]. This chapter focuses on developing lumped-element equivalent circuit models for vibration of continuous one-dimensional and two-dimensional resonators. These models allow the designer to predict various aspects of resonator behavior such as the natural frequency and vibration amplitude of the various vibration modes.

Sensors are found in a wide variety of applications, such as smart mobile devices, automotive, healthcare, and environmental monitoring. The recent advancements in terms of sensor miniaturization, low power consumption, and low cost allow envisioning a new era for sensing in which the data collected from multiple individual smart sensor systems are combined to get information about the environment that is more accurate and reliable than an individual sensor data. By leveraging such sensor fusion it will be possible to acquire complete and accurate information about the context in which human beings live, which has huge potential for the development of the Internet of Things (IoT) in which physical and virtual objects are linked through the exploitation of sensing and communication capabilities with the intent of making life simpler and more efficient for human beings.

This chapter describes basic properties and modeling of resonators using acoustic waves propagating in a solid body. This type of waves is called the bulk acoustic wave (BAW), which can be excited and detected efficiently using piezoelectricity. The resonators are widely used in various applications such as clock generation, frequency filtering, and sensing. The most popular ones are crystal quartz resonators [1] for relatively low frequency applications (<100 MHz). Recently BAW resonators fabricated by thin film and micromachining technologies, i.e., film bulk acoustic resonators (FBARs) [2], are getting popular for relatively high frequency applications (>1–3 GHz) such as duplexers used in mobile and smart phones.

High Q, low loss, small size, and the potential for high-frequency operation are the primary benefits of shear piezoelectric MEMS resonators. The mechanical Q of a device strongly determines the overall frequency stability of the resonator/oscillator under test. In particular, piezoelectric crystals such as quartz can be carefully oriented to provide frequency stabilization over temperature. Thickness-shear-mode (TSM) piezoelectric resonators have been fabricated in the MHz–GHz frequency range using wafer-level processing suitable for volume manufacturing of frequency control devices. This chapter discusses the fundamentals of shear-mode operation and provides examples of simulated and fabricated devices and their performance and applications.

Piezo-MEMS resonators have been used in timing applications to provide reference signals. In such applications, the temperature-induced drift of generated signals is often required to be minimized. A resonator’s resonance frequency drifts as a result of temperature change because of the intrinsic material properties and resonance mode characteristics. Without special design, such frequency drifts usually well exceed what is required by the applications. To reduce this variation, compensation techniques can be applied to either the resonator itself, or the oscillator circuitry, or the package. In this chapter, we discuss the various temperature compensation techniques applied to the resonator itself, and their advantages and drawbacks.

Accurate modeling of piezoelectric MEMS Resonators is key to reducing fabrication cycles or diagnosing issues with fabricated devices. Analytic models provide a first pass at assessing device performance but are idealized and do not represent actual device performance to sufficient accuracy, especially for the high-order bulk modes utilized in MEMS oscillators and filters. Computational modeling by the finite-element method has taken great strides in capturing greater detail in both geometry and physical behavior of these resonators. While continuing improvements in computational power have aided in this progress, it is the advances in both algorithms and methodology in coupling physical domains that have enabled greater accuracy compared to fabricated devices. In this chapter we review the key challenge areas in resonator design for which advances in computational modeling provide predictive value.

Since the first demonstration of thin-film piezoelectric resonators [1], high performance MEMS devices including low insertion-loss resonators and filters [2, 3], small form factor energy harvesters [4], large-force actuators [5], and highly sensitive resonant sensors [6–8] have been successfully demonstrated using piezoelectric materials such as AlN, ZnO, and PZT. Thin film AlN has been of great interest mainly due to the high quality of its film growth and CMOS compatibility, as well as well-developed process recipes. The processing advantages in addition to the superior piezoelectric and acoustic properties of AlN, including large wave propagation velocity [9] and low thermoacoustic dissipation [10], have made it a popular choice for piezoelectric transduction of MEMS resonant devices.

The every growing human demand or need to communicate has been the driving force that began to define the need for frequency control devices. The need to control a radiated signal with information imbedded and then to be recovered by a receiving device (all forms of transmitted and received communications). The technical community found that devices that used mechanical stability to set a resonant frequency were capable of offering the needed performance if the mechanical element could be excited and operate in an electrical circuit. The piezoelectric devices offered this solution where mechanical and electrical operation is coupled together.

Large volume testing is expected in manufacturing for production. It is one of the operations that could be the determining factor for the success of an MEMS product. The function of the testing is to ensure that every part manufactured is a good part. Today, MEMS products, such as motion sensors and microphones for smart phones, are often just a portion of the many components used in the final product form—if any one of them is defected, the entire system may fail at any time. Hence, the impact of a defected die could be much bigger than just a component failure and would possibly cause situations such as production line down, schedule delay, unhappy customers, and moreover, a tarnished company reputation which affects future sales and even the sales of other product lines. The product quality is, therefore, one of the most important measures of the usability of an MEMS product. Additionally, the control of the test cost is quite often an on-going task for MEMS manufactures. Testing for MEMS could be many times more complex than that of typical semiconductor products. If not managed well, it could significantly affect the cost margin and render the product uncompetitive. In this chapter, we will use the piezoelectric MEMS resonators developed by Integrated Device Technology, Inc. (IDT) as examples to discuss the techniques used in MEMS manufacturing testing in details. Methodologies used in both test rejections and test time reduction (for test cost management) are discussed.

This chapter will introduce the challenge of developing a quartz crystal replacement based on piezoelectric thin film resonators fabricated using microelectromechanical system (MEMS) technology. Very often quartz crystal resonator technology is dismissed as 90 year old technology, suggesting it is outdated, which could not be more wrong. At present, all wireless communication functionality found in a handset, including cellular, WiFi, FM, Bluetooth and GPS relies on quartz based frequency references. There has been an immense development to reduce size, package height, cost and frequency errors over the years. At the same time, there has been a strong trend to consolidate the different frequency sources.

While there is no broad consensus RF bulk acoustic wave (BAW) technology should actually be listed as an RF-MEMS technology, most experts would agree BAW is a true success story. In fact—at this point in time—no other “true” RF-MEMS technology comes even close to BAW in terms of market penetration and revenue generated. It is a prime example for an innovation having a profound impact on size and performance of devices most persons consider a necessity in daily life: the smartphone. Without the selectivity of high-performance acoustic filters, none of the advanced cellular voice and data services would exist. There is an insatiable demand for higher wireless data rates across all boundaries of society and geography. Why has BAW flourished way beyond the predictions in recent years?