Acoustic Resonance in an Independent-gate Finfet

This paper demonstrates the acoustic resonance of an Independent-Gate (IG) FinFET driven with internal dielectric transduction and sensed by piezoresistive modulation of the drain current through the transistor. An acoustic resonance at 37.1 GHz is obtained with a quality factor of 560, corresponding to an f.Q product of 2.1x10. The demonstrated hybrid NEMS-CMOS technology can provide RF CMOS circuit designers with high-Q active devices operating up to mm-wave frequencies and beyond. INTRODUCTION As we scale to deep sub-micron (DSM) technology, transistor unity gain frequencies increase, enabling the design of CMOS circuits for RF and mm-wave applications up to 95 GHz. However, such high-frequency CMOS transistors have limited gain, resulting in poor output power efficiency. Successful implementation of DSM CMOS in mm-wave applications therefore requires high-Q, low-power components operating at high frequencies. Another challenge facing DSM circuits is the increasing density of devices, projected to reach 10 devices/cm. At such densities, clock distribution and the power consumption associated with it necessitate implementation of low-power local clocks with the potential for global synchronization. The Resonant Body Transistor (RBT) presented in this work is a fundamental building block that addresses both these challenges. Field effect transistors (FETs) were first used for sensing mechanical motion in one of the earliest Micro Electromechanical (MEM) devices. In 1967, Nathanson et al. demonstrated the Resonant Gate Transistor (RGT), driving resonance in a gold cantilever with an air-gap capacitive electrode [1]. The RGT cantilever functioned as the gate of an air-gap transistor, with output drain current modulated by the cantilever resonant motion. This device achieved a resonance frequency of 30 kHz with quality factor of ~70 despite the limited processing capabilities of the time. Fabrication limitations prevented the proliferation of these and other MEMS devices until the advent of silicon-based surface micromachining. FET sensing has only recently regained momentum as a means of mechanical detection, and has been implemented in a variety of micromechanical devices. Resonant Gate Transistors similar to Nathanson’s device have been demonstrated in silicon air-gap resonators up to 14 MHz [2,3]. Mechanical resonators sensed through direct elastic modulation of a transistor channel have also been demonstrated. Such devices include air gap resonators with FETs embedded in the resonator body up to 71 MHz [4], mechanical mixing in single electron transistors up to 245 MHz [5], and piezoelectric high electron mobility transistor (HEMT) channel modulation in GaN resonators up to 2 MHz [6]. Internal Dielectric Transduction To improve electrostatic transduction efficiency and scale MEM resonators into the GHz domain, we previously demonstrated longitudinal silicon bar resonators using a novel method to drive and sense acoustic waves in the bar. This mechanism, termed ‘internal dielectric transduction’ [7], incorporates thin dielectric film transducers inside the resonator body for capacitive transduction. Internal dielectrically transduced resonators have yielded acoustic resonance frequencies up to 6.2 GHz [8] and frequency‐quality factor products (f.Q) up to 5.1x10 [9]. Moreover, these dielectrically transduced resonators demonstrate improved efficiency as resonance frequency increases, providing a means of scaling MEM resonators to previously unattainable frequencies. However, at multi-GHz frequencies capacitive feed-through becomes significant and prevents capacitive detection of MEM resonance without three-port mixing measurements. Unlike capacitive sensing employed in these resonators, FET sensing can amplify the mechanical signal prior to any feedthrough parasitics. Combining the benefits of FET sensing with the frequency scaling and high-Q capabilities of internal dielectrically transduced resonators, the authors recently demonstrated a Resonant Body Transistor (RBT) [10] operating at 11.7 GHz with Q of over 1800. This device incorporates a field-effect transistor into the resonator body for internal amplification of the resonant signal. The best RBT geometry for optimal dielectric transduction at high frequencies is an internal dielectrically transduced longitudinal-mode resonator, with dielectric films positioned at points of maximum strain. As we scale to higher frequency, the width of the resonator decreases, eventually converging to a geometry very similar to that of Independent-Gate FinFETs [11]. THE RESONANT BODY TRANSISTOR Theory The principle of operation of the internal dielectrically transduced RBT is shown in Fig. 1. The region in light grey represents the active region of the resonator, while the blue region is highly doped. The active region near the drive gate is biased into accumulation (red), so that a large capacitive force acts across the thin dielectric film (yellow), driving longitudinal vibrations in the body. A gate voltage is applied to the opposing gate, generating an inversion channel (blue) which results in a DC drain current. At resonance, elastic waves formed in the resonator modulate the drain current both by physically changing the gate capacitance and by piezoresistive modulation of carrier mobility. The internally amplified RBT has significantly lower output impedance than capacitive detection mechanisms, simplifying impedance matching with active circuits. Fig. 1. Top-view schematic showing principle of operation of a bulk-mode dielectrically transduced Resonant Body Transistor. The RBT geometry, similar to that of an IG-FinFET, incorporates FET sensing with a dielectrically transduced bar resonator. VD > VG-VT VG Vacc > VD VG VS = GND accumulation inversion