A Radiofrequency Threshold Temperature Sensor Using a Hf0.5Zr0.5O2 Device and a Microacoustic Piezoelectric Resonant Sensor

This article presents the first prototype of a novel threshold sensing system operating in the radiofrequency (RF) range, capable of passively detecting and recording instances of temperature violations with high sensitivity and without the need of DC-biased memory devices. The system comprises a microfabricated hafnium zirconium oxide (HZO) ferroelectric varactor, a microfabricated lithium niobate (LiNbO3) shear-horizontal (SH0) Lamb wave microacoustic resonator used as a temperature sensor, and an inductor. All components have been soldered onto a printed circuit board for testing and characterization. Through this threshold sensing system, we also provide the first demonstration of ferroelectric switching in a HZO varactor triggered exclusively by an RF signal without requiring any DC voltages. The findings reported in this paper offer a promising pathway for future RF passive tags leveraging micro- and nanosystems to implement a programmable threshold sensing functionality with high sensitivity and with embedded memory capabilities. [2023-0215]

I. Introduction

In the last decades, the Internet of Things (IoT) has generated a multitude of opportunities in various emerging applications, including remote health monitoring, environmental monitoring, smart manufacturing and more [1], [2], [3], [4]. Just recently, the growth of artificial intelligence and cloud-computing has triggered a paradigm shift in the design of IoT-systems, introducing new capabilities in wireless sensor networks like distributed data-processing, machine learning and real-time decision making. This has further increased the need for wireless sensor nodes (WSNs) to collect and process unprecedented amounts of sensing data. With nearly a trillion WSNs expected to be deployed by 2025, the WSN market is now projected to reach a worth of billion by 2026 [5]. Unfortunately, the WSNs predominantly used today are active. Hence, they rely on onboard batteries to sustain their radiofrequency (RF) and sensing operations, which need to be periodically replaced. Consequently, using active WSNs in massive WSN-deployments is not practical due to the unsustainable costs associated with periodic battery replacements and due to the significant environmental challenges posed by the disposal of depleted batteries in landfills. A growing interest is being recently paid to passive WSNs. Unlike the active ones, passive WSNs do not require onboard batteries, which makes them suitable for large-scale WSN-deployments [6]. Nonetheless, the current passive WSNs offer a limited range of functionalities. For example, they cannot store the occurrence of violations in a targeted parameter of interest (PoI) in a non-volatile but refreshable manner [7], [8], [9]. This is different from what active WSNs can do, as these components can embody memory devices that typically require a battery to work, like electrically-erasable-programmable-read-only memory (EEPROM) devices [10]. This limitation currently prevents passive WSNs from being used, for example, to accurately identify and permanently mark food and drug items that have experienced temperature irregularities along the cold chain, or to detect and mark areas of buildings or other structures that show unexpected early signs of failure [11], [12], [13].