Micromachined piezoelectric harvester drives fully autonomous wireless sensor

As part of the Holst Centre programme on micropower generation and storage, imec researchers have developed a temperature sensor that can wirelessly transmit data in a fully autonomous way. imec says it has developed the piezoelectric energy harvesters by using cmos compatible mems processes on 6' silicon wafers. The harvester consists of a Si mass that is suspended on a beam with aluminum nitride (AlN) as piezoelectric material. The researchers claim that by changing the dimensions of the beam and mass, the resonance frequency of the harvester can be designed for any value in the 150-1200Hz domain. According to imec, the use of AlN as piezoelectric layer is a notable achievement as AlN has several advantages in terms of materials parameters and ease of processing compared to the commonly used lead zirconate titanate (PZT). AlN can be deposited up to three times faster while composition control is not an issue due to the stoichiometric nature of the material. The research has been completed by the development of a wafer scale process to protect the piezoelectric devices. IMEC claims that the power output significantly increases by the use of the vacuum package compared to packaging in atmospheric pressure. In a three step process, glass covers are coated with an adhesive, vacuum bonded on top and bottom of the processed wafer and diced. In one test. the piezoelectric harvester was connected to a wireless temperature sensor, built op from off the shelf components. After power optimisation, the consumption of the sensor was reduced from 1.5mW to ±10µW. When subjected to vibrations at 353Hz at 0.64g (indicating a realistic amplitude of the vibrations), the system generated sufficient power to measure the environmental temperature and transmit it to a base station with an interval of 15 seconds. The result, say the researchers. proves the feasibility of building fully autonomous harvesters for industrial applications. Future applications could include power sensors in industrial applications such as tire-pressure monitoring and predictive maintenance of moving or rotating machine parts.