May 24, 2012, MEMS Business Forum, Santan Clara, CA—Al Pisano from UC Berkeley describes research into Wireless MEMS for harsh environments for energy and power monitoring. These sensors are not something for a smart phone and they cannot use standard silicon.

The economic impact of additional sensor capabilities is fairly high. Gas turbines are used for backup and main power generation and geothermal generation accounts for 17 percent of the power in Iceland. The percentage would be higher, but is transmission limited. A small change in operating parameters can result in $ millions in savings.

To measure the operating conditions in a power generator, a sensor has to operate at 600° C, possible with silicon carbide and aluminum nitride sensors. The energy from combustion is 100 exajoules. In a gas turbine, the 15th stage is running at 1.315 atmospheres. These generators need highly localized temperature sensors that are less than 150 microns thick or they will affect the airflow.

In automotive engines and running equipment, there is a need for temperature sensors on the brakes to determine the coefficient of friction which varies with temperature. Tire pressure sensors need to withstand the vulcanization processing temperatures of 500° C in steam. Most sensors are damaged by the steam, more so than by the pressure and temperatures. Silicon carbide is immune to these environmental hazards.

In geothermal wells, the energy is generated by the earth's internal heat 4km below the surface. Again, the temperatures are in the 600° C range. For efficient energy recovery, you need to have a sensor that can function at the bottom of a well.

For many of these applications, self-powered low-rate RF is the way to transfer the data from the sensor to the monitoring equipment. The sensor can be low resolution, but must operate at high temperatures and pressures. For some of the generating equipment, the temperature sensor must operate in a high G acceleration environment.

They are working with collapsing diaphragms as differential capacitors for energy harvesting. These capacitors have to work at 12k psi +/- 100 psi and are implemented using 20nm graphene as the contact for energy harvesting functions. They are using pressure changes and not mechanical motion to generate power.

They use a AuSn eutectic die bond to match the operating temperature requirements. All the components are basic and crude compared to the more conventional sensors. They have developed a complementary SiC process that includes lateral JFETs. Vertical devices are still in development, but the individual devices are sufficient for the circuit requirements as the elements for a radio front end. Other materials like GaN need material stack development, so they are not ready for circuit designs.

They are focused on high temperature devices that will sell for high prices, rather than the more common sensors that are driving for high volume. For the environments and working conditions, the key is to have all pieces working. All of the sensors are in the same fab process. This process is low performance, but will work at very high temperatures. They are starting with temperature and will add other sensors in the future. This research is geared for big companies in energy generation and will help them with their logistics supply chains.