Jim Hwang (left) and Ph.D. candidate Steven Peng in the Compound Semiconductor Technology Lab.
From inkjet printers to cell phones, from airbags to electronic games, it seems that MEMS are everywhere.
If James Hwang has his way, the smart miniature devices may become even more pervasive.
MEMS—short for MicroElectroMechanical Systems—are prized for their size, functions and low cost. MEMS sensors detect incredibly tiny fluctuations in motion, speed and sound, while MEMS actuators manipulate matter with extreme precision. MEMS electronic components achieve unprecedented feedback and control, making it possible to put entire engineering systems—wireless communications, biometric identification, even chemical and DNA analysis labs—on a single computer chip.
Thus, says Hwang, it is MEMS accelerometers that detect the sudden change in motion caused by a collision and trigger a car’s airbag to deploy. On high-end digital cameras, a MEMS device senses and instantly corrects for vibration, compensating for a shaky hand on a zoom lens. MEMS motion sensors made a big hit out of Wii, the latest game console from Nintendo, which is equipped with magic wands.
Arranged in arrays, MEMS gain strength and efficiency. The array of tiny nozzles on an inkjet printer head enables a printer head to be made at low cost and to operate at high speed. An array of MEMS microphones suppresses ambient noise to improve the reception on your cell phone. And an array of tiny MEMS mirrors projects high-contrast slides in conference rooms and vivid movies in home theaters.
MEMs are mainstream
“MEMS affect everyone’s daily life without people realizing it,” says Hwang, a professor of electrical engineering and a materials scientist by education who also directs Lehigh’s Compound Semiconductor Technology Laboratory.
“Whenever you need a large array of mechanical devices that are small, lightweight, low-power and low-cost, that’s where MEMS really delivers.”
Hwang confesses that he came late to the MEMS revolution, having begun his first MEMS research project just four years ago.
Late or not, however, Hwang and his colleagues at MEMtronics
, a Dallas-based spinoff of Texas Instruments, have overcome two critical challenges holding up the use of MEMS in military applications, focusing on radio-frequency (RF) MEMS switches, which make or break circuit connections, as demonstration vehicles.
In October 2006, Hwang’s group became the first to operate an RF MEMS switch for more than 100 billion cycles. In 2005, the group developed an improved method of packaging the switches to protect them from environmental stresses.
The research was funded by the Defense Advanced Research Projects Agency
(DARPA), the research arm of the Pentagon, through its HERMIT (Harsh Environment, Robust, Micromechanical Technology) program.
DARPA, which set the 100-billion-cycle target for RF MEMS switches, is seeking improved packaging and reliability for the switches as part of its goal of developing space-based radar surveillance systems that can operate for at least 15 years.
“It is critical to package MEMS devices properly,” says Hwang, “because they are very susceptible to ambient conditions, including moisture, which can cause the components of the device to stick together.”
To improve the packaging of the RF MEMS switch, Hwang’s group first developed a state-of-the-art RF MEMS capacitative switch that operates with minimal power loss. When the researchers fitted the new switch inside a traditional package, however, power loss increased more than three times.
“We went from less than .1dB insertion loss, which represents high performance, to insertion loss of .3dB or greater,” Hwang says.
“Small, lightweight, low-cost and low-loss”
To solve this problem, Hwang’s team developed a package of novel dimensions.
“Historically, a package was created as an afterthought and was much bigger than a device. Now package and device are the same size. We have gone one step further. Our package is now actually smaller than our device. We achieved this by forming a transparent bubble over the critical area of the device at the same time that the device is made.
“The beauty of this is that it is small, lightweight, low-cost and low-loss. The additional [power] loss from the packaging amounts to less than .03dB, which is almost not measurable.”
Hwang and his colleagues next set their sights on the 100-billion-cycle target. They found that the RF MEMS switch was degrading after multiple cycles because the applied voltage (which controls the motion of the switch) was trapping an electrical charge inside the insulator that covers the bottom electrode of the switch. The trapped charge caused the switch’s top electrode to stick to the bottom electrode instead of springing back to its suspended position, after the application and removal of the voltage.
“We investigated the charging problem and modeled its effect. Then we optimized the material, along with the electrical and mechanical design, to minimize the charging effect to a tolerable level. This allowed us to operate the switch for more than 100 billion cycles without failure.
“Our group was the first to surpass the 100-billion mark for MEMS capacitative switches. We passed the previous record of 30 billion cycles, which we had established earlier in 2006.”
Defense-related applications for RF MEMS switches, says Hwang, include smarter, faster steering of phased-array communications and radar systems, which, like the eye of a fly, contain dozens to thousands of antenna elements. MEMS devices will enable the antennae to be guided electronically, with low power consumption, thereby eliminating many mechanical and heating problems. Furthermore, by electronically adjusting the signal phase of each antenna element, multiple communication links or radar targets can be tracked simultaneously.
“Phased-array radar is standard equipment on fighter aircraft and cruiser ships,” says Hwang. “Phased-array communication systems will be standard equipment on helicopters and tanks.
Cell phones represent a potential non-military application of RF MEMS switches, says Hwang.
“Multiple wireless communications standards and frequencies exist in different regions of the world. People have dreamed of an intelligent RF front end that will enable a cell phone to be quickly reprogrammed to switch between different standards and frequencies and to communicate with wired, Bluetooth, Wi-Fi, and GPS systems, without having to cram multiple sets of RF antennas, filters, tuners, amplifiers and so forth in a single phone.
“If we do develop an intelligent RF front-end, we can build a software radio, which has many applications, including one radio that will work at all standards and frequencies. MEMS will be the enabling technology for a software radio.”
Hwang’s achievement, which occurred at Lehigh, has been verified at MEMtronics and at the Air Force Research Lab, which has now exceeded 260 billion cycles using Hwang’s technology.
In the third phase of his current contract with DARPA, Hwang’s group hopes to achieve 500 billion cycles and to integrate dozens of MEMS switches in a phase shifter, which is the heart of phased-array communications and radar systems.
Hwang’s collaborators at Lehigh include Steven Peng, a Ph.D. candidate in electrical engineering, and Xiaolin Yuan, who earned a Ph.D. in electrical engineering in 2006 and now works for IBM.