Semiconductors: The next generation
Nanotechnology researchers pursue hybrid silicon chips with life-saving potential
Today’s cheetah can become tomorrow’s turtle in the changing world of semiconductors—those increasingly tiny switches that say yea or nay to electrons as they jump through the microscopic gates of a technological society.
The rule of domination, semiconductor style, is that he or she wins who can make the devices cheaper, smaller, faster and less power hungry. Scientists worldwide are involved in this highly competitive research, within which The University of Texas at Arlington is a major player. Consider the work of electrical engineering Professor Zeynep Celik-Butler or materials science and engineering Assistant Professor Seong Jin Koh.
Both researchers continue to refine the incredibly small semiconductors created by companies like Texas Instruments and Intel. But a technological snag looms. The vast majority of semiconductor applications are based on silicon technology, and “there’s eventually a limit on what you can do,” says Dr. Celik-Butler, director of UT Arlington’s Nanotechnology Research and Teaching Facility.
“Once you start to make it smaller and smaller, what you end up with is that the size becomes so small that you run into a problem—the dreaded Heisenberg Uncertainty Principle—with managing your electrons.” In other words, the devices become unreliable.
Some chips are already less than 100 atoms thick. The smaller the chips get, the more difficult it is for them to function reliably. Celik-Butler believes that the platform on which the commercial semiconductor industry is built will reach its limits within five years.
Dr. Koh says 15 years, maybe 20. “The end of what you can do with silicon is not here yet,” says Koh, who designs hybrid chips that add components to basic silicon wafers. The components, called nanoparticles, have a diameter only a few tens of times larger than that of a single atom.
“The biggest problem I have now is figuring out how to put these individual nanoscale building blocks exactly where we want so that we make devices that have superior capabilities that current silicon devices cannot produce,” he said.
Although Koh has a $400,000 National Science Foundation grant to develop technologies for precise placement of nanoscale building blocks, he’s also involved with more immediate applications focusing on next-generation sensors and other devices. He has a grant from the Office of Naval Research for which he’s trying to create hybrid silicon-nano semiconductors that will be both efficient in achieving tasks and will use tiny amounts of power.
Such devices might be used extensively in very small, unmanned surveillance aircraft. That kind of downsizing delivers a double bonus.
“It cuts down not only the size of the device but also the power source, which can be much heavier and bulkier than the device itself,” Koh says. “We want devices that can be operated with little power, like in a submarine or missile or in space, devices that can be powered for a long time without new energy sources.”
And that technology, Koh says, ultimately filters down to the kind of everyday technology found in cellphones or in the potentially life-saving gadgetry at the doctor’s office or hospital.
A third Koh project involves the Texas Advanced Research Program portable detection devices that are so efficient that they can detect even single DNA molecules.
“This kind of device has many medical applications as well as possibilities for homeland security because it can quickly detect even minute quantities of chemical or biological threats,” he says.
Koh is far from alone in his efforts at UT Arlington to develop ever-smaller chip technology. Electrical engineering Associate Professor J.C. Chiao recently received an NSF grant related to creation of implantable devices in the brain that could affect behavioral modification. The implications for treatment of some forms of mental illness are vast but depend on having reliable, tiny and energy-saving semiconductors like those that UT Arlington researchers are designing.
Much of this research takes place at the UT Arlington nanotechnology facility, which has more than 30 tenured or tenure-track faculty and more than 100 users.
“We’re conducting some critical research in areas that range from molecular electronics and atom-wire devices to device-level packaging,” Celik-Butler says. “Sensor area research ranges from biomedical and chemical sensors to smart skin. Other research focuses on electronic noise and electromigration—both crucial considerations in creating and packaging very small, precise semiconductors.”
Celik-Butler, for instance, is working on a cardiopulmonary resuscitation patch that relies heavily on semiconductor technology. The patch provides feedback to someone administering CPR. The device is being tested by the School of Nursing and shows promise.
So what’s the future of the silicon-based semiconductor? Both Celik-Butler and Koh believe that its demise has been overstated. Even as new semiconductors evolve, it’s likely that the old standby silicon models will be extensively utilized.
“There are many reasons, but silicon technology is well advanced and the investment in it is huge,” Koh says. “People have been working on it for 30 or 40 years, so in my view, the next generation of semiconductors will still be silicon based but with hybrid technologies.”
Celik-Butler sees an enormous future for hybrid silicon chips involving hafnium oxides. Hafnium is a metallic element found near the middle of the periodic table. With oxidation it becomes insulative, an essential quality for semiconductors.
She also is encouraged about developments in “disruptive” technology, a new way of thinking about the on and off of electron flow that appears to expand the potential of silicon-based chips.
“If there is a technology that overcomes silicon, it has to be practical, affordable and work at room temperatures because otherwise it would require scrapping billions of dollars of equipment invested in making silicon devices,” she said. “Or it has to be manufactured using the same equipment. Otherwise, the economics of it simply aren’t going to work.”
Can Celik-Butler predict the future of semiconductor technology? She’ll give it a shot.
“The most likely future of new semiconductor applications is in spintronics, an incredibly complex technology that involves use of electron spins to control switching. Other slightly less esoteric possibilities include single electron transistors and nanophotonics,” which can provide high-bandwidth, high-speed and ultra-small optoelectronic components.
Those three categories all happen to be disciplines in which numerous other UT Arlington researchers have expertise and that no doubt will provide grist for future research beyond the dimensions of silicon-based semiconductors. But for now, UT Arlington researchers continue to squeeze all the efficiency possible from the next generation or two—or three—of hybrid silicon semiconductors.
— O.K. Carter