When considered only suitable for high-speed communication systems, an alloy InGaAs may one day be found with silicon in high-performance computations.
For decades a material has dominated the production of computer chips and transistors, where the world’s technology capital – Silicon Valley – bears its name. But perhaps the silicon mandate will not last forever.
WITH ONE researchers have found that an alloy called InGaAs (indium gallium arsenide) may have the potential to make transistors smaller and more efficient. Previously, researchers thought that the performance of InGaAs transistors deteriorated on a small scale. New research shows that this apparent deterioration is not an inherent property of the material.
The findings may one day help boost computational power and efficiency beyond what is possible with silicon. “We are very pleased,” said Xiaowei Caik, lead author of the study. “We hope this result will encourage the community to continue using InGaAs as a channel material for transistors.”
Cai, now with Analog Devices, has completed his research as a PhD student in MIT Microsystems Technology Laboratories and the Department of Electrical and Computer Engineering (EECS) with Professor Jesús del Alamo Donner. Its authors include Jesús Grajal of the Polytechnic University of Madrid, as well as Alon Vardi and del Alamo of MIT. The article will be presented this month at the IEEE International Virtual Device Virtual Conference.
Transistors are the building blocks of a computer. The role of switches, which stop or discharge electricity, creates a huge set of computations – from simulating the global climate to playing videos of cats on Youtube. A single laptop can have billions of transistors. To improve computing power in the future, as in decades, electrical engineers will need to develop smaller, narrower transistors. Until now, silicon has been the semiconductor material for transistors. But InGaAs has shown signs of becoming a potential competitor.
Electrons can be easily encrypted using InGaAs, even at low voltage. The material “is very well known [electron] transport properties, “says Caik. InGaAs transistors can process signals quickly, creating faster calculations. In addition, InGaAs transistors can operate at a relatively low voltage, improving the energy efficiency of a computer. So InGaAs seems like a hopeful material for computer transistors. But there is a catch.
The good electron transport properties of InGaAs seem to be deteriorating on a small scale, the scales needed to build faster and denser computer processors. The problem has led some researchers to conclude that InGaAs nanoscale transistors are not suitable for this task. But, says Cai, “we’ve seen that misunderstanding.”
The team found that small-scale performance problems in InGaAs are partly due to oxide trapping. This phenomenon blocks electrons when they try to pass through a transistor. “A transistor seems to function as a switch. You want to turn on the voltage and have a high current, “says Caik.” But if you’re trapped by electrons, all you have to do is turn on the voltage, but you only have a very limited amount of current in the channel. So the ability to change is much lower when you have that oxide capture. ”
Cair’s team blamed the oxide capture for studying the frequency dependence of the transistor – the speed at which the electrical pulses are emitted from the transistor. At low frequencies, the performance of InGaAs nanoscale transistors appeared to be impaired. They worked well at frequencies of 1 gigahertz or more; trapping oxides was no obstacle. “When we use these devices at a very high frequency, we notice that the performance is very good,” he says. “They’re competitive with silicon technology.”
Cai hopes his team’s findings will provide new reasons for researchers to obtain computer transistors based on InGaAs. The work shows that “the problem to be solved is not really the InGaAs transistor itself. It’s a problem catching oxides, ”he says.“ We think it’s a problem that can be solved or can be solved. ”He adds that InGaAs has shown that it has shown both in the classic and quantum computing applications.
“This [research] it remains very, very exciting, ”says del Alamo.“ We strive to push transistors to extreme performance. ”One day, that extreme performance could come at the hands of InGaAs.
This research was partly supported by the Defense Threat Reduction Agency and the National Science Foundation.