Removing a roadblock to smaller transistors using 2-D firstname.lastname@example.org
In 1965, Intel co-founder Gordon Moore noticed that the number of transistors per square inch of integrated circuit doubled roughly every two years. Moore’s Law posits that this trend would continue into the foreseeable future — and it has. Certainly the ability to hold a smartphone in the palm of your hand — in the process availing yourself of the computing power of a circa-1990 supercomputer — is a testament to Moore’s prescience.
But 50 years later, Moore’s Law is bumping up against the laws of physics. As transistor density increases, so does performance-degrading heat. This year, chip manufacturers are rolling out a 10-nanometer transistor, but their path to even smaller transistors is blocked by their inability to keep these chips cool.
Assistant Professor Mona Zebarjadi, collaborating with Assistant Professor Stephen McDonnell, is developing an active electronic cooling device using two-dimensional (2-D) materials that could remove this roadblock. “The cooling we think we can get from these devices beats the best coolers on the market,” Zebarjadi said. “We are very excited about the possibilities.”
Zebarjadi has a joint appointment in the Charles L. Brown Department of Electrical and Computer Engineering and the Department of Materials Science and Engineering, while McDonnell is a faculty member in the Department of Material Science and Engineering. Both are involved with UVA Engineering’s Multifunctional Materials Integration initiative, a $10 million interdisciplinary effort to develop new, advanced and complex materials and devices that – from their atoms all the way to their finished products, and systems of products – have a built-in level of energy efficiency and functionality that does not exist today.
Generators and Coolers
Zebarjadi began exploring 2-D materials as a result of her interest in thermoelectric devices. In 1821, Thomas Johann Seebeck, a German physicist, discovered that a difference in temperature in a material linking two dissimilar conductors could produce electricity. Two hundred years later, Zebarjadi and other researchers view the ability to generate electricity from waste heat as a way to increase the efficiency of everything from automobile engines to power plants.
Two-D materials like graphene have a crystalline structure that consists of a single layer of atoms. For thermoelectric generators, the 2-D materials have a fatal flaw. Their lattice structure, the source of their superior electrical conductivity, also makes them excellent conductors of heat. Because of this thermal conductivity, it is impossible to maintain the temperature differential that is the basis of the thermoelectric effect.
However, Zebarjadi and her colleagues realized that they could turn this thermal conductivity to an advantage by, in essence, running the generator backwards, creating a device known as an active Peltier cooler. Apply current to an active Peltier cooler, and heat flows from the hot side of the device to the cool side.
The problem with most existing Peltier coolers is their power inefficiency. Zebarjadi and her colleagues quickly saw that graphene’s combination of superior electric and thermal conductivity would make it excellent for cooling. “To cool a hotspot on a chip, you want a flexible material that can conduct heat and electricity at the same time,” she said. “Two-dimensional materials like graphene are perfect.”
Zebarjadi found that placing graphene on a boron nitride crystal substrate increased its cooling properties even further. “We had a thermoelectric power factor about two times higher than in previous thermoelectric coolers,” she said. “It significantly exceeded the theoretical values that we had calculated.”
Building Cooling Capacity
Compared to 3-D materials, however, 2-D materials lack the volume needed to pump sufficient heat away from a hotspot. The solution is to create thermoelectric materials comprised of multiple layers of 2-D materials. McDonnell will help Zebarjadi meet this challenge.
In his lab in Wilsdorf Hall, McDonnell has recently finished building an ultra-high vacuum system that will enable him not only to synthesize multilayered 2-D materials, but also to investigate their structure, composition and electronic properties. “There are just a handful of labs in the country with a set-up like it,” McDonnell said.
A critical advantage of McDonnell’s system is that his growth and characterization systems are coupled, allowing him to characterize these devices without exposure to air that might change their properties. He can also move from synthesis to characterization quickly.
“Narrowing the gap between Mona’s theoretical models and the actual properties of the devices is an iterative process,” he said. “Being able to fabricate and analyze a device in the span of a few days will greatly accelerate the process of developing devices that provide the cooling that Mona requires.”
For her work on thermoelectric devices, Zebarjadi has been awarded a National Science Foundation Career Award and a Young Investigator Award from the Air Force Office of Scientific Research.