Research Could Lead to Smaller Computer Chips

High-tech devices seem to be constantly shrinking — from the tiny iPod nano to the latest laptop computers that are no larger than a notebook. Manufacturers, however, can only make devices as small as scientists can build computer chips, so Texas A&M University physics researcher M. Suhail Zubairy and two colleagues at the Max-Planck Institute in Germany have proposed a way to write circuits on chips smaller and more precise than ever before.

Over the past decade, scientists have been trying to get around a problem called the “diffraction limit” that prevents them from writing smaller circuits onto chips, said Zubairy, who is currently teaching at Texas A&M University at Qatar. He and his colleagues, M. Kiffner and J. Evers of the Max-Planck Institute, have found an efficient way to get around this limit, however, and potentially double or even quadruple the number of transistors scientists can put onto a chip.

The team’s work has been published in the journal Physical Review Letters and was reviewed in Physical Review Focus and highlighted in Nature, the highly respected British science journal.

One of the ways scientists write circuits onto chips is called optical lithography. In this method, scientists use beams of light to etch the image of the circuit onto a chip that is coated with light-sensitive material. When scientists shine light on the chip, the light-sensitive material is activated, and they can etch the circuit pattern onto the chip. But the size in which scientists can write the circuits is limited by the diffraction limit.

The diffraction limit occurs because of the wave nature of light, Zubairy said. “How close, for example, you can draw two lines is limited by the wavelength of the light you are using,” he explained.

You can think of it in the sense that you can only write as small as the size of the point of your pen. In the same way, scientists can only write as small as the wavelength of the light they are using.

“In the last 10 years, there has been a great interest in finding a way to get around this diffraction limit in order to shrink the size of chips,” Zubairy said.

Many researchers have submitted proposals to get around the limit, but the problem with all the proposals so far is that they involve the simultaneous absorption of multiple rays of light to take atoms on the chip from the ground state to an excited state, Zubairy said. Instead of using one beam of light to etch the circuit pattern, these methods use multiple beams to increase the atoms’ excitation energy, which reduces the light’s effective wavelength and allows scientists to write smaller.

The problem is that this process requires the photons — particles of light — to be in the exact same place at the same time. This is a very unlikely event, so you need a very large number of photons to achieve it, Zubairy said, and this requires very intense light beams that could damage the chips or equipment.

Zubairy and his colleagues, however, have proposed a way to get around the diffraction limit without using high-intensity light beams. To do this, they prepare the chips for etching by putting them in something called a “dark state” using a method called coherent population trapping.

In this method, Zubairy and his colleagues shine two or more light beams onto the chip in a particular pattern to “trap” atoms in the chip in a specific dark state. By doing this, they can create patterns where some of the atoms in the chip are put into a state where light can no longer excite them while other atoms are left susceptible to light. Regions of the chip where the atoms cannot be excited cannot be etched. In other words, instead of trying to decrease the size of the “point of their pen” (the etching light’s effective wavelength), Zubairy and his colleagues use complex patterns of light beams to block out the parts of the chip that they do not want etched.

It’s like using a stencil and spray paint — the researchers create a “stencil” of the circuit pattern that blocks out areas on the chip that they do not want etched. By doing this, they only leave certain very small regions on the chip susceptible to the light used to etch the pattern, so when they shine light onto the chip, only the open parts in the stencil are etched, Zubairy explained. This is how they are able to write patterns that are smaller than the light’s wavelength.

Zubairy and his colleagues’ proposed method surpasses the perceived fundamental limit for etching circuits on chips, and their research could lead to chips that are smaller than ever.

“In principle, our method can go beyond the best precision that scientists have been able to achieve up until now, and we can, in principle, double or quadruple the number of transistors that can be put on a chip,” Zubairy said. “The research that we presented in our paper is the basic idea of how the process would work, so now we just have to find the right kind of materials and the right combination of laser fields to do that.”

-aTm-

Contact: Amelia Williamson at (979) 845-4641 or aaw11@tamu.edu or M. Suhail Zubairy at zubairy@physics.tamu.edu