Plasma startup creates high-energy light to make smaller microchips

Jun 28, 2012
The lab equipment includes a small system that measures plasma for electronics applications, attached to a larger tank containing plasma for energy research. Credit: U. of Washington

A University of Washington lab has been working for more than a decade on fusion energy, harnessing the energy-generating mechanism of the sun. But in one of the twists of scientific discovery, on the way the researchers found a potential solution to a looming problem in the electronics industry.

To bring their solution to market two UW engineers have launched a startup, Zplasma, that aims to produce the high-energy needed to etch the next generation of microchips.

"In order to get smaller feature sizes on silicon, the industry has to go to shorter wavelength light," said Uri Shumlak, a UW professor of aeronautics and astronautics. "We're able to produce that light with enough power that it can be used to manufacture microchips."

The UW beam lasts up to 1,000 times longer than competing technologies and provides more control over the million-degree plasma that produces the light.

For more than four decades the technology industry has kept up with Moore's Law, a prediction that the number of on a computer chip will double every two years. This trend has allowed ever-smaller, faster, lighter and less energy-intensive electronics. But it's hit a roadblock: the 193-nanometer ultraviolet light now being used cannot etch circuits any smaller.

The industry has determined that the future standard for making microchips will be 13.5-nanometer light, a wavelength less than 1/14 the current size that should carry the industry for years to come. Such light can be created only from plasmas, which are high-temperature, electrically charged gases in which are stripped from their . The is trying to produce this extreme in various ways. One takes a droplet of tin and shoots it with a laser to make plasma that releases a brief spark of light. But so far this spark is too brief. use a $100 million machine to bounce light off a series of mirrors and eventually project the light onto a silicon wafer; each step absorbs some of the light's energy.

"Over the past decade, the primary issue with these sources is they just can't produce enough power," Shumlak said. "It's a stumbling block for the whole semiconductor industry."

Four capacitors supply up to 10,000 volts of energy to separate the xenon electrons from their nuclei, creating an electrically charged plasma. Credit: U. of Washington

Fusion scientists, it turns out, are plasma experts. The hydrogen plasma in the sun is so hot that hydrogen nuclei fuse together and release energy. Scientists around the world, including at the UW, are working to replicate this on Earth. A fusion reactor would use hydrogen as its fuel and emit helium as a waste product, a technically challenging

but clean source of energy. The UW group's specialty is a lower-cost version of a fusion reactor that uses currents flowing through the material, rather than giant magnets, to contain the million-degree plasma. Their method also produces plasma that is stable and long-lived.

This video is not supported by your browser at this time.
Uri Shumlak, UW professor of aeronautics and astronautics, describes the equipment that can produce extreme ultraviolet light to make smaller microchips. Credit: University of Washington

"It's a completely different way to make the plasma that gives you much more control," said Brian Nelson, a UW research associate professor of electrical engineering.

The first time they triggered the experiment in 1999, an engineer looking through the glass said, "That was really bright!" That was when the proverbial light bulb went off, Nelson said, and the team began to explore applications for bright high-energy light.

They may have found that application in the microchip industry. Light produced through techniques now being considered by the chip industry generate a spark that lasts just 20 to 50 nanoseconds. Zplasma's light beam lasts 20 to 50 millionths of a second, about 1,000 times longer.

"That translates directly into more light output, more power depositing on the wafer, such that you can move it through in some reasonable amount of time," Shumlak said.

An initial grant from the UW's Center for Commercialization allowed the team to verify that it could produce 13.5-nanometer light. A gift last fall from the Washington Research Foundation helped the team shrink the equipment from the size of a broomstick to a new version the size of a pin, which can produce a sharp beam.

The company was established last year with help from the UW's Center for Commercialization and Henry Berg, a technology entrepreneur who met the researchers through the center's Entrepreneurs in Residence program. Berg is now CEO of Zplasma.

The company is seeking "smart money" from corporate investors who can integrate the new technology with existing industrial processes.

"I hope this gets implemented into the industry and has an impact," Shumlak said.

The group will continue its fusion research project funded by the U.S. Department of Energy. Raymond Golingo, a UW research scientist in aeronautics and astronautics, is co-author of the patent for the technology issued in 2008.

Explore further: Ray tracing and beyond

Related Stories

Making wafers faster by making features smaller

Dec 14, 2010

The manufacturing of semiconductor wafers used in all types of electronics involves etching small features onto a wafer with lasers, a process that is ultimately limited by the wavelength of the light itself. ...

Scientists Shed 'Light' on Semiconductor Quandry

Jul 14, 2009

(PhysOrg.com) -- UC San Diego scientists are using laser plasma-produced light sources to explore performance improvements of critical inspection tools for the semiconductor industry, which ultimately will ...

Nuclear fusion research key to advancing computer chips

Aug 18, 2009

(PhysOrg.com) -- Researchers are adapting the same methods used in fusion-energy research to create extremely thin plasma beams for a new class of "nanolithography" required to make future computer chips.

NIST Light Source Illuminates Fusion Power Diagnostics

Oct 11, 2007

Using a device that can turn a tiny piece of laboratory space into an ion cloud as hot as those found in a nuclear fusion reactor, physicists at the National Institute of Standards and Technology are helping ...

Recommended for you

Ray tracing and beyond

17 hours ago

Ray tracing is simple to explain at one level: "We all do it all day long: That's how you navigate the world visually," Gene Tracy explains. "The fact that I know that you're sitting there and not over there is because the ...

User comments : 8

Adjust slider to filter visible comments by rank

Display comments: newest first

guiding_light
5 / 5 (2) Jul 05, 2012
Zplasma's technology is Z-pinch plasma, which is already worked on in Japan (Ushio). They may have good results with deuterium/tritium plasmas, but they do not have the headstart in xenon or tin plasmas.
guiding_light
5 / 5 (1) Jul 05, 2012
With pulse width of 30-40 microseconds, the minimum stable dose is on the order of 1000 mJ/cm^2, which is really too much for resist to handle.
plasma_guy
5 / 5 (2) Jul 06, 2012
If you read their study a few years ago on the stability of Z-pinch plasma, their data shows it's not really that stable.

https://fusion.ga...lson.pdf
nullcodes_
5 / 5 (1) Jul 07, 2012
@guiding_ light you said
With pulse width of 30-40 microseconds, the minimum stable dose is on the order of 1000 mJ/cm^2, which is really too much for resist to handle.


If that's the case, why can't some sort of darkened glass filter be be placed in front of the light (such as a glass neutral density filter that can reduce the brightness considerably)? Do you have a special interest in this btw?
guiding_light
5 / 5 (2) Jul 08, 2012
@nullcodes: within a period of say 10 microseconds, the right amount of EUV energy must be collected, no more, no less. This depends on how long the plasma is on during that period as well as the power of the plasma during that period. Of course, the plasma power should be stable, not fluctuating during that period, but it does not appear so from the measured data. It rarely is.
Vendicar_Decarian
3.7 / 5 (3) Jul 08, 2012
My guess is that multi-emitter electron beam lithography is the next major step in chip manufacturing.

Apparently masks are now a very significant cost of manufacture, and electron beam lithography doesn't use them.

Unfortunately the process is still slow.
guiding_light
not rated yet Jul 10, 2012
My guess is that multi-emitter electron beam lithography is the next major step in chip manufacturing.

Apparently masks are now a very significant cost of manufacture, and electron beam lithography doesn't use them.

Unfortunately the process is still slow.


The electron beam lithography without masks is only acceptable in very very low volume, where the slowness doesn't matter, and when there is someone already paying.
Vendicar_Decarian
5 / 5 (1) Jul 10, 2012
Multi beam electron lithography involving several hundred to a thousand parallel beams is now possible.