Cymer Expands into Extreme UV

Aaron J. Hand

As the semiconductor industry debates next-generation lithography, the players are scrambling to develop components for the narrowing field of choices. Cymer Inc. recently announced that it has entered the game, working on a 13.5-nm light source for extreme-UV lithography.

The company has developed a prototype EUV light source based on a dense plasma focus device. Although Cymer has departed from the laser source used in its offerings for deep-UV lithography, there are several similarities between the two light sources. The latest prototype, for example, uses an all-solid-state pulse power drive that is virtually identical to the power modules used in the company's 248-nm sources. "This gives us a tremendous comfort factor," said William Partlo, director of EUV technology for Cymer.

In Cymer's dense plasma focus prototype, the insulated gate bipolar transistors (IGBTs) used to transfer energy are the same as those in the pulse power system of the company's excimer lasers.

Based on electric plasma thruster technology designed for space applications, the device consists of a coaxial electrode set that forms a conductive plasma sheath between the inner and outer electrodes. The lithium plasma is constricted into a very small volume, Partlo said, which raises the temperature to the 100-eV level necessary for EUV lithography. Cymer's device -- scaled down from comparable x-ray versions -- operates at powers below 1 kV and has the potential for a pulse repetition rate in the kilohertz range.

Simplicity and scalability were two likely advantages that led Cymer to choose this design. Laser-based EUV light sources developed by Sandia National Laboratories in Albuquerque, N.M., are more complex, Partlo said, and have yet to be proved. A capillary discharge device developed by researchers at the University of Central Florida (see Photonics Spectra, November 1998, p. 39) in Orlando seems to suffer from scalability problems, he said.

Partlo estimates that the final light source will be 200 to 300 µm in diameter and several millimeters long.

It will be several years before the technology is made commercial, noted Pascal Didier, senior vice president of worldwide customer operations for Cymer, because its development -- like EUV lithography development on the whole -- is still in its preliminary stages.

Although there is always some doubt about such a new technology, Didier said, Cymer has proved the feasibility of its device and minimized its risks by using some central functions from tried DUV technologies. "The biggest challenges come from all the different core competencies being put into place as a group," he said. "How are they going to evolve over the next 12 to 24 months?"