LBL team taps SSC know-how for electronics applications

March 10, 1995

By Lynn Yarris, [email protected]

Technology used in the development of an ion source for the Superconducting Super Collider is now being applied to the electronics industry.

LBL's new Ion Beam Technology (IBT) group has turned its scientific know-how to developing ion beam projection lithography, a processing technique that could play a central role in the future production of computer chips and microelectronic-based flat panel displays.

Optical lithography, the use of a beam of visible light to etch integrated circuit patterns onto the surfaces of computer chips, is a multibillion dollar industry that was once ruled by the United States. Domestic chip makers hope to reclaim their former pre-eminence with the next generation of integrated circuits which will feature patterns too small to be etched from beams of visible light. Production of these chips will require lithography that uses beams of ions or x-rays.

The most promising route for using either ion beams or x-rays to produce integrated circuits is a technique called projection lithography. In this approach, a beam passing through a stencil-like mask projects a circuitry pattern onto the surface of a semiconductor wafer. Although x-ray projection lithography is currently at a more advanced stage of development than ion beam projection lithography, ion beams offer certain advantages that make them an attractive alternative.

"Ion beams can be used on a variety of surface shapes whereas x-rays can only be used on flat surfaces," says principal investigator Ka-Ngo Leung, a physicist in the Accelerator and Fusion Research Division (AFRD). "An ion beam source is also much less expensive than the best source of x-rays, which is LBL's Advanced Light Source."

The major components of an ion beam projection lithography system are the source that produces the ion beam and the transport column and electrostatic lenses that accelerate and focus it through a mask and onto the surface of a wafer. For the ion source component, Leung and his colleagues are developing a radio frequency (rf)-driven ion source similar (though smaller) than the source they were developing for the SSC.

An rf-driven ion source consists of a chamber inside which a radio frequency electric field transforms a gas of hydrogen or helium into a plasma. The plasma, which is confined within the chamber by a multicusp magnetic field (named for the shape of its force field), contains electrons and both negative and positive ions, but only the positive ions are extracted and formed into a beam.

"An rf-driven ion source is simple to make, rates high in all aspects of performance, and is easily adapted to an automated operation," says Leung. "Its lifetime is practically unlimited, which means it is economical, and it can provide the cold (low energy) ions that lend themselves to forming an intense and easily focused beam."

Leung's group is working now on optimizing the geometry of their rf-driven ion source to maximize the current and the stability of the beam it generates. In ion beam projection lithography, an intense, stable beam is one of the keys to producing a clear, sharp image on a wafer. The other crucial factor is the system of lenses that focus the beam.

Leung and his colleagues are participating in a CRADA that LBL signed last fall with the Advanced Lithography Group (ALG), a Maryland-based consortium of private industry, government laboratories, and universities. The objective of the CRADA is to develop a U.S. capability for ion-beam projection lithography by the year 2000.

The ALG group will use a beam "cross-over" technique in which the mask and a series of electrodes serve as converging and dispersive lenses. An ion beam expands in size as it emerges from the source so that it can illuminate the entire mask. After passing through the mask, the ions in the beam are focused down to a point where they cross over one another and begin expanding again before they strike the wafer and recreate the image of the mask. With this technique, the ALG group can reduce the size of the pattern that is transferred from the mask to the wafer as much as ten times.

Leung's LBL colleagues on this project include Chun-Fai Chan, Wulf Kunkel, Bud Leonard, Tom McVeigh, Margit Sarstedt, Steve Wilde, and Don Williams. UC Berkeley students who are participating on this project include Paul Herz, Yvette Lee, Luke Perkins, Dick Pickard, and Mike Weber.