Manufacturing process will give `ball` room a whole new meaning
By Kelly Sewell
Allen, TX — A startup that hopes to revolutionize the semiconductor industry by replacing semiconductor wafers with spheres could alter the contamination control industry because it employs a proprietary manufacturing process that substitutes cleanrooms with alternative clean environments.
Ball Semiconductor (Allen, TX) proposes to manufacture semiconductor spheres in a series of hermetic clean pipes and tubes in place of cleanrooms by the year 2000. “The balls will be in constant motion as they are continuously processed, treated, and transported at high speed through hermetic pipes during various processes for the crystal-growing, grinding and polishing steps and then for the repeated cleaning, drying, diffusion, film deposition, wet and dry etching, coating, and exposing steps of the integrated-circuit manufacturing process,” explains Dr. Ram K. Ramamurthi, vice president of R&D operations. The lithography process is the only one that will require a clean environment, which will be some kind of minienvironment or portable cleanroom station, he says.
Company researchers have constructed an experimental set-up in which to test the technology. The balls are fed into a station feeder, which is a rotating disk with loose balls on it. Argon gas pushes the balls from the disk into the 2-mm diameter quartz tube, and once the ball is in the tube, it`s a sealed clean environment, Ramamurthi explains. The floating technique by which the balls travel keeps them from touching the sides of the tubes or each other to prevent contamination. The process is continuous — once a ball enters the maze of tubes, it is transported at a rate of 5 meters per second (about 20 miles per hour) from one station to the next. The tubes run vertically, seven to eight meters high, in continuous loops from one station to the next. The cycle time is 5 days, compared to the three months it takes to process a wafer.
Because the tube configuration takes up so much room, the manufacturing space will not be any smaller than a typical wafer fab; however, it will not be expensive cleanroom space.
What could this mean to the cleanrooms industry? Initially, not much.
“Over the years we`ve seen lots of technology come and go,” says Dr. Robert N. Castellano, president of The Information Network (New Tripoli, PA) market research concern specializing in the semiconductor industry. “You always have newer concepts, but until the technology is proven, you`re talking 10 years before it`s going to be a factor.”
He explains that the existing semiconductor technology is a “moving target,” and for companies to switch to a new technology takes a long time, costs a lot of money and could result in lower product yields — a risk most companies aren`t willing to take.
Ramamurthi admits the transition will take time. “When a technology takes hold, the people will try to think how to use their current expertise in the new technology area. So we expect the cleanroom suppliers will come to us with new solutions on how to keep the pipes cleaner, or come up with other technologies to keep the tubes and pipes and minienvironments clean.
Company officials announced the concept at their two-day technology conference held in conjunction with Semicon West in July. Researchers offered proof that they`ve successfully exposed and developed an integrated circuit metal pattern on a silicon ball. In recent months, the researchers have proven four out of five critical components needed to produce spherical semiconductors: spherical lithography, no-contact processing, spherical single crystallization, and 3-D VLSI design. They have yet to prove the remaining component, 3-D VLSI clustering. However, Ramamurthi says he expects this will be accomplished by year`s end. Once it is, the pilot line will be implemented with commercial production slated for the year 2000.
But whether it will work and gain acceptance remains to be seen.
“Look at history at how some things coming into play have affected the industry,” Castellano says. “The same factors that kept SMIF from entering the mainstream are factors that will keep Ball out of the mainstream, too.” CR