300-mm is hot at tokyo semicon; SEMI standards will play in Japan

02/01/1997

300-mm is hot at Tokyo Semicon; Semi standards will play in Japan

SEMICON/Japan generated much news related to 300-mm wafer technology, including word that Japanese organizations have agreed to work through the SEMI standards process for 300-mm instead of developing a parallel set of standards, and discussions of cooperation between the Japanese SELETE consortium and the International 300mm Initiative (I300I).

Other highlights included updates on SELETE`s equipment evaluation schedule, and a cost analysis from NEC suggesting that a 30-40 percent cost benefit will be obtained with 300-mm substrates.

The disclosure that the Japanese J300 group has agreed informally to allow SEMI to handle standards development was hailed as a major breakthrough by many equipment makers, who had feared the emergence of two sets of conflicting standards. Noting that SEMI is an international body, not an American organization, SEMI chairman James Greed said J300 leaders portray the SEMI standards process as "knitting together the whole (300-mm) effort. Not only is there a functional infrastructure, but the quality of the products is pretty much right on."

He explained that over the years, there has been some difficulty in getting chipmakers to help in the standards development process. However, given the cost and time frame of the 300-mm effort, "You look at J300 and the International 300mm Initiative - the members are all users and they`re all saying, we`ve got to have standards and have them quickly. This is a chance to do it in an organized way." J300 and I300I, which represents chipmakers from the US, Europe, and Asia outside Japan, have held discussions, he added, and set out areas where they have agreed to disagree, such as the question of whether to use open cassettes or front-opening unified pods for wafer transport. "Those two approaches are being done in a way that`s compatible," noted Greed. "The healthy thing is that they`re sharing positions."

Much remains to be done in getting actual standards in place; Greed suggested that the model followed by SEMATECH in its early days might be worth revisiting by J300 and I300I. "[SEMATECH] was able to assign people to work full-time on standards, and then submit them as proposals. That might not be the exact working model, but standards development has many dimensions and we need to have a plan in place. For instance, who will keep the standards evergreen? We may need to revise 300-mm standards every three years instead of every five."

Leaders of the Japanese SELETE 300-mm tool evaluation consortium and the I300I met during the conference to begin defining areas for cooperation. Frank Robertson, head of I300I, said, "There has been a lot of convergence with SELETE." He cited fundamental areas like fab modeling and interface standards as good candidates for joint work as both groups perform tests on 300-mm production tools. "After being in existence for a year, we`ve gotten a better understanding of where cooperation can take place," said Robertson. "Both consortiums have had to define what they want to do." Robertson added, "Having the same equipment metrics and making sure there is no duplication of effort are the important things."

Murray Bullis of SEMI said, "The desirability of having a common set of specifications and metrics for equipment evaluation cannot be overstated." He noted that tool vendors would much prefer to conduct just one demonstration run of their products, and "it would be nice if the results could be adapted to the other organization." He said current thinking is that 60-70% of tool specs could be set in consortium evaluations, with the rest determined in traditional negotiations between customer and supplier.

In other 300-mm news:

SELETE general manager Haruo Tsuchikawa said work on his group`s cleanroom had been completed in mid-November, and the first piece of 300-mm equipment, a vertical furnace, had been delivered. About a dozen tools are to be in place by the end of the fiscal year in March, and some 35 by September. A Canon deep UV wafer stepper, the first such system to be delivered by a major

supplier, is expected in April. Tsuchikawa said SELETE should be able to produce patterned test wafers by October.

Frank Robertson, general manager of I300I, said his organization will conduct "about 40-50 equipment demonstrations during 1997, focusing first on 0.25 micron, and also characterize wafers from six or seven suppliers." He warned, however, that "very clearly in lithography, both exposure and pattern defect detection tools are late, and that needs to be remedied." Robertson also cautioned equipment developers to include funding for 300-mm test wafers in their 1997 budgets. "Be sure it`s in place - you don`t want a half-million dollar surprise," he said.

Masanori Kikuchi of NEC`s Semiconductor Group presented a cost analysis model for large-diameter wafers, concluding that a 30-40 percent benefit can be obtained by moving from 200- to 300-mm wafers, but that "cost benefit is likely to level off with a size of approximately 400 mm." A technological breakthrough will be needed to make 400-mm wafers cost-effective, as higher wafer costs offset the manufacturing cost advantage. - P.N.D.