Stanford Research Institute scientists say they have discovered a new, low-cost process for producing pure silicon, a breakthrough that could dramatically accelerate the commercial use of solar energy.

The new process, SRI says, would reduce the cost of silicon used for solar cells by as much as 90 percent, and meets a goal Energy Department scientists had not expected to reach until 1986.

"It's a simple process. It's cheap, fast and gives a pure product. You can't ask for anything more than that," said Dr. Leonard Nanis, adding, "In the area of materials alone, it's a major breakthrough."

Silicon, one of the most widely distributed elements on earth, is used to fabricate photovoltaic cells, solar cells made of thin sandwiches of silicon crystals that convert sunlight into electric current. To date, however, the high cost of pure silicon -- nearly $65 a kilogram -- has posed a major obstacle to economical solar cells for commercial use. SRI claims its new process will lower the cost to about $5 to $7 a kilogram.

Nanis said in a phone interview yesterday that SRI, a non-profit West Coast research facility, is now negotiating with private industry to sell the rights to the process. "Commercialization clearly is the next step," Nanis said.

Silicon accounts from one-fifth to one-fourth of the total cost of producing photovoltaic cells.

Reaction to the SRI announcement ranged from optimism to caution and outright skeptism.

Herb Epstein of Solar Lobby, a Washington-based public advocacy group said, "It's very important if true."

Ralph Lutwack of the Jet Propulsion Laboratory in Pasadena said he was hopeful about the SRI'S claims, adding "The estimates are based upon very preliminary information." Lutwack also said that the JPL, which has a contract from the Energy Department to evaluate silicon and pyhotovoltaic research, does not have detailed data on the new process.

Nanis said the new process utilizes a dry mix of sodium and sodium fluorosilicate -- a by-product of chemical fertilize production -- and a catalyst. An added advantage of the SRI process, he says, is that in production -- further reducing the costs of the silicon.

Most of the 3,000 metric tons of silicon produced annually worldwide is used by the semiconductor industry. Because of the expected takeoff in demand for silicon resulting from the growth of the photovoltaic industry during the early 1980s, many industry officials have predicted that silicon prices would rise dramatically.

Paul D. Maycock, head of the Energy Department's Office of Photovoltaic Systems, says it will likely take up to four years before a commercial plant could be in operation using the new process.

According to DOE, the current cost of producing electricity from solar cells is between 50 cents and one dollar a kilowatt hour, compared with DOE'S 1986 goal of 5 to 6 cents.

Nanis said he and other SRI scientists had been working on the new process "on and off for the last year," and that it stemmed from earlier silicon research SRI was doing under a DOE contract.

A major advantage of the new process, Nanis said is that it is one step, uses low-cost materials, and, unlike the current processes does not require extensive heating.

As for the discovery itself, Nais said it resulted from a joint effort undertaken with Dr. Angel Sanjurjo, and Dr. Robert Barlett, other SRI scientists.

"He (Sanjuruo) said let's try this, and it was successful right off the bat," Nanis said, adding that the discovery was somewhat unexpected.