Two academic scientists working in a basement laboratory at Catholic University here are locked in a tight race with the giant Corning Glass Co. to develop a cheap glass fiber that eventually will replace copper wire and bring about a multibillion dollar revolution in the communications industry.

Corning, Bell Laboratories and others are producing small amounts of "wire" from glass now, but the costs are prohibitive. Physicist Theodore Litovitz and chemist Peter Macedo say they can reduce the costs to a fraction of Corning and Bell's and even below the current price of copper wire, which is about $15 a kilometer.

If they are correct, the telecommunications lines of the future will be spun from glass and Litovitz and Macedo will be rich.

Corning is betting on its wire which now costs between $900 and $4,000 a kilometer to produce, depending on its quality. David Duke, the Corning project manager, said that with mass production the cost could be reduced in a few years, to about $100 a kilometer for medium grade wire.

Litovitz says that with the process he and Macedo have developed the cost of production should be no more than $10 a kilometer.

A key problem in cutting the cost of manufacturing the wires, or optical wave guides, as they are technically known, is finding an inexpensive way to remove iron from the sand used to make the glass.

Because the fine, light wires transmit light waves rather than electricity, they must be as clear as possible. The iron, which is what makes Coca-Cola bottles green, has to be completely removed.

Litovitz and Macedo have not yet succeeded in making their glass fiber as pure as Cornings, but they say that is simply a "quality control" problem that has nothing to do with their basic process.

Corning has spent over $10 million in developing its product, Duke said. Litovitz and macedo, whose work has cost about $500,000 to date, most of it in federal grants, contend that when their process is used in a full-production setting as opposed to a laboratory, the quality difference will disappear.

The higher the quality of the wire, the larger the load of messages it can handle and the greater the distance it can transmit them. Corning's product is now suitable for long telephone lines. The Litovitz - Macedo wire is not because it loses more light than the Corning wire.

Litovitz said that he and Macedo have been trying to sell their patented process since 1973. "ITT (International Telephone and Telegraph) turned us down," he said, "and now they want a license (to manufacture the wire). General Cable turned us down, too."

In recent months, Litovitz and Macedo have entered into licensing arrangements with Canada Wire in Montreal, Pilkington Glass in England, Sumitomo Electric in Japan and American Cystoscope in the United States.

When aksed why it took so long, Litovitz said, "who's going to believe two guys when Corning's put millions into it?"

The principle on which the wire works is relatively simple. Light waves like electrical waves can be modulated to carry messages. The signal is transmitted either by s ource of concentrated light such as a light-emitting diode (LED) for short distances or a laser for long distances.

The advantage of glass strands over copper wire, aside from the fact that sand is much more plentiful and cheaper than copper, is that they can carry far more messages while occupying far less space. A cable containing glass-fiber wire instead of copper wire, Duke said, "will be about the size of your little finger instead of the size of your thigh," like the coaxial cables in use today.

"The capacity of a telephone line," Macedo said, "is about 2,000 bits of information a second. A line feeding a color television set transmits about 5 million bits a second. But today's state-of-the-art glass can transmit a billion bits per second."

Possible future uses, aside from large telephone systems, include communication between computers, cable television, undersea cables, instrumentation systems and certain kinds of medical instruments.

Another advantage of glass wire is that sinc it transmits light instead of electricity, it is not subject to electromagnetic interference. A new phone line, for example, could be laid next to a major power line.

Duke contends that there will be some limited commercial use of glass wire before 1980, probably in instrumentation and computers.

Bell Laboratories currently is setting up a field trial in Chicago in which glass wire cables will be run from two central offices to customer premises.

According to Robert Laudise of Bell, the company already has produced hundreds of kilometers of the wire at a Western Electric facility in Atlanta, and tests "have proved successful beyond our greatest expectation."

Laudise would not disclose the cost of producing the wire, but he said Bell could do it "at least as cheaply as Corning."

Duke said that aCorning, which has been working on the process since 1966, has produced "thousands of kilometers" in the last 10 years. Litovitz estimates that he and Macedo have turned out "hundreds of kilometers" in their CU laboratory.

Ironically, Corning has been producing a type of high quality glass called Vycor for 30 years using the process Litovitz and Macedo have adapted for making glass wire.

The Litovitz-Macedo process, which is known as phase separation, works something like this:

Just as oil separates from water at a certain temperature, the same happens with iron deposits in glass. Glass rods are heated, and when te separation occurs an acid is used to leach out the iron, thereby purifying the glass.

In contrast, the system now used by Corning, Bell and others, involves the use of a gas called silicone tetrachloride which, when it is formed, leaves behind a deposit of solid iron chloride. Afterwards, the silicone tetrachloride gas is burned with oxygen to form silicon dioxide, or pure sand, minus the iron. The pure sand is then made into glass giber.

When asked whether he thought the process he and Macedo had invented would make them wealthy, Litovitz answered, "If it works, we won't be hurt by it. And there seems to be no reason why it shouldn't work."