Berkeley - Marvin Cohen has been thinking about physics since he was a child in Montreal, before he realized it was called physics. In fact, every day for the past 50 years physics has occupied his mind.

When he's not trying to predict what kind of material you'd get by mixing different atoms, he's devising problems for his students to solve, or trying to understand a new discovery in a totally different area of physics. He may jump from considering the effects of rolling a sheet of atoms into a cylinder, to wondering whether you can turn a tangled clump of nanotubes into a computer.

And since theoreticians need only pen and paper, he can do his thinking anywhere; a café in Paris or a Honolulu beach as easily as his fifth-floor office in Birge Hall, overlooking the Campanile.

"My favorite spots to do physics are Paris and Hawaii," he admits.

This relentless quest to understand the physics of the world, in particular the solid stuff - metals, semiconductors, insulators and superconductors - this week has garnered the 67-year-old Cohen the National Medal of Science, an award from the President of the United States to the nation's premier researchers.

Cohen's ties to UC Berkeley go back to his enrollment as an undergraduate in 1953, the same year he became a U.S citizen. Back then, before the free speech movement, his life revolved around fraternity dances and playing the clarinet and jazz saxophone. He didn't lose sight of his plan to become a physicist, however, and finally began to apply himself in 1957, when he enrolled as a graduate student at the University of Chicago.

"I got serious at that point," he said, hooking up with a mentor, Jim Phillips, who steered him into theoretical solid state physics. Choosing his own thesis topic, he showed theoretically that semiconductors could be superconductors and successfully predicted the first superconducting oxide, the precursor to the high temperature superconductors discovered 23 years later.

After a year with Bell Laboratories in New Jersey, Cohen migrated back to UC Berkeley, joining the physics faculty in 1964 and the research staff at Lawrence Berkeley National Laboratory in 1965. He continued to think about a problem proposed by Phillips when he was at Chicago, the solution to which has had ramifications throughout the world, ranging from the design of new semiconductors for the electronics industry to the search for zero-resistance superconductors and new forms of matter, such as nanotubes.

In the 1960s, solid state physicists were trying to understand why solids with different atomic structures have different properties. Some mixtures of atoms turn out to be metals and conduct electricity, other mixtures are insulators, and some, semiconductors have properties in between. Throwing together still other chemicals creates materials that, at low temperatures, conduct electricity without any resistance. A theory to explain these superconductors had been proposed in 1957 by Bardeen, Cooper and Schrieffer, and in 1972 won its originators a Nobel Prize.

Years later, in 1987, Cohen worked with the university's dance program to illustrate the pairing motion of the electrons in a superconductor for a NOVA television show on superconductivity, which aired the following year and won an Emmy Award.

To explain material properties, theoretical physicists began with the theory of quantum mechanics and fundamental principles of physics, but because each solid contains billions upon billions of atoms, and each atom carries dozens of electrons, the mathematical challenge was too great. The only solutions obtainable were for very idealized structures which only vaguely resembled real world materials.

Phillips and others had suggested that simplifying the picture might help by using an idea introduced by Enrico Fermi in 1934. Assume that the only electrons you need worry about are those floating relatively freely on the surface of the atom. Ignore the inner electrons, which are so tightly bound to the nucleus that they don't get involved in chemical reactions anyway.

Cohen took the hint and developed a "pseudopotential theory" that treats the inner electrons and nucleus as one entity creating an electrical field in which only the surface or valence electrons roam. With the advent of more and more advanced computers, he and his students created generations of computer programs to calculate the physical properties of almost any solid based on the movement of valence electrons in a pseudopotential field.

The model he created works so well that, in cases where theory and experiment disagree, Cohen and others have come to trust the theory.

"It's spooky that you can actually use these equations to predict new materials that never existed in the laboratory," he said. "Quantum mechanics is spooky anyway, but this is spookier."

The computer programs were quickly picked up by researchers in academia and industry - Cohen makes them available free of charge - and now "band-gap" engineers use them to design custom electronic components. Today, they are used around the world.

"Using the programs we developed, scientists and engineers are able to tell how the properties of a material will change when they put it under pressure, or add other alloys," he said. "The applications of the science are very important."

Cohen and his group also used the programs to predict successfully that silicon under pressure would change from a semiconductor to a superconductor and new superstrong solids and fibers.

The model even works at the scale of a nanometer - the size of a mere 10 hydrogen atoms laid end to end - which has steered Cohen into the burgeoning field of nanoscience to work alongside his former student and now UC Berkeley professor of physics, Steven Louie.

Cohen came up with one successful prediction while flying cross-country thinking about buckminsterfullerenes - molecules formed from 60 carbon atoms arranged in the shape of a soccer ball, reminiscent of the geodesic domes designed by the late architect Buckminster Fuller. Carbon does this easily, and these molecules can be isolated out of the soot from an electrical arc. But what about other atoms? He thought about substituting boron and nitrogen, but while they fit nicely in a hexagonal arrangement, they wouldn't normally form the pentagons also needed to fully cover the surface of a soccer ball. If you arranged them on a sheet of hexagons, like chicken wire, and rolled them up, he thought, perhaps they would form a stronger and easier-to-make structure than the carbon nanotubes.

It took six months to convince his graduate students that the idea wasn't off the wall, and when they ran the structure through their modeling program, it predicted not only that it would be a semiconductor but also that it would have a strength greater than that of steel. When UC Berkeley colleague Alex Zettl, professor of physics, finally synthesized boron nitride nanotubes in 1995, they had all the predicted properties.

He and Zettl later teamed up to found a company, now called Nanomix, to develop nanotubes as sensors and hydrogen storage systems for use in fuel cells.

"I have known and worked with Marv for 20 years, and I still am inspired every time I talk with him," Zettl said. "He's a delight to work with, he gets the best graduate students, and he is full of interesting ideas. He's an inspiration to us all."

His pseudopotential theory has led his and Zettl's groups to many new nanotube configurations, including an oxygen sensor, a diode and other electronic circuit devices.

"I joke that I can make a radio with just 50 atoms, but it operates in the X-ray," Cohen said. "That's not very useful at present, but it shows we can make components atom by atom."

Cohen now is investigating materials called nanocrystals that contain a mere 100 to several thousand atoms, and which exhibit properties that larger clumps made of the same material do not.

These accomplishments have earned Cohen many honors, including election to the National Academy of Sciences and elevation to University Professor, which essentially makes him a faculty member of the whole 10-campus University of California system. According to formal surveys of citations, Cohen is one of the most cited condensed matter physicist of the past 30 years. Now, as one of UC Berkeley's 25 National Medal of Science winners, he adds another laurel, one he will receive at the White House on June 13 in the presence of his wife, Suzy Locke Cohen, his son, Mark, and daughter and son-in-law, Susie and Jerry Crumpler.

"During the past 35 years, there has been a revolution in the use of quantum theory to predict the existence of new materials and properties, and Dr. Cohen is the individual most responsible for this advance," praised the citation from the White House.

"This is really an award for Berkeley - the campus and the lab," Cohen said. "I've had numerous students and post-docs, and they certainly share in this award too.".