UCLA Chemists Report Significant Progress Toward Molecular Computers

In the Aug. 18 issue of the journal Science, the UCLA chemists resolve the challenge of a reconfigurable molecular switch that they presented last year in a paper on molecular computers. The earlier paper, published in the July 16, 1999, issue of Science, received worldwide attention.

"Last year's paper was the first experimental step toward molecular computers," said James R. Heath, professor of chemistry at UCLA. "This is the second experimental step, and the steps are no longer a slow walk, but a fast jog."

The team is led by Heath; J. Fraser Stoddart, who is UCLA's Saul Winstein Professor of Organic Chemistry; and Pat Collier, a postdoctoral scholar in Heath's laboratory.

"This molecular approach could have failed early on in many places, and it's not failing," Heath said. "These are still early days, but right now, it seems that everything is beginning to work together well and that many avenues are opening up. That means that if our next set of experiments doesn't work, we will have a number of other options to pursue. We're getting there. Overall, the progress is faster than any of us expected."

Stoddart agreed, saying, "When I joined UCLA's faculty three years ago, if someone had asked me how far off molecular computing was, I would have said on a scale of a quarter-of-a-century. Things I was only dreaming of are suddenly becoming a reality in Jim's lab."

In last year's Science paper, the researchers could switch the molecules only once; now they have done so hundreds of times.

"Last year, we published an architectural demonstration with molecules and demonstrated that it is possible to do simple mathematical operations," Heath said. "However, the switches used in that work switched only once, and this limited their relevance to any serious technology. Now we have taken another class of Fraser's molecules and demonstrated that they may be repeatedly switched on and off over reasonably long periods of time in a solid-state device under normal laboratory conditions. For the first time, we are able to turn the molecular switches on and off repeatedly."

Closing in on molecular RAM Developing molecular RAM (random access memory) was one of the major challenges the researchers cited last year, and now they are close to reaching it.

"With this paper, we have achieved a critical step on the way to molecular RAM at room temperature - which is required if this is to be a viable technology," Heath said. "From here, molecular RAM may well be just a matter of time. While there are many pitfalls between the demonstration of a technology and the actual invention, at the moment, we can't see how any of those pitfalls could prove fatal."

The research team gives much of the credit to the unique molecules that Stoddart and his team develop. Stoddart has been working for more than a decade on interlocking molecules with recognition sites.

"Not only do the molecular components in these interlocked molecules communicate efficiently with one another but the molecules also interrelate to one another in the solid-state device," Stoddart said. "It's significant that we have moved from the incoherence we had in the solution state, and have progressed from molecules that behaved like pedestrians walking in a shopping mall to molecules that march like a united, well-trained army. Where molecules were previously swimming around incoherently in solution, now we see coherent molecular motion in a solid-state device. The molecules have the ability to recognize one another and line up in an efficient manner."

Stoddart's molecules, called catenanes, consist of two tiny mechanically interlocked rings made up of atoms linked in a circle. To interlock the ring components, Stoddart and his research team build chemical groups that recognize one another into the components' precursors so that, when the appropriate pieces are brought together in solution, they self-assemble to form a pair of interlocked rings. In catenanes containing the right two rings, one ring can be stimulated to move between two different states with respect to the other reference ring, giving the catenane molecule its bistability. It is the switching motion, which can be induced by taking away and giving back an electron that is the molecular basis for the present device. In last year's Science paper, the chemists used another kind of interlocked molecule synthesized in the Stoddart labs. Called a rotaxane, it has a dumbbell-shaped component that is encircled by a ring component after the style of an abacus. Rotaxanes can also be made to switch their molecular states.

The catenanes reported in the new Science paper work much better than last year's rotaxanes, said the chemists, who added that they are already working with a new class of molecules that lead to significantly better switching performance than the catenanes. In fact, the chemists are working with more than a half-dozen different kinds of molecular switches, each with its own unique characteristics.

Molecular computers hold the promise of being far less expensive, much smaller, and much more energy-efficient than today's silicon-based computers. The UCLA groups, in collaboration with Hewlett-Packard researchers, are working on making molecular computers that may "learn and improve the more they are used," Heath said.

While the research could dramatically influence the computer industry, it may also have a significant impact on very different uses of information technologies, Heath and Stoddart said.

"Molecular machines will lead to other new technologies beyond molecular electronic computers," Stoddart said. "Other fields besides computers may be revolutionized by this molecular approach, although it is too soon to say precisely which ones will be the first to benefit."

The unimaginable may be possible Heath believes eventually there may be a new molecular manufacturing technology. "What once seemed like science fiction is now looking more and more like actual science," Heath said. "A molecular computer will enable us to do things we cannot even imagine now."

In addition to Heath, Stoddart and Collier, the research team includes Francisco Raymo and Gunter Mattersteig, former postdoctoral scholars in Stoddart's laboratory; Eric Wong, a postdoctoral scholar in Heath's laboratory; graduate student Chris Beverly; and Jose Sampaio, a visiting professor on sabbatical at UCLA.

"We're trying to learn to make a computer from the bottom up, and learning how to do such manufacturing is really what this research is all about," Heath said. "The power of Fraser's chemistry is that it's like a Tinkertoy set, and various desirable physical and chemical properties can be generated by piecing together the appropriate molecular components. Until very recently, it wasn't clear whether these unique molecular properties would translate into exploitable solid-state device characteristics. In fact, this research provides compelling evidence that this does happen. As we rearrange the Tinkertoys, we find that we can modify and improve the properties of these devices in ways that make complete chemical sense. This opens up a lot of possibilities for new devices and technologies."

Within a few years, Heath thinks the research team will develop circuits that have molecular logic, molecular memory and nano-size wires. A hybrid computer that interfaces with molecular memory with silicon logic is only a few years away, and a scientific demonstration of a nano-scale computer that is largely molecular - with molecular logic and molecular memory - will likely happen within the decade," Heath estimated.

Stoddart came to UCLA from England's University of Birmingham, where he was head of the school of chemistry and professor of organic chemistry.

"I tried to get collaborators to work on a molecular computer in Europe, but I drew a blank," Stoddart said. "It was all a dream until I came to UCLA."

"I liken Southern California today to Paris in 1900, which was the place to go if you were an artist," Stoddart said. "When I was working in England in the '90s, I felt that the place to make things happen as a scientist was Southern California, and I have been proved right."

In the July 1999 paper in Science, Heath, Stoddart and their colleagues, including researchers at Hewlett-Packard Laboratories, demonstrated highly effective molecular-based logic gates for the first time, and showed that, for certain tasks, molecules can effectively achieve results the same as, or better results than, silicon. For example, they showed that molecular circuitry, when assembled into the appropriate configuration, can tolerate manufacturing defects. That paper was widely considered the first real step toward making a molecular computer.

In addition to conducting this research, Heath is a leader of a joint proposal by UCLA and UC Santa Barbara to create a wide-ranging California Nanosystems Institute that could help revolutionize many fields of science. The project has been chosen as a finalist in the competition for one of the California Institutes for Science and Innovation proposed by Gov. Gray Davis. Nanosystems involve the science, engineering and manufacturing of molecular-based structures. The UCLA-UCSB collaboration focuses primarily on molecular medicine and information technology.

The research reported in Science was funded by the Defense Advanced Research Projects Agency.