Self-assembling Circuits Using DNA May Represent Next Computer Breakthrough
The same DNA that carries genetic information may assemble electronic components when they become so minuscule that current manufacturing techniques no longer work, said researchers working on a $1.2 million project funded by the National Science Foundation to develop processes for submicroscopic DNA assembly. Their aim is to use the innate self-assembling properties of DNA to transport submicroscopic carbon "nanotubes" into place to function as transistors and connectors in computer circuitry.
"Reducing the size of features in electronic circuits isn't just a technical issue for the semiconductor industry," said Associate Professor of Computer Science Alvin Lebeck, the grant's principal investigator. "Our whole consumer culture depends on having smaller, faster circuits every year, and so does progress in much of medicine and science."
According to Lebeck, circuitry on such a small scale could lead to the creation of minute implants to monitor medical conditions, imperceptibly small sensors to detect and report hazards, and gadgets combining the functions of a voice-activated electronic organizer, desktop computer, wireless phone, game console, and television in a package the size of a wrist watch.
The semiconductor industry has grown to $140 billion in annual revenues by constantly producing smaller, faster, less expensive components that enable superior new products, Lebeck said. Key steps in the manufacturing process include shining light through a stencil-like mask and a lens onto a wafer of silicon, printing the pattern of the mask on the wafer, and using chemicals to etch the pattern into the silicon.
"The lithographic process can produce features only 90 nanometers wide. ("Nano" means billionth.) That's a thousand times thinner than a human hair," said Lebeck. "But there's a growing consensus that in ten to fifteen years, the process will hit a red brick wall."
The "red brick wall" phrase comes from a roadmap prepared by the semiconductor industry each year to focus researchers on barriers to producing still smaller features, showing problems with "no known manufacturable solution" in red.
"There's so much red up ahead that people are saying, 'Red brick wall! Red brick wall! The sky is falling! The sky is falling!'" said Lebeck. "If progress did stop, the industry would stagnate, computers, cell phones, and game consoles would stay the same year after year, and advances in things like medical imaging would stop, too."
Co-Investigator and Professor of Computer Science John Reif said, "The red brick wall may turn out to be a toll gate with a toll that's just too high. IBM spent nearly $3 billion on its latest chip manufacturing facility. If the current high rate of increased costs is sustained, in perhaps fifteen to twenty years a new semiconductor facility may cost the equivalent of the U.S. Gross Domestic Product. Basic research is needed to find an alternative."
Lebeck said the researchers named their project Troika after the three-horse Russian sleigh. Three universities are participating -- Duke, the University of North Carolina - Chapel Hill, and North Carolina State. "There are also three critical factors we have to balance in designing circuits for DNA self-assembly -- regularity, complexity, and defect tolerance," said Lebeck.
"The regularity refers to the waffle-like DNA nanostructures that will be scaffolds for building circuits," he said. "The complexity comes from laying out complex circuits on the waffles. Since DNA manufacturing sets up some initial conditions and allows devices to assemble themselves, errors happen. Defect tolerance is essential."
According to Lebeck, the Troika team is planning to build circuits from carbon nanotubes, which are rolled-up sheets of carbon atoms with walls only one atom thick. The researchers will use carbon nanotubes made with atomic arrangements that enable them to behave like semiconductors as the transistors that perform basic functions in a circuit such as acting as gates, which change the value of their output depending on inputs. They will use other nanotubes designed to be highly conductive to connect the transistors.
"Carbon nanotubes are only one nanometer in diameter compared to the current 90-nanometer state-of-the-art lithography," said Lebeck.
Reif said, "If we can build effective circuits at this scale with DNA and carbon nanotubes, the amount of circuitry that could be packed into a given area on a chip might be anywhere from 10,000 times to 1,000,000 times greater. That should mean dramatic advances in performance and reductions in cost. There's high risk in this research, but there may also be a high payoff."
The self-assembly process will use DNA "tags" applied to the end of nanotubes to place the nanotubes as desired on the DNA scaffolds. Because of its chemical structure, a single-stranded DNA molecule seeks to bind to a "complementary" strand to form a double strand, like two pieces of a puzzle fitting together. The tags work like Velcro, Lebeck said. "By putting a single strand of DNA on the end of a nanotube and the complementary single strand at one end of a waffle, we can get the nanotube to bind there."
"We know self-assembly of complex molecular-scale devices is possible because our bodies are made of them," said Reif. "But developing techniques to build complex and useful nanoscale devices of our own design is one of the great scientific challenges of the 21st Century."
Lebeck said the challenge of precisely self-assembling computer circuits becomes apparent when considering the problems that self-assembly of cars would present. Huge numbers of every part would be sent into an environment that allowed the parts to move around freely. Sticky tags on the parts would make the steering wheel "want" to attach to the steering column and the wheels "want" to attach to the axles, and so on.
"Some cars would be sure to come out with a steering wheel on an axle and a big wheel that should be on an axle stuck to the steering column," said Lebeck. "A car like that probably wouldn't work, but the architecture we're developing has to be able to design a circuit that works despite assembly errors."
The researchers will advance in stages toward the goal of DNA self-assembly of complex electronic devices such as computers, said Lebeck. The project will start by fabricating simple components that perform basic operations and then combine those components into working circuits.
Reif's laboratory has already used DNA self-assembly to create the waffle-like scaffolds and has used such scaffolds and tagged DNA strands to create a variety of structures. Assistant Professor of Chemistry Jie Liu's research group has developed techniques for creating unusually long and uniform carbon nanotubes, grids of intersecting nanotubes, and functioning transistors.
In addition to Lebeck, Reif and Liu, the Troika team includes Assistant Professor of Electrical and Computer Engineering Daniel Sorin, Assistant Professor of Computer Science Thomas LaBean, and Assistant Professor of Computer Science Hao Yan of Duke, Professor Sean Washburn and Assistant Professor Dorothy Erie of UNC, and Professor Paul Franzon of NC State.
The potential application of DNA self-assembly and exotic nanomaterials to computing and electronics has attracted the attention of researchers all over the world, said Reif. He and collaborators at New York University and the University of South Florida have received a National Science Foundation grant for DNA self-assembly of devices in which the DNA itself does the computing. A team of researchers from Rice, North Carolina State, and Penn State has created a memory circuit from clumps of gold atoms in disordered arrays. University of Tokyo scientists have developed techniques for using DNA to position metals with great precision. Researchers at Harvard have created nanowires from materials that are already used by the semiconductor industry, including silicon.
"These are exciting early days in building nanoscale computers based on DNA, carbon nanotubes and nanowires," said Reif. "Much remains to be done, but there is no shortage of ideas."