Nanotechnology
Nanotechnology refers to the construction and use of atomic machines—functioning consortia of atoms that are less than about 1,000 nanometers in size. A nanometer is one billionth of a meter. The technology is relevant to computer science as it could potentially be used to construct molecular computers, which could be smaller than the size of a bacterial cell and capable of performance exceeding technology currently in use.
Molecular machines are in the early developmental stages. Several reports have described the physical rearrangement of atoms into a desired configuration, such as the use of a scanning tunneling microscope by IBM (International Business Machine) researcher John Foster to push xenon atoms around to spell "IBM." Other reports have described the atomic assembly of crude interlocking gears. And, in 1999, a molecular electronic research team successfully constructed a "molecular wire" circuit based on a polyphenylene polymer and a functional molecular switch.
In the future, it may be possible to use atomic assemblers—robotic arms analogous in function to a construction crane--to position atoms. This is called positional assembly. There is precedent for molecular positional assembly. Ribosomes, which position and link together amino acids in a specified sequence to form a specific protein, are biological assemblers. Atomic assemblers would be capable of assembling atoms into three-dimensional configurations and attaching the atoms together by a variety of chemical means. Assemblers would theoretically also be capable of constructing copies of themselves, or replication. Self-replication at the atomic scale was first studied by John von Neumann in the 1940's.
The construction of molecular computers hinges on the refinement of techniques of molecular electronics, in which chemistry is used to assemble circuit components, such as the building of the polyphenylene circuit. Molecular electronics began with the realization that molecules comprised of long chains of atoms could conduct electricity. Individual electrons hop from one molecule to the next in the chain. This movement differs from conventional current, which results from the bulk movement of electrons through the material of the wire. To harness the molecular current, it is necessary to have controllable "gates."
Several gating designs exist. One feasible design for a molecular computer is based on sliding rods that act to block or unblock one another's sliding action as they interact at certain sites, or locks. The controlled interaction of the locks would allow information--electrical or otherwise--to pass from rod to rod. It has been calculated that the number of locks necessary to build a simple 4-bit or 8-bit general-purpose computer would occupy only a few nanometers of space. Another design is to attach a molecular "side chain" to a molecular wire. The molecule could both accept and donate electrons. Control of these processes would allow for the controlled movement of electrons through the molecular gate. A functional molecular gating device has been constructed at Bell Labs.
The next goal is to find a chemical means to construct what is called a three-terminal transistor. These would enable electricity to flow in a web-like pattern, rather than in a straight line. In turn, complex logic devices could then be made. Current research is exploring the use of chemical processes to construct the transistors, rather than the use of atomic assemblers.
A drawback to nanotechnology is the problem of random defects. While molecular structures self-assemble very accurately, compared with human-directed assembly processes, perfection is impossible, and defects in assembly will occur in a random way. For a molecular wire, which requires molecular precision, such imperfections are lethal to the function of the wire. The question of how molecular defects will be handled is an important design constraint. The focus is to find circuit architectures that will be able to function with fidelity even with defects present. Perhaps circuit design will incorporate the presence of a certain number of defects. Or, analogous with biological systems, perhaps parallel redundant molecular wires could be used. Also, molecular computers may be designed to operate similarly to the human brain, where networks of neurons permit information flow, even with abnormally functioning neurons.
Shrinking computers to atomic size would allow computerized control of processes inside many currently inaccessible niches, such as the human body. Medical treatment could potentially be revolutionized. If chemical reactions can be harnessed to create the logical electrical pathways needed for computer operations, then molecular computers could become a reality.
This is the complete article, containing 709 words
(approx. 2 pages at 300 words per page).
