It’s almost a law of nature: computers just keep getting smaller. This phenomenon is so widely recognized that it even has a name: Moore’s Law (of no relation to the author), which states that the number of transisters that can fit on a chip doubles every 18 months. A welcome by-product of this ongoing process of miniaturization is a steady increase in computers’ power and speed.
Moore’s Law has held true for 40 years. However, computer technology is approaching what many engineers consider the “brick wall” for existing technology. In less than ten years, we may be faced with the fact that computer components as we know them simply cannot get any smaller. Such a scenario could cause problems for a society that has grown to expect more powerful computers (and cooler electronic gadgets) every year.
Against the Wall
Just what is this technological limitation? To understand, let’s look at how today’s computers work. Essentially, computers are just very dense electrical circuits made up of two main components: wires and switches. The wires conduct electricity, while the switches enable a computer to make decisions. Anything a computer does can be expressed in a language of ones and zeros, and it is the on/off nature of the switches, or “transistors,” that allows ones and zeros to be stored, added, multiplied, and logically combined in electrical circuits.
In modern computers, these devices are now so small that millions of them could fit on your fingernail. Both wires and switches are made of materials that conduct electricity. Wires are made out of ordinary metals, like copper or aluminum, while switches are fabricated from materials such as silicon, termed “semiconductors” because of their selective conductivity. Through a process called photolithography, wires and switches are photographically imprinted onto silicon wafers to create microchips.
The flow of electrons through wires and switches results in the transmission of electrical information. Computers get faster as wires and switches get smaller and are positioned closer together. As the distance electrons must travel becomes smaller, they take less time to traverse a circuit, and a processor can move on to the next piece of information more quickly. For the past half century, a continual process of ever greater miniaturization of wires and switches has allowed the industry to create faster and faster computer chips.
In a few years, however, wires and switches will become so small—just a few nanometers wide—that they approach the size of individual atoms and molecules. (A nanometer is one billionth of a meter, or about 10 atomic diameters.) At those sizes, the conventional laws of physics that govern the flow of electricity through metals break down and give way to the laws of quantum mechanics. In short, the switches stop working.
A New Approach
Scientists are already at work on alternative technologies that would make use of quantum mechanical effects rather than be defeated by them. More than just perpetuating Moore’s Law, technologies now in development may bring unprecedented leaps in computing power.
One such technology, molecular electronics, may bring about a fundamental change in the way we build computers. Unlike current computer technology, molecular electronics will use individual molecules as the building blocks. Since molecules bond and interact according to the rules of quantum mechanics, they are the logical choice for the building blocks of nanometer-scale electronics. By constructing wires and switches out of individual molecules, molecular electronics may permit the design of computers up to a million times smaller and many times faster than today’s.
To construct a nanometer-scale computer, we will need to build molecular wires and switches—that is, single molecules capable of conducting and controlling electrical currents. Wires are the simpler circuit element, and they will be the easier one to construct on a molecular scale. By linking a number of hexagonal molecules called benzene rings (C6H6) into a long strand, a conductive wire can be created. The atoms in benzene (and its chemical cousins) have the unusual property of sharing electrons in a “delocalized molecular orbital”—a region of freely-moving electrons above and below the plane of the molecule. The natural fluidity of its electrons makes benzene highly conductive to electricity. When linked in long chains, called polyphenylene chains, benzene rings can transmit electrons freely from end to end. Such a molecular wire has actually been synthesized, and experiments have shown that it is much more conductive than metal wires of the same scale.
Switches are the more complex elements in a circuit, and they will be the more difficult ones to design and fabricate on a molecular scale. They will be constructed from the same polyphenylene backbone; but by adding molecular subunits or “substituents” to that backbone and by introducing insulating subunits into it, molecular devices like switches can also be created. A few molecular-scale switches have been synthesized and tested; however, much of the research on molecular switches is still only theoretical because of the difficultly at present of constructing and testing the molecules of interest.
Modeling a Molecular Switch
One switching device important in logic circuitry is a “rectifying diode,” which conducts electricity through itself in only one direction. When combined with other diodes, it can emulate the switching abilities of its more complex cousin, the transistor. The summer before my senior year of high school, I participated in a research project to predict the behavior of a possible design for a molecular electronic diode.
That summer, I was an intern in the Nanosystems Group at The MITRE Corporation in Reston, Virginia, where I was part of a research group led by Dr. James Ellenbogen. My project was to use computer modeling to predict the electronic behavior of this possible molecular electronic diode—a molecule that had never actually been built. My goal was to calculate quantitatively the behavior of the molecule as if it were in a circuit, to see whether it would in fact act as a rectifying diode. I used a commercially-available computer simulation program to estimate the chemical properties of the molecule. Then, once I had the chemical properties, I used a computer program I wrote to derive the molecule’s electrical properties—its behavior in a circuit. My research showed that the molecule probably would operate as anticipated if it actually were synthesized and used in a molecular circuit.
Remaining Mysteries and Challenges
While this and other research moves us closer to the first molecular electronic computer circuits, much work remains before they will be the basis for a feasible alternative to existing microchip technology. As yet, no one knows how to build a molecular electronic computer chip. Synthesizing some of the complex chemical structures needed for molecular circuits seems difficult given our current knowledge of chemical synthesis. Even more daunting is the challenge of precisely positioning individual molecules on a nanochip and connecting them to other circuit elements. The processes used to manufacture contemporary silicon microchips are not precise enough to manipulate individuals molecules. And even if chemists could build the molecules and engineers could organize them in orderly circuit patterns, it will require technological breakthroughs before we can interface such tiny circuits with larger-scale devices like keyboards and monitors.
Despite the remaining hurdles, molecular electronics shows great promise and, if realized, would bring about radical changes in the way we view computers. We may find ourselves living with “smart dust”—computers that are nearly invisible to the naked eye, yet almost as powerful as current PCs. Such microscopic computers could be integrated with everyday objects in our daily lives to make our lives easier and better. Potential applications of this kind of miniaturization include medical monitoring devices, lifelike human prosthetics, and “smart paint” that could act like a giant TV over an entire wall.
Get Involved
Exploring this future face of computing by doing research was a highly rewarding experience, and I encourage others interested in science to find ways to get involved. You can always find research organizations or corporations in your area that are interested in having summer students. Whether you are interested in computers of the future, biotechnology, chemistry, or some other area of science, nothing can compare with real-world experience.
About the author: When he wrote this article, David Moore was a freshman studying electrical engineering at Caltech, after having graduated from the Math/Science Magnet Program at Montgomery Blair High School in Silver Spring, MD. For his research on molecular electronic diodes, Moore was awarded Second Place in the 1999 Intel Science Talent Search. He is now a graduate student at MIT in the Computer Science and Artificial Intelligence Lab. His web site is davemoore.org.
Further reading