The History of Electronics

The first major leap in electronics was in 1937, when Claude Shannon published his thesis. In his thesis, he introduced switches for arithmetic and logic. The first object to use these switches was Konrad Zuse’s Z3, which used mechanical switches. The first programmable computer was the British Colossus, which was used to crack the Enigma. It could be re-programmed by rewiring the system. The first digitally programmable computer was the revamped ENIAC, which incorporated Von Neumann architecture, which basically means that a computer’s programs are held electronically, and not through hardware.

Today’s electronic chip is made by connecting millions of tiny switches called transistors. Transistors are minute “gates” that can have electricity (1) or be empty (0). These transistors, combined, create the wonderfully complex object known as the personal computer. Today’s chips are made of silicon, element #14 on the periodic table. Silicon is a semiconductor, which means it gives some resistance to electrons, but still allows it to flow. Chips are made through a process called optical lithography. In optical lithography, light and a stencil are used to “cut” layers of silicon away to create an image of the stencil onto the silicon. This can be done perfectly all the way down the wavelength of light.

Creating “Nanocomputers”

Creating “nanocomputers” should be very simple because they’re smaller right? Well, making a transistor smaller is not very difficult, but not simple enough to call easy. To make a small transistor, we just shrink a normal chip. To shrink it, we must get silicon smaller than the wavelength of light, so we use electron lithography. That part is easy.

The problem lies in the mysterious mists of quantum mechanics.

Transistors must keep certain proportions. If the proportions are too big or too small, the transistor will not work. When we make transistors as small as atoms, they become small enough to be affected by electron tunneling. Basically, as electrons get more and more constrained through certain passageways, they can “jump” a few nanometers away. This “jumping” can cause other electrons to “jump” as well, making these ultra-thin transistors virtually nonexistent, allowing electrons to bypass them as simply as ducking through a low door.

This same effect is used in the Atomic Force Microscope (AFM), an instrument used to view atoms. The only problem with this is that it now becomes necessary to find new materials that will allow different proportions, making the gates thicker while leaving the rest at the nanolevel.

An easier way to create “nanocomputers” is to allow Nature to do much of the work. Nature starts things small and builds them up, starting with a few atoms and adding more, creating large structures. For example, in 1974, Avi Aviram and Mark Ratner created a molecule. At one end, the molecule was full of electrons. At the other, it was empty, wanting electrons. When the molecules were made into a liquid form, they would line up in rows like molecules in a magnet, allowing electrons to go one way, but not the other. This was called a diode.

So how will we string these minute pieces together? Carbon nanotubes (CNTs) may help us solve this, by creating minutely small pathways for electrons to pass from chip to chip. Unfortunately, we have a few obstacles to overcome. The creation of carbon nanotubes are not easily controlled, creating a variety of different lengths, shells, and properties. Also, if CNTs are heated up quickly like in a computer processor, it reacts with oxygen violently and may explode.

Courtesy of Molecular Biophysics Group, TU Delft

Materials for Electronics

We have discussed different ways to make silicon chips, but why aren’t other materials used?

Many materials have semiconductor properties, and one area that has favorable properties is the area of organic materials. Organic semiconductors are made of the same things living creatures are made of. These polymer semiconductors are flexible, thin, and smooth. This is good for small, flexible computers we can make in fabric or other wearable materials. It is also good for roll-up computers and monitors. Some limitations occur. As it is well known, organics break down, such as a rotting apple. Also, oxygen attacks organics, as it is seen when fruits turn brown. To keep organic semiconductors, they must be protected.

But if we use different materials, why do we still have to use electrons?

Electronics are based on on/off switches, so as long as we can use other materials that allow something to either “be there” or “not be there”, we can use the same type of programming/hardware. A good example would be of how we now use fiber optics. Optic fibers are the thinnest, purest pieces of glass formed in mile-long wires. Laser light, the purest form of light, is sent through a glass wire, traveling at 186,282.397 miles per second, or about one foot per nanosecond. At this rate, a signal sent through the optic fiber could travel around the entire earth seven times in a second! So why don’t we use it now? In actuality, we use it every day for sending information from one end of the earth to the other, creating our super-speedy Internet. Still, optic fibers are expensive to make, and their speed is slowed by electronic switchers, which operate about three times slower than optic fibers. To make this faster, we must be able to re-direct photons (particles of light) with efficiency.

Scientists have recently created a material that allows such change of directions. In the year 2000, Mark Kuzyk used quantum mechanics to see how well light-switching could be done, theoretically. To his surprise, he found that the physical limit was at least 30 times bigger than optic fibers in use. Using his calculations, scientists bonded buckyballs, two-nanometers carbon spheres, to a special polymer that can be controlled. This controlling factor allows us to redirect the light, allowing totally light-based networking.

While we decide to totally change the materials computers are made of, we could also change the way computers work. Our modern computer uses a single processing unit to control where information goes and what processes to run. Newer computers use two or more processing units, allowing more work to be done. The problem with this kind of computer is that each bit of data must run through this unit, creating delay when running many different processes, or a few large programs.

How do we solve this delay? Perhaps we can get rid of the use of central processing units (CPUs). The human body runs amazingly well on this kind of processing, in which millions of switches operate without instruction from a single source alone. By creating a system of immense parallel switches, a computer could do all its processes almost instantaneously, without each bit of data waiting in line for a turn at the CPU.