Lithography and Technology
In the computer industry, a continuing goal is to make computers faster and more efficient. No matter how fast a computer may be, no matter how efficient; one that is faster and more efficient, smaller and better, is always something to work at and strive for. Even today, with computers performing at such incredible speeds, there is a continual hunt to improve upon them.
As a computer science major, I have some experience with computer speed and lack thereof. Programming algorithms, which are the step by step methods performed in order to gain the desired output, are judged as efficient, based on the amount of time taken to run them in the worst possible case. This is known as the “Big O.” These values range from n to 2^n and possibly even larger, where n is the size of the problem to be solved. As your problem size gets larger and larger, your Big O values can shoot through the roof. To the average user, this translates into longer waits for operations to be completed and more frustration. In some cases, if the algorithm is bad enough, or the computer slow enough, you have computer crashes and we all know how unpleasant that can be. With faster computers that have more memory, this value becomes less and less significant and the intense urge to throw your computer out the window becomes less and less frequent an experience.
Knowing this, how can we make computers faster and how does chemistry tie in? In researching this topic, I kept coming across the same term over and over. All these computer experts kept mentioning something called lithography.
Lithography, according to the Dictionary of Computing & Information Technology, is a printing technique by which an image is fixed on a stone or metal plate. The image areas on the plate are coated with a water-repellant ink, while the non-image areas are protected from the ink by a film of water. Once this is prepared, it is possible to stamp this image on other objects. There is no raised surface to imprint the image on your object. The concept is, therefore, more like photography than like a printing press.
According to D. S. Halacy Jr. in Science and Serendipity; Great Discoveries by Accident, lithography is a concept that was discovered quite by accident. The person responsible, a young man named Aloys Senefelder, was an aspiring printer. He was determined to invent an inexpensive way of printing, and went through many sheets of paper in his endeavors to do so.
As the story goes, while toying with the idea of using stone tile for printing, he was down to his last sheet of paper when his mother asked him to write a list for her. Reaching for something handy on which to make the list and not wanting to waste his last sheet of paper, Senefelder grabbed a stone tile, and started to write his list, only to discover that his ink, made of soap, wax, and lampblack had dried up. In a hurry, he broke off a chunk of the dried ink and wrote on the stone tile anyway.
After using the list, Senefelder wanted to use the stone tile in his printing experimentations, so he attempted to wash the ink off. At this point, he made the discovery that the ink would not wash off. The water simply ran off of the ink. He also noticed that the stone absorbed the water. Playing with this some more, he discovered further that, once the stone was wet, it was impossible to draw on. Here was the solution that he was looking for. Not only were these products inexpensive, in using them, he would no longer have to etch away the stone to get a raised surface. He could simply draw on the stone with the ink and lock the image in with water. This discovery was the beginning of the process we now call lithography.
The chemistry behind this idea is very simple. It is based upon the idea that oil and water do not mix and also on the phenomenon known as “adsorption.” Adsorption, not to be confused with absorption, which also comes into play, is a chemical reaction where a molecule bonds to a surface without combining with it. The general process is as follows. An image is made on the particular surface with a fatty crayon or ink. The fat adsorbs on the underlying plate and produces a greasy substance. This substance is insoluable, meaning it will not dissolve in a particular solvent. The areas around the image are treated to seal them from any unwanted contact with the grease. Depending on the application, both sections can be reinforced, making them more or less soluable. At this point, when the solvent or ink is applied, the non-image portion will absorb it, while the image portion will not. It can then work similarly to a stamp where the portion that is printed is only the portion that does not absorb the ink.
Wait a minute. Lithography is about reproducing a stone image on paper. How does this tie into computers and computer speed and efficiency in particular?
“Integrated
circuits (semiconductors) are the key components of modern computers,
communication systems, consumer electronics, and the new generations of smart
machines and instruments. Microlithography is one of the most critical elements
of the semiconductor manufacturing process because it determines the minimum
feature size and the functional capabilities of the semiconductor. The quality
of the microlithography process is critical in determining the yield and cost
of semiconductors and hence the competitiveness of the electronics industry”
(Clemens, etc).
Printing microchips is very similar to other kinds of printing. Microchip lithography, although more technical than conventional printing, follows many of the same principles in copying a master design. Currently, optical lithography dominates in microchip manufacturing, as it has since it became necessary to produce microchips. Optical lithography involves shining light through a stencil, or mask, reducing the image through optics. The design is then imaged on a light-sensitive material called a photoresist. After this, the portion of the material that does not have the image is removed and the process is repeated. One microchip may have more than 20 lithographic layers with each one gaining in complexity.
In the earlier days, mercury lamps were used for the light source; however, laser technology is taking over the industry as the light source for microchips, especially for the more intricate layers. The “critical dimension” of a microchip refers to the size of the smallest feature on it. These features make up the transistors, electronic gates, and beltways that funnel information in the form of electrons through a computer. Cutting-edge lithography techniques use KrF excimer lasers at 248 nm to create critical dimensions of 0.25 µm. This is such a good lithographic source because, while it is wavelength stable and offers good power at UV wavelengths, the light is not overly coherent, and is therefore easier to work with.
Increasing how much can be positioned on one microchip increases the speed and/or the amount of memory the chip can store. About two decades ago, Gordon Moore, the cofounder and chairman of Intel at the time, stated that the number of transistors on a microchip should double approximately every two years, and industry has taken this to be the law of computer advancement. That means that the computers of today are roughly 2^10 or 1024 times faster than they were a mere 20 years ago. Current optical lithography techniques can only keep up with this pace for so long before they will simply be incapable of attaining the necessary critical dimension of halving.
Although we are coming close to reaching the physical limits of using lasers, the microchip industry believes that this form of optical lithography will be able to keep up with the decreasing sizes for another estimated 20 years. Optical techniques such as phase sensitive masks allow critical dimensions smaller than the wavelength of the source, but the cost of these techniques adds up considerably. As a result, some manufactures are hoping to utilize the ArF excimer, which emits at 193 nm, for the next step up in optical lithography. Others are looking for new solutions.
This brings me to the most interesting portion of my research. I am going to introduce a few of the possible solutions that are being researched. At present, I have information on 5 different options, with four of the possibilities being different types of lithography, often referred to as “Next-Generation Lithographies.”
The first one, I will look at is call EUV (extreme UV) lithography. This concept is very similar to optical lithography, except that it uses light rays from the spectral region between 40 nm and 1 nm. It has the ability to reach 1/3 the size of the current microchips with a critical dimension of about 0.1 micron. This can increase the computer chip’s speed 10-fold and increase the computer’s memory 1000-fold.
Another
next-generation lithography that may come into play is electron beam
lithography often referred to as SCALPEL (SCattering with Angular Limitation
Projection). They claim to be able to
make microchips with a critical dimension as small as 0.08 microns. This is accomplished by using, as the name
implies, high-energy electron
beams. Rather than using the mask to
block the unwanted beams, as occurs in most other forms of lithography, SCALPEL
uses elements with high atomic numbers in the desired locations to scatter the electrons
away. Since the mask does not need to
absorb the extra beams, it will not wear out as quickly.
The next lithographic solution is
X-ray lithography. This is also quite
similar to optical lithography, using synchrotron X-rays instead of UV light waves. The waves in this type of beam
are nearly parallel; they do not fan out as waves from X-ray tubes do, and so
require no focusing. Other properties include extremely high intensity, a broad
spectral band width in a smooth, featureless continuum (unlike other sources,
which are characterized by peaks, or emission lines), high polarization, and
small beam size, typically about 1 sq mm. It is
believed to be capable of producing microchips with a critical dimension of
about 0.1 micron. This is the option of
choice in Japan. In the United States
and Europe, however, there is very limited interest due to the cost of
production.
The final of the next-generation forms of lithography is ion lithography. Currently, they are working on developing a focused ion beam using He ions. The biggest advantage here is that, with the higher mass of an ion verses an electron, the chances of getting backscattering are quite a bit smaller.
The last possible solution is in synthetic chemistry. This is a very new idea and very little is actually tested. The idea behind this is to reduce the electrical contacts within a computer down to the size of one molecule. Thanks to the scanning tunneling microscope, this has potential. The key would be a technique called “self-assembly” where all a person would need to do is attach the endgroups to the desired molecular device and they would spontaneously assemble onto the metallic contacts, forming the device with no lithography needed. Given the potential that this could possibly lead to computers 100,000 times faster than our current ones, this idea would be the most optimal, however, there are so many hurdles to overcome that it is not very feasible to imagine that we will be ready with this technology for quite some time.
With all of these possibilities, the future of computers looks promising and exciting. With the advances in technology, computers could soon be so fast that the effects of bad programming will be negligible. Whichever road ends up being the most feasible, it is doubtful that these will be the last in microchip advancement.
Bibliography:
Bjorkholm, John E. “EUV Lithography—The Successor to Optical Lithography?” Internet. http://developer.intel.com/technology/itj/q31998/articles/art_4.htm (3rd Quarter, 1998.
Clemens, James T., et al.
“X-ray Lithography in Japan.”
Internet. http://itri.loyola.edu/ar93_94/xl.htm (March,
1994).
Halacy, D. S., Jr. Science and Serendipity; Great Discoveries by Accident. Philadelphia: Macrae Smith Company, 1967.
Hardin, R. Winn. “Optical Lithography Gets Nod From Microchip Industry.” Internet. http://www.spie.org/web/oer/february/feb99/cover.html (February, 1999).
Meadows, A. J., et al. Dictionary of Computing & Information Technology, 3rd Ed. London: Kogan Page Ltd, 1987.
Reed, Mark. “The Promist of Molecular Computing.” Dr. Dobb’s Journal: Software In the 21st Century. Centura Software Corporation, 2000.
Scott, Karen. “Ion Beam Lithography.” Internet. http://diva.eecs.berkeley.edu/~tking/ibl.html (August, 2000).
“Synchrotron Radiation." Microsoft® Encarta® Online
Encyclopedia 2000.
http://encarta.msn.com © 1997-2000 Microsoft Corporation.
Vicary, Richard. Manual of Lithography. New York: Charles Scribner’s Sons, 1976.