Classic Computer Magazine Archive CREATIVE COMPUTING VOL. 11, NO. 1 / JANUARY 1985 / PAGE 30

The National Submicron Facility; the little laboratory that does big things. Hillary Rettig.

In the beginning, there was ENIAC. ENIAC, short for Electronic Numerical Integrator and Computer, was the world's first electronic digital computer. It was built less than 40 years ago by a team of engineers at Bell Laboratories and was considered a marvel of its age, a herald of things to come.

Now, with the wisdom of hindsight, we see ENIAC as an important, essential development, but also as nothing more, nothing less than a 30-ton 20,000-square-foot-pocket calculator. It took more than 50 engineers and technicians and nearly 18,000 vacuum tubes to keep it running, and at best it could perform 5000 calculations per second.

By contrast, today's microcomputers, which fit on a desktop and contain not vacuum tubes but solid-state circuitry, routinely perform more than 1,000,000 calculations per second: a factor of 200 improvement in less than 40 years--an unparallelled achievement in the history of human invention. And, as if to impress even the most jaded science fiction fan, researchers have developed experimental silicon chips which themselves can perform more than a million calculations in a single second--all in an area the size of an infant's fingernail. Elements of the individual electronic switches (or "gates") in these highly integrated silicon chips can be as small as 1500 Angstroms (one Angstrom equals one ten-billionth (10.sup.-10) meter, or approximately 1/25,000,000 (2.5 X 10.sup.-7) inch).

These developments herald the emergence of a new technology: the technology of the ultrasmall. This "hot spot" of current research is a world whose largest dimension is one micron, or one millionth of a meter, and whose smallest dimension is equal to the width of a small cluster of hydrogen atoms. In this world, engineers and scientists study the intermediate range between bulk solid-state physics and subatomic physics--atoms and ions in small clusters or crystals, or in ultrathin layers--where the physics and potential applicaitons of matter may be startling different from anything we have seen before.

Through innovations and improvements of technologies such as digital electronics, chemical and biological microsensors, optical wave-guides and fiberoptics technologies, submicron research is laying the foundation for the next wave in what is now commonly referred to as the "microelectronics revolution." A laboratory at Cornell University, the National Research and Resource Facility for Submicron Structures (or National Submicron Facility), is leading the way for the United States' progress in this exciting and important new technology.

The National Submicron Facility is a pioneer institution. Established by the National Science Foundation in 1977, it remains the only laboratory where any qualified U.S. researcher can come to use the highly specialized and expensive tools of submicron science and technology. These researchers--scientists and engineers in fields ranging from electrical, chemical, and materials engineering to physics to medicine and agriculture--come to the Facility to utilize equipment and other resources unavailable elsewhere. Together with the staff of the Facility, they explore the microworld with a freedom heretofore only dreamed of. The results so far have been stunning:

* The world's smallest artifacts: letters so small that using them you could reproduce all thirty volumes of the Encyclopaedia Brittanica, on a postage stamp;

* A device that can measure the change in the earth's magnetic field caused by the blink of an eye;

* A wire so thin that it "traps" electrons--it no longer conducts electricity, but acts as an insulator instead. Or, if you prefer, a gas so dense that it becomes a metal and conducts electricity; and

* Electronic devices so small that 30,000,000 of them would fit on a single 1/4-inch-square chip.

These results and others have established the National Submicron Facility as a laboratory that defies superlatives and where the extraordinary begins to seem almost commonplace.

Not the least of the marvels of the National Submicron Facility is its research laboratory, the Lester B. Knight Laboratory, named for a Cornell alumnus and patron. The building itself stands out from other on the Cornell campus. Visitors who fly into Ithaca, NY, on a clear day, or who view the Cornell campus from its bell tower or another lofty building, invariably comment on the giant, 30-foot <[mu] logo outlined in colored crushed rocks on the laboratory roof: the National Submicron Facility is, quite literally, submicron!

To those uninitiated in submicron research, a visit to the Knight Laboratory is apt to hold one surprise after another. You must meet the submicron world on its own terms, and the Knight Laboratory is designed to provide an environment in which macroscopic scientists and engineers can adapt to the needs of their microscopic research. For example, inquisitive visitors often wonder why there are no windows throughout most of the laboratory. The walls stretch unbroken and white, and the building, glistening in the sun, resembles a giant sugar cube. The windows were left out, we are told, to minimize thermal imbalances caused by weather.

You enter the Knight Laboratory via a reception area that is ordinary enough and sign a guestbook. To your right are the computing and computer design facilities, the only technical facilities situated in a "normal," unaltered environment. A quick look around reveals an impressive array of computing power: a Digital Equipment Corporation VAX11/750 computer with several graphics peripherals, including a Grinnell Color Frame Buffer for ultra-high-resolution color graphics imaging, plotters, and graphics terminals. There is also a CALMA computer-aided design system with high-resolution color terminals, run by a Data General Eclipse S280 computer. The computer facility is generally one of the busiest places in the laboratory, with people working in it 24 hours a day, seven days a week.

Submicron research encompasses three general operations: the design, fabrication, and analysis of microstructures. Design work using computers is the only technical operation in the National Submicron Facility that does not require a specialized environment. Fabrication and analysis of microstructures are done in the main research area within the Knight Laboratory, called the clean room.

The Knight Laboratory clean room is rated Class 400, which means that there are fewer than 400 half-micron or larger particles in every cubic foot of air (for contrast, normal office air is Class 300,000). The clean room also contains 24 superclean laminar-flow hoods in which the air is filtered down to Class 10 or better. The precautions are understandable: A single small dust particle can look like Mount Rushmore when you see it sitting on top of your carefully developed experiment.

You can enter the clean room only through an airlock, a special passage that isolates the laboratory from the dirt and grime of the outside world. The air pressure inside the airlock is higher than that in the reception area, so that dust is blown outward when the door is opened. A large sign warns, "No Smoking, Eating, or Drinking." Other rules include: no cosmetics, no writing implements other than ballpoint pen, no bare legs (in the summer time), and no cleated hiking boots (in winter).

You do special clean room clothing: booties, bonnet, and a lab coat. You pass through yet another door (more air blows outward), and enter the laboratory proper.

People pass by, shuffling in their booties. Your guide explains that there are usually anywhere from 15 to 30 people in the laboratory at any given time: Facility staff, Cornell faculty and graduate students, and students and researchers from other university, industrial, and government laboratories who have come to the Facility to conduct research. Regular Regular staff and students wear blue lab coats; novices and visitors wear white. If there is a problem, you look for someone wearing blue.

The laboratory from floor to ceiling is a panorama of stark, spotlessly white surfaces. There is a low but continuous humming noise from equipment and air conditioning--after a few minutes in the laboratory, you cease to hear it. A common sight is the hose to the vacuum cleaner, a giant, snakelike, contraption that coils around itself as the janitress disconnects it fron one outlet of the immense vacuum system and plugs it into another. (Needless to say, in a laboratory so dedicated to cleanliness, such mundane functions as vacuuming and laundry take on a new importance. The Facility's laundry bill alone comes to several hundred dollars each month.)

The 7500-foot-square clean area is composed of a large central room that houses machines for such "large scale" processes as evaporation and ion milling. Around it are 12 smaller rooms which house the instruments that perform the very delicate and precise fabrication and analysis of structures so small that the largest is 1/100 the diameter of a human hair.

Because vibrations from traffic on the roads near the Knight Laboratory could ruin a delicate experiment, each of the smaller rooms rests on its own pad of "floating" concrete, 2-1/2 feet thick, isolated from the foundation and adjacent rooms. In addition, the entire building rests on a foundation of specially-compacted earth to insulate it further from vibration.

The basement of the Knight Laboratory houses the support system for the building, including special air and water purification systems and a complicated electronics system that feeds each room individually. It is a huge room, a jungle of pipes and wires and cylinders that seems to stretch on in all directions. The central air conditioning unit, AC-1, with its banks of high density filters, sits along one side like a stranded Pullman railroad car. It recirculates the air within the entire Knight Laboratory twice each minute. There are also four auxiliary air conditioners for specific areas within the building, as well as a 700 gallon deionized water tank. A monitoring panel for all of the support systems is visible through the glass-walled airlock. It has dozens of dials and lights--green when all systems are "go" and red if there is a failure, in which case there is also what the guide calls an "impressive" alarm system.

The Knight Laboratory is an engineering and design marvel in itself, and yet it is only the setting for something even more marvelous: an array of highly specialized equipment that fewer than ten laboratories in the world can match. Scientists use one machine to build submicron devices by sandwiching together superthin layers of materials. They use another to take a single ion and "implant" it at a specific position within a crystal: the properties of the crystal change dramatically with the position of the impurity. Yet another machine is used to etch amazingly tiny and intricate patterns into semiconductor material to study techniques for making new and better integrated circuits.

Some of the machines fill entire rooms with their wires and coils--a scientist working at one looks like Captain Nemo at the helm of the Nautilus. Visitors are often surprised to see that such big machines are used to study such little things. The machines are large because, despite the elaborate precautions taken within the Knight Laboratory itself, they must be incredibly stable when working with single atoms and atomic distances.

"Users come to our laboratory and are often astonished at the variety of equipment that is here for their use," comments Edward Wolf, director of the Facility and professor of electrical engineering at Cornell. "Nor just equipment, but expertise: our staff and, of course, Cornell faculty and students, provide a unique and exciting interdisciplinary research environment. Many of the people our visitors meet are acknowledged world leaders in their fields."

Wolf feels that a national laboratory is an idea whose time has come, especially in a field such as submicron research, where even the most basic tools of the trade are enormously expensive. The facts certainly seem to bear him out: less than three years after the Knight Laboratory was completed, the Facility now hosts over 40 user projects from university, government, and industrial laboratories throughout the United States, as well as nearly 40 more from within Cornell. Another primary objective in establishing the Facility was to foster graduate research in submicron fabrication. Here, too, numbers speak louder than words: more than 100 advanced degrees have been granted to students working at the Facility, and more than 150 students may be conducting research at the Facility at any given time.

Many of the non-Cornell users visit for one to three weeks several times a year, working intensively with the equipment and expertise not available elsewhere. The Facility also participates in Cornell University's Program on Submicrometer Structures, Prosus, an industrial affiliate program, which seeks to establish closer ties between industry and the academic community.

Wolf sees the alliance between government, academe, and industry as a crucial one for the United States. "The recent tax credit legislation for corporations interested in university research and instrumentation has really stimulated our university-industry interaction," he explains. "The exchange of ideas, resources, and personnel allows us to get much, much more out of every dollar that we spend on research--and I don't have to tell anyone how important that is."

As has been the case in other frontier fields, for example, astronomy, technology and basic science go hand-in-hand in submicron structures research. Advances in one lead to advances in the other, and progress occurs quickly and in often unexpected directions. It is like the old chicken-and-egg story: which comes first, novel computing techniques which allow the creation of new microstructures, or novel microstructures which lead to advances in computing? For this reason, nearly all of the research in the National Submicron Facility involves some innovation in computing and computer engineering. Some examples of work at the National Submicron Facility which may lead to new submicron structures for integrated circuitry include:

* The development of materials specially configured for ballistic transport in semiconductors and metal silicides with unusual properties or with properties which make them suitable for specific uses;

* The development of new fabrication techniques, such as reactive-ion-beam etching, an anisotropic process which combines both chemical and physical etching mechanisms to create submicron structures impossible to achieve through traditional "wet" etch processes;

* The pushing of "conventional" fabrication and analytical technologies, such as electron-beam lithography and scanning transmission electron microscopy, to their limits;

* The application of "conventional" technologies to new areas, such as bubble-memory devices;

* The investigation of basic physics as applied to microcircuitry: such problems as electromigration (current-induced atomic transport, a significant failure mode in highly integrated circuitry) and parasitics (the creation of unintentional coupling and other false elements in complex circuitry), as well as field-emission studies of ions, which could eventually lead to ion-beam lithography; and

* Improved image processing design and analysis capabilities.

Researchers interested in using the Facility's resources should first contact Professor Wolf. A User's Manual is available upon request; outlines both the criteria for projects of the Facility and lists the equipment and other resources available to users. Criteria for Facility user projects include:

* A project should involve microminiaturization, especially in the submicrometer regime, in a substantial and innovative way;

* A goal of the project should be to advance significantly the art of submicrometer technology or its application to engineering or to scientific research;

* The chief purpose of the project should not be to make use of services that are available commercially;

* The nature of the project should be such that the specialized equipment or expertise available at the Facility will make an essential and important contribution to the outcome of the work;

* The project should provide educational opportunities for personnel associated with the work.

The user community of the National Submicron Facility now includes university, industrial, and government laboratories from throughout the United States, including: Carnegi-Mellon University, University of Florida--Gainesville, University of Pennsylvania, University of California--San Diego, Drexel University, Howard University, Bell Laboratories, Eastman Kodak Company, Vought Corporation, Sandia Laboratories, Lawrence Livermore Laboratory, and Woods Hole Oceanographic Institute.

Wolf encourages anyone with an interest in using the resources of the Facility to get in touch with him: "In addition to academic, goverment, and industrial laboratories, small businesses involved with advanced technology might benefit greatly from interaction with the Facility, as might researchers in "nontraditional" fields such as biology, medicine, and agriculture." Anyone interested in receiving more information about the National Submicron Facility is invited to write him: Professor Edward D. Wolf, Director, National Submicron Facility, Knight Laboratory, Cornell University, Ithaca, NY 14853.

The National Submicron Facility has generated much excitement throughout the technical community, and there is a steady stream of visitors--as many as 30 in a single week--from laboratories throughout the world. The popular media have also adopted "the little laboratory that does big things," and articles have appeared in such diverse publications as Newsweek, Barrons, Popular Science, and National Geographic. WCBS and WIXT-Syracuse have both managed to produced television segments about the Facility despite the difficulties of preparing camera equipment for the clean laboratory. One reported, when meeting Ed Wolf for an interview, exclaimed, "But your're such a big guy to be working on such little things!" Wolf, who once played basketball for Kansas State, grins, "I guess he thought it was a lab for small scientists instead of small science."