"Nanotechnology is a natural evolution of research that has occurred over the last 40 years," says Clifford Pollock, Ilda and Charles Lee Professor of Engineering at Cornell University and the director of the University's School of Electrical and Computer Engineering. "One of the questions almost any researcher can ask about any problem is: what's happening at the nanometer scale?"

Simply being aware of the nanometer dimension has helped people look at old problems from a new perspective:
• Someone making a fabric realizes that if the fiber has the right nanometer structure on its surface it won't collect dirt or grease.
• A pharmaceutical researcher realizes that a drug works only when it hits the protein of a certain type of cell. Instead of flooding the entire body with a drug, she can attach the drug to a nanostructure that attaches itself preferentially to the protein.

Since 1996, Pollock has served as the faculty leader on the design team for Duffield Hall, a new nanotechnology facility currently under construction at Cornell. Upon completion in 2004, Duffield Hall will house at least four areas of highly specialized cross-disciplinary research:
• Nanofabrication—making small structures
• Nanocharacterization—seeing small things
• Materials growth—synthesizing new materials at the molecular level
• Nanobiotechnology—exploring the science of life at the dimension of DNA molecules

(For more information on Duffield Hall, see Duffield Hall Nanotechnology Research Moves Forward.)

Even in an era of tightened budgets, nanotechnology—the science and engineering of assembling materials and components atom by atom, or molecule by molecule, and integrating them into useful devices—continues to attract federal and state funding. The Clinton administration devoted $1 billion to nanotechnology research programs and the Bush administration shows no signs of turning off the tap, recently proposing to increase the number of federally-funded nanotechnology centers from one to three.

Accommodating these research activities requires a very specialized facility, one with ample cleanroom space and very low vibration criteria in its physical structure. Materials management is also an important consideration, given the nature of the chemicals employed in nano-scale research and the potential for adverse interactions. Finally, the fact that researchers from multiple disciplines are involved also presents a challenge to planners. According to Pollock, Duffield Hall will serve every science department at Cornell, as well as more than 400 visitors from industry, universities, and national labs.

"It can't just work for undergraduate research," he says. "It has to work for industry, for graduate students, and for post-docs, too."

Not Your Father's Cleanroom

A cleanroom in a nanotech research facility varies in a number of respects from a traditional cleanroom, according to Pollock.

The traditional cleanroom is designed for manufacturing semiconductor devices, so chip yield is the driving force. One speck of dust in the wrong place can kill a chip. A trace impurity, like sodium from a fingerprint, can destroy the performance of a transistor.

"A nanotech research facility is not concerned with manufacturing, so typically the cleanliness requirements can be relaxed. Its yield is students or scientific results, not chips."

Where the traditional cleanroom is primarily run by semiconductor engineers, the nanotech cleanroom will accommodate a broader mix of people (ranging from biologists to engineers to chemists and so on) working with a wider variety of materials (glass, semiconductors, plastics, epoxy, biomaterials, polymers, etc.). Students and novices will also be in the cleanroom, which can reduce throughput, increase the dirt, and complicate training and supervision. As a result, there need to be methods established to handle unconventional materials and a dynamic population.

"You can't just take an existing cleanroom and convert it into a new facility," says Pollock.

Users Drive the Design

According to Pollock, the nano user is the key to designing an effective facility. Different types of users will have different expectations based on the specific requirements of their research.

Semiconductor research—This research typically requires the most extreme cleanliness for a number of reasons. While semiconductor manufacturing is focused on yield, in a research lab the work addresses novel new devices, fabrication techniques, and electronic materials. Cleanliness in a research lab is not needed necessarily for chip yield, but for operations such as bonding two pieces of material together. Any dust that gets between the layers can potentially ruin a bond. The semiconductor fabrication facility (or fab) usually requires very clean tools (meaning that no trace impurities are on them). A few atoms of the wrong dopant will ruin a transistor, so transistor processing requires a dedicated set of tools where nothing else goes in them. Most processes are on the surface, so the need for deep etching is limited.

Microelectromechanical systems (MEMS)—Researchers making MEMS require tools to remove material and etch deep trenches. MEMS research grows thick layers of oxides and other materials, so rapid growth and rapid etching are critical. MEMS researchers generally aren't as concerned about impurities at a fine scale.

Biology—Pollock points out that biologists present the biggest challenge in a shared cleanroom. In exploring ways to use biomolecules such as DNA to create self-aligned structures, biology researchers employ chemical solutions containing organic and biological molecules not usually allowed in cleanrooms.

"Segregating biologists and semiconductor people is critical," says Pollock. "Special hoods and handling processes need to be installed. This also requires careful placement of air handlers so that different airstreams are not mixed during recirculation. Walls and doors need to be placed between areas preventing access to certain areas by the biologist."

Cleanliness is also a critical issue to biologists, but in terms of biological sterility rather than dust.

"If you're going to put all these functions together in a cleanroom, know up front that you have to do some very careful designing to segregate the dirt and impurities," says Pollock. "And make sure you know both sides are satisfied."

Vibrations and Magnetic Fields

In a nanotech fab, vibrations must be minimized to assure the efficient operation of the extremely sensitive instruments (high-resolution microscopes, etc.) used in the research. Because many instruments employ electron beams, controlling magnetic fields is also critical.

"You use electron beams to write small patterns or to do your microscopy," says Pollock. "An unexplained magnetic field will deviate the beam."

The major cause of extraneous magnetic fields in a building is ground current in the building's steel structure. Eliminating steel is not a viable alternative in most structures. To eliminate stray fields in Duffield Hall, concrete columns and floors have epoxy-coated rebar tied together with plastic wire. Prior to pouring any concrete, each section of rebar was tested with an ohm meter to make sure there was no direct steel-to-steel contact.

"You also need to be very careful in how the building is wired," says Pollock. "It's not unusual for a ground wire and neutral wire to be connected near the load. That can lead to ground currents. So, carefully monitor the building's wiring to make sure the ground and neutral are kept isolated."

Building Management Plan

Duffield Hall has a management plan, something no other building on the Cornell campus has at present. The plan details all the steps necessary for chemical safety in the building.

"Our emphasis is on safety training and safety management," says Pollock. "We will institute a programmed safety training course for all occupants, with yearly update training, primarily focused on chemicals handling and disposal."

All new experiments will be required to submit a pre-operational safety and environmental review documenting the hazards and operation of the experiment, and describing the worst-case effluent from the experiment. If the exhaust is toxic or dangerous, the amount of exhaust and method of exhaust will be studied to ensure that it doesn't create an environmental problem outside the facility.

"If necessary, point-of-use devices like scrubbers or incinerators will be installed prior to starting the experiment to maintain environmental safety. Duffield Hall will have several hundred toxic gas sensors distributed throughout the facility to safeguard against leaks," says Pollock.

Use of highly toxic or dangerous chemicals is restricted to a handful of specially-trained people. Access to chemical storage will be restricted and controlled by card-reader locks. Chemical transport in the building occurs only in special utility corridors and tunnels, never in public space. Floor-to-floor delivery of chemicals will occur on a chemical dumbwaiter. Card-access is required for all labs.

"If someone is found to be violating the safety rules, access can be disabled immediately," says Pollock.

Other Considerations

Nanotechnology is still very much in a state of flux. Pollock points out that tools and chemicals anticipated during the programming and design phase will likely change dramatically before the project is occupied. This can raise fit-out and design costs significantly.

He also cautions that criteria developed for the semiconductor industry may not be optimal for non-semiconductor work.

"It's very tempting to say, 'Let's see what they're doing in Silicon Valley.' But cleanrooms for Silicon Valley are really not a good starting point for research labs in a biomedical or university setting. They have completely different functions and the cleanliness requirements are very different," says Pollock.

Finally, be prepared to temper user demands for the extreme in cleanliness. In reality, they may not be necessary.

"We discovered that our existing cleanroom had been running at worse than Class 10,000 for 20 years," says Pollock. "Everyone thought it was Class 100. In 20 years, no one had ever complained about dirty cleanrooms and nothing had ever been attributed to dirt."

By Lee Ingalls

We welcome your Questions and Comments

Copyright 2008 Tradeline Inc.
All Rights Reserved
ISSN: 1096-4894