To fully understand the scope of such projects, it's essential to consider both the general building components and the mission of the research that will be conducted inside the facility, says Curt Finfrock, the Chicago-based vice president and director of architecture for M+W Zander.

Nanotechnology research facilities are divided into three parts: a clean area, which can be either a major or minor portion of the structure; the utility systems necessary to support the clean area; and the rest of the building—everything from conventional labs and offices to conference spaces and public amenities.

The infrastructure is both the least visible and the most underestimated of the three components, contributing significantly to the cost and planning effort, according to Finfrock.

"Nanotechnology cleanroom costs range widely depending on infrastructure and utility support concepts selected," he says. "To avoid surprises, cleanroom costs should be evaluated with respect to the entire facility budget."

Nano Needs

Scientists engaged in nanotechnology research deal with particles and features that are smaller than 100 nanometers in size. ("Nano" means one-billionth.) Materials at this miniscule level display characteristics that can be quite different from the properties they exhibit on a larger, more normal scale. In this minuteness lies great promise for breakthrough and innovation.

Nanoscale investigation can involve many scientific disciplines and entails a wide variety of equipment needs. According to Finfrock, while the facility itself does not represent new territory, the convergence of so many fields in one setting does pose novel design (and budgetary) challenges.

"From a technical standpoint, the facilities required have been built before, at least in part," he explains. "What's new is the combination of situations or research interests in one location. Biotech, pharmaceutical, or semiconductor experimentation is typically isolated, but in nanotechnology research they all occur in the same building.

"Each field has different and demanding space requirements," he continues, "and when you combine them, it increases the complexity of the entire facility."

Research Drives Process Drives Equipment

When it comes to design, the nanotech cleanroom can take several different forms, depending on key variables like the basic cleanroom type, the number of floor levels, and the air management strategy. But even before these can be determined, in fact before any planning process begins, a clear idea of the research mission needs to emerge.

"A nanotech cleanroom must be designed from the bottom up or from the inside out, starting with the research its occupants will conduct—not only today and next year, but what they expect to be doing 10 years down the road," says Don Yeaman, M+W Zander director of industrial engineering. "The nature of the research they want to do drives which processes they want to perform, and the processes then drive the equipment set, which in turn determines everything else: the clean room environment, size, utility distribution system, and even the quantity and quality of the facilities."

More than a dozen different kinds of machines in various quantities can populate the clean laboratory, from massive wet processing systems 20 feet long by 12 feet high to multi-chamber deposition and etch tools. The quantity and ratio of equipment can vary significantly from one facility to another. As an example, Finfrock cites two nanotech research centers currently in the design stage at Argonne and Oak Ridge national laboratories.

"The Argonne facility is about 30 percent larger than Oak Ridge in square footage, and it has two and a half times the number of wet processing tools, include plating baths, sinks, and the like," he says.

Size and Cost Estimates

With the tool set requirements established, the planning team should then have a sound basis for selecting the most appropriate cleanroom concept and sizing utility needs. Like the overall building, the clean environment breaks down into three components: the clean laboratory, in its designated classification; dedicated space for the ventilation/fan systems; and the clean utility systems. Included in this last component are services such as ultra-pure water, acid waste neutralization, chemical storage, chemical waste, clean power, etc.

Finfrock offers some rules of thumb for estimating the square footage to allocate to each area. For the lab, figure on anywhere from 120 sf to 300 sf per tool, he advises. Typical space requirements on the air side are between 0.5 sf and 1.5 sf per square foot of clean space, while the ratio for utility space runs between .25 sf and 1.0 sf for each square foot of clean space.

"It's generally acknowledged that these facilities can be inefficient in terms of net usable area when compared to conventional laboratory facilities," he remarks.

Estimating construction costs is a different matter. Unlike more traditional structures, for which architects can quickly calculate a ballpark figure (seats in a classroom times a standard sf per unit times an average sf cost), on-the-spot cost projections for a clean environment are notoriously inexact.

Finfrock suggests the following rough guidelines:

• $600 to $1,200 per sf for clean lab, including the supporting ventilation and utility systems,
• $100 to $150 per sf for ventilation space,
• $150 to $250 per sf for utility space only.

But he also discourages planners from using this approach.

"I'd add a 40 percent contingency to the total because it's a poor way to estimate clean facility costs," he says, adding, "In the semiconductor industry, the cleanroom cost is quoted in many ways and can go as high as $2,000 per sf or more. We found it is simpler to think about the real cost: the clean space cost plus utility systems divided by cleanroom area. We have to educate clients about this thinking because they have heard so much contradictory information."

The best plan of attack, he recommends, is to invest in a pre-design study, a preliminary analysis of the user needs and potential cleanroom concepts. Even then, the resulting estimates have roughly a 25 percent margin of error.

Layout and Ventilation Options

Nanotech cleanroom layout is based on one of two fundamental types: the ballroom, a clear span of universal space offering maximum flexibility for equipment placement; and the bay/chase configuration, featuring an interlocking relationship between the clean space (bay) and the utilities area (chase), with a return air path through them. Occasionally, a hybrid scheme, combining elements of both concepts, will appear.

Bay/chase arrangements follow any number of patterns, from the conventional center aisle to alternatives with irregular bays and chases and asymmetrical aisles. Because each room represents an enclosed environment, the cleanliness level can vary from bay to bay. The ballroom scheme, in contrast, is typically engineered to only one classification. The one limit to the bay/chase layout is that generally, depending on the return air path, a bay cannot be more than about 14 feet wide.

"As a rule, the bay-and-chase design provides a more cost-effective opportunity for higher cleanliness classifications, plus it offers more flexibility in varying cleanliness classes within the facility," says Finfrock. "The main drivers of choice here are the cleanliness class required, the type of tools to be installed, and the project budget."

Air management strategies, from the type of fan unit to the supply and return air paths, will also influence the cleanroom layout choice, or vice versa. The return air path can travel through sidewall returns or via a variety of raised floor schemes, including a sub-floor level where process equipment can be installed. In geographic regions where the climate is moderate, some air handling equipment can be located outdoors, reducing the need for more expensive interior space.

"Typically we would evaluate the options for what's appropriate and then compare initial and long-term costs, the functionality of the clean room, ease of maintenance, flexibility, and so forth," says Finfrock.

Process and Operations

The types of processes to be run will significantly influence facility layout. For example, asking users what kind of substrate materials they want to process is critical. If both gallium arsenide and silicon will be in use at the same facility, the clean area must keep the substances separated to avoid cross-contamination and the potential for poisoning.

Operational practices must also be taken into account. It's important to know whether outside user groups will need to be accommodated and what kind of security might be necessary between certain portions of the cleanroom before developing a detailed layout.

Logistics and hazardous occupancies are other significant issues. Moving bulky and expensive machinery in and out of the cleanroom requires up-front planning, and chemical moving paths need to be considered early.

"The code for H-5 and H-6 categories mandates that the chemical moving path cannot cross one of the exit paths," advises Yeaman. "Planners must devise a way to get chemicals in while allowing people to exit without crossing paths."

Electromagnetic vibration and acoustic- (EVA) sensitive equipment is also a major driver of the layout, he points out. Sensitive equipment should be located as far away from EVA-generating machinery as possible.

Utilities Duty Factor

Utilities comprise a significant portion of the overall clean facility cost. For example, power and associated cooling water requirements are generally much higher than in a typical wet or dry lab. At the same time, users tend to overestimate their needs, adding together manufacturers' guidelines for individual equipment to reach an exaggerated total.

Many utilities are only used part-time for a given tool, especially in the research environment, Yeaman points out. Even in high-volume manufacturing, which demands maximum product yield, utilization rates rarely top 70 percent. In contrast, M+W Zander interviews with user groups in the R&D setting put process equipment usage at about 20 percent. However, some process utilities systems don't scale, meaning that sinks or exhaust machinery inside the cleanroom don't turn off when equipment is not in use.

To avoid going overboard on capacity, Yeaman recommends developing an equipment utility matrix that instead of relying on nameplate values incorporates a duty factor and realistic utilization rate for each piece of machinery.

By Nicole Zaro Stahl

We welcome your Questions and Comments

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