The Next Big Wave

It is a fairly safe bet that nanotechnology will be one of the next big industries in terms of technical and commercial development. The potential benefits of nanotech have been well described both in the technical literature as well as in general circulation magazines and newspapers. Some of these include:

*

A possible merger involving physics, chemistry and biology (on both the cellular and molecular levels).

*Possible new micro-machines that will alter the nature of our production processes by manipulating matter on the molecular level.

*Potential new classes of pharmaceuticals with the promise to perform analysis and repair of body parts at the cellular level (micro-surgery), thereby revolutionizing the practice of medicine.

*Significant advances in computing speeds and memory capacities.

In an effort to bring these into existence, vast sums worth billions of dollars are being spent to force feed this new technology. Of course, the potential returns that result from a successful R&D effort would be worth orders of magnitude greater than the required investment. And so every major university, science center and laboratory seems to be engaged in nanotechnology research at some level.

Before getting too far ahead of our current capabilities, it might be a good idea to review a few of the underlying facts. We are at the very beginning of the nanotech roadmap. What we donít know far outweighs what we do know. From this intuitively obvious statement, a few generalizations follow:

*The processes and recipes for creating nanotechnology structures are generally unknown and remain to be worked out.

*The tool sets that will be used to create these devices are in their infancy.

*Similarly the materials that will be used are generally undefined and some new materials with currently unknown properties will be utilized.

*The process utilities required to support these production processes are similarly vague and will remain in flux for the foreseeable future.

*Contamination control efforts for nanotech are just beginning. Cleanliness levels for air, process gases, and liquids are also unknown and will be determined by forecasting (extrapolating from known applications) followed by trial and error review.

*There are no clearly established winners or losers, yet. This refers to individuals, institutions and whole market sectors that have yet to develop.

What We Do Know

Given this level of ignorance, it would be helpful (and perhaps an ego-boost) for us to review some of the things we do know with at least some degree of confidence. Here is a likely list of some of those axiomatic starting points that we can probably agree upon to form a common frame of reference:

*Based on the chemistries and the geometries, we know the products and processes will be extremely contamination sensitive.

*Contaminants will fall into all categories: particulate, molecular, biological, and other (vibration, lighting levels and frequencies, ESD, EMI/RFI, etc.).

*The tool sets used for research and development will change rapidly as processes are defined.

*There will be a wide variety of gases and liquids with an equally wide variety of properties that will be required for fabrication of nanotechnology devices.

*For the foreseeable future, the required utility set will change as rapidly as the tool set.

*Flexibility will be required in the laboratory facilities developing these processes. Tools and processes will evolve quickly. Layouts will have to adapt to these changes.

With this background, a few general statements can be asserted. But even these are immediately followed by more questions.

*We will need flexibility, flexibility, and more flexibility. This is akin to semiconductor production, but even more so since we know so little about the process. Perhaps a better analogy would be the state of semiconductor facilities and processing in the 1960s and 1970s. But we have learned the hard way that this flexibility is expensive. Can we afford it? And how much can we afford? Can we afford to be without it?

*Most nanotech facilities in the next few years will be research oriented. But the need for financial returns will lead to some limited development and pilot production efforts as soon as feasible. What impact should we expect of these ìdual-useî requirements?

*At university research labs, we can reasonably expect a wide variety of facility users: undergraduate and graduate students, post-docs, faculty, technicians and operators. Will this have much of an impact on the facility layout and design? Can research and teaching peacefully coexist in these laboratories?

*Showcase or hidden gem? Remember when computers were new and we put them in fishbowls where everyone could see and be suitably impressed? Early semiconductor fabrication facilities followed the same pattern. Then we learned about risk and security. What lessons should we employ for nanotech labs? What about physical security driven by safety factors? Or information security?

*Recall the numerous gases and liquids to be used in the development and production process mentioned above. Some are extremely toxic or corrosive or pyrophoric. In a university environment, how do these characteristics affect our decisions for nanotechnology facilities location and layout?

*Universities often tend to be collegial places with few hierarchical structures (other than the professor-graduate student one). How can this ìopennessî co-exist with the needs for informational security? Or the safety requirements driven by the presence of numerous highly hazardous substances? (Probable answer: it canít.)

*Over the past 35 years, cleanrooms for semiconductor applications have bounced back and forth between ballroom and bay-and-chase approaches. Eventually, it became clear that neither was optimal under all conditions and that most projects called for some combination of the two. And then super-clean minienvironments located in a much less clean ballroom began to gain wide favor. How will this evolution affect nanotechnology facilities development?

 Recent Developments

With so many colleges and universities getting involved in nanotechnology research at various levels, many organizations are trying to obtain funding for their research environments. What is becoming clear is that some institutions are taking a minimalist approach while other, more affluent or perhaps industry-sponsored institutions are investing far more heavily in their facilities. Variations of two orders of magnitude are not uncommon between the haves and the have-nots. Figures ranging from just over one million dollars up to $100,000,000 have been reported. While we must take care to assure that we are comparing apples to apples, we should probably expect to see these wide ranges for the foreseeable future.

One key decision point is the cleanliness level specified for our research lab space. ìNanotechnologyî is commonly defined as working with line sizes of less than 100 nanometers (0.10 micrometers). Despite this commonly used definition, there is wide variation in the cleanliness levels that are being specified. Research cleanliness levels of from ISO class 7 (FS 209E class 10,000) down to ISO class 4 (FS 209E class 10) have been called out. In some rare cases specific areas are defined at even higher levels of cleanliness. And to date, few good reasons for this wide range have been offered. Similar disparities are being seen in purity levels being called out for process gas and liquid distribution systems. Technology decision makers should probably review these cleanliness and purity specifications because they are critical for defining the facilityís capabilities. Equally important, such cleanliness and purity increases can also exert a major cost impact on project financial requirements.

At the end of the day, whatever is spent must provide a vehicle for a successful research project. In addition, there must be enough forethought to allow for some generational adjustments without having to re-design and rebuild the space every time this rapidly evolving technology jumps forward. This is somewhat troubling because it appears that this issue may have escaped scrutiny by the institutional decision makers.

In other words, the amount of money available should not be the vehicle for a decision to use one level of technology over another. We have seen cleanliness specifications that are typical of those used in the 1970s and 1980s. In some cases this may be appropriate if the researchers are planning to use glove boxes and minienvironments (small ultra-clean enclosures surrounding the more sensitive tools or processes). But if this is the case, there seems to be little in-depth understanding of the integration requirements that this hardware forces on the facility, the materials handling systems, the input/output access to the tools and the layout of the facility. Perhaps more fundamentally, the key decision makers need to decide (quantitatively) what cleanliness and purity levels they require. As previously stated, a research facility that does not allow for the research projects to be accomplished, is not worth building. Too many times these choices have lead to cost overruns to fix what could not be used. The delays in determining who was at fault, which party pays for the fix as well as the reconstruction time, often cause inordinate delays in performing to the original research intent of the laboratory.

Another major issue concerns the use of existing facilities to provide an envelope for the new laboratory. With issues like EMI (Electro-Mechanical Interference) and vibration that can impact accuracy of highly sensitive metrology devices, there is little forethought as to whether the existing systems can provide for a workable lab. It should be noted that some existing spaces might never be suitable for nanotechnology research. For this reason, the same basic process used to predetermine the viability of a ìGreenfieldî or new site for a new laboratory should be used to qualify an existing site. (Note: Other potential environmental considerations will be discussed below.)

The Fundamentals

Some elements of the facility and its associated systems are fundamental to project success. For that reason alone, the following points should be considered:

*ISO 14644-1 and 2 describe the airborne cleanliness levels in use today as well as procedures to be used for cleanroom certification and monitoring. The cleanest classification defined is ISO class 1 which implies a maximum of 10 particles, 0.1 micrometers or larger per cubic meter of air. Knowing that this is the cleanest available, it is tempting to specify this cleanliness level for our laboratory, or at least its most critical areas. Big mistake. This cleanliness level may be appropriate for minienvironments and process chambers but it is probably not suitable for large rooms with people in them. First, there is serious doubt whether it can be sustained in rooms in the operational state. Second, sampling and testing requirements dictated by the ISO procedures would make Class 1 certification process extremely long, difficult and expensive. Also, such stringent particle control would likely make the cleanroom prohibitively expensive. Lastly, such control is seldom needed, except in small critical and localized places. Under these conditions, isolation technology-including glove boxes, minienvironments and pods-usually provides better protection at lower cost. Rather than specifying a cleanliness level for all areas of the lab, it is far better to decide what product protection is required in which places and then decide how to best provide that protection.

*High purity process support systems such as ultra-pure water, gas, and liquid distribution systems require equally careful scrutiny. Typical state-of-the-art purity levels for these substances are often in the low part-per-billion (ppb) to part-per-trillion (ppt) range. Of course, there is a major cost difference between common industrial grades and these levels. Similarly, the cost of the distribution systems themselves also increases significantly with purity levels. Although expensive, these are often driven solely by the needs of the process. As with airborne cleanliness, it is important to carefully consider the process requirements and not over-specify purity levels. At the same time, key decision makers must carefully consider how chemicals will be distributed. For gases, the solution almost always involves a central dispensing area with high purity pipe or tubing to carry the gas to the use-point. With liquids, similar approaches can be used or individual bottles can be delivered to the tool and then poured into the toolís reservoir. Although bulk dispense systems are more expensive than bottle delivery, they provide much better consistency in terms of purity as well as improved safety. As with most things in life, one size does not fit all and one approach is not likely to suit all projects.

 Space Considerations

As mentioned above, many of the laboratories being designed and built today are going into existing buildings. Although this is a common practice going back generations, it is not always appropriate for cleanroom facilities. Because of airflow and tool characteristics, certain room geometries must be observed. Some examples:

*ISO class 5 (and cleaner) labs that are wider than around 15 feet will probably require a raised (flow-through) floor for uniform airflow. This will add a minimum of 12 inches to the facilityís height requirements. It may be possible to achieve ISO class 4 without a raised floor if the width is held to around 10 feet. Rooms that are much wider than that (or cleaner) generally require raised floors.

*In terms of the effect on facility height, routing of utilities must also be given careful consideration. A common approach is to run them under the raised floor and up into the bottom or the back of the tools they are feeding. Some utilities such as exhaust ducts are quite large and can block the under-floor return airflow. To prevent this, the raised floor height must be increased, further affecting the overall facility height.

*The cleanroom must accommodate the tool with the greatest overall height, plus the material handling equipment used to move the tool into place. Lights and sprinklers can also get in the way and must be included in the planning process. Also recall that flat panels on the top of tools cannot be placed immediately against the ceiling filters without seriously obstructing their airflow.

*Although it may appear to be ìwasted spaceî from a process perspective, valuable area must be dedicated to return air plenums.

When all these factors are taken into account, it becomes clear that laboratory location decisions must be very carefully thought out. The decision to place a lab within an existing building because there is area available may turn out to be sub-optimal.

Code Compliance

Code compliance is a serious issue but often not readily visible on the radar screen of universities. Experience in building classroom buildings, libraries, and dormitories does not prepare them for the multitude of different chemical substances that will be used in nanotech labs. Even chemistry labs with their numerous acids, bases, solvents, and exhaust systems are significantly less complex than typical semiconductor labs.

Most labs are considered ìBî occupancies. This means (in part) that the total quantity of chemicals and gases is relatively small and below certain set limits. Quantities above these limits will often drive the facility to an ìHî occupancy that requires much more expensive design and construction features.

There are two factors to remember with respect to code compliance: First, the subject is driven by life safety considerations. Due to the potentially hazardous materials and processes within these labs, life safety issues should always be at the top of everyoneís priority list. Second, local code officials drive code compliance. Although codes themselves may be universal, they are interpreted and enforced by local officials. It is therefore not surprising to see somewhat different approaches to code compliance in different jurisdictions. An experienced design team can make all the difference when working with these local code officials.

More Things to Consider

After life safety, versatility, and flexibility are probably the most important goals for nanotech lab facilities. Not many institutions have all the funding necessary for all of the process equipment that may eventually be required. To permit this future growth and likely changing research directions, the facility planning function must plan for this flexibility in two ways. First, the comprehensive layout must contain enough floor space to contain the future hardware additions. Second, the non-process support systems must be sized and located to handle the future requirements including utilities such as steam, chilled water, electrical power, scrubber capacity, and exhaust fan capacity. With proper precautions, these systems can be ìhot-tappedî to allow future tie-ins if the proper valve, dampers, tees, and associated fittings were not included in the original installation. Not so with the process utilities such as high purity bulk and specialty gases. Future tie-in valves must be left in place during the initial construction phase. Cutting into high purity gas lines results in severely contaminated systems and is not acceptable procedure.

Almost as important for future expansion and tie-ins is the topic of documentation. During the initial build-out of the facility, numerous tie-in points will be left blanked off or valved off for these future services. It is critical that these connection points be tagged and identified on drawings to avoid ìlosingî them in the interim.

Note that the facility may be going into an existing building and tying into existing building systems such as chillers, boilers, pumps, compressors, etc. In this instance, it is critical to know that there is adequate capacity in the existing systems before depending on them. As most of our older laboratory complexes are not always well defined with a complete set of ìas builtî drawings, there needs to be some care to investigate current usage and existing tie-in issues to prevent later conflicts. At a minimum, these lead to cost escalation and schedule slips. Here again, experienced programming and design teams can reduce the number and severity of these hard to foresee problems.

Other Contaminants

Up to this point, we have only considered primary contaminants such as airborne particles and process systems impurities. However, cleanrooms are often called upon to control many other forms of possible contamination. Some of these include temperature and humidity levels in cleanroom air, vibration and noise levels, ESD (electro-static discharge) levels, RFI (radio frequency interference), AMC (airborne molecular contamination) and lighting levels (both frequency and intensity). The problem here is that the nanotechnology products and processes are not well defined and it may be extremely difficult to determine what levels of what environmental parameters will constitute real contaminants that must be controlled and which can be ignored. What if we guess wrong? In this instance, future surprises are likely to be handled inside the tool or with minienvironments.

Conclusion

This paper was not intended as a how-to attempt to specify the design for nanotechnology research labs. Rather, it was our intention to make sure that when the construction ìdustî settles, you have a flexible usable space. It is hoped that the issues raised will be of use to the reader, primarily in the facilities planning phase of their project.