Building Vibration Can Contaminate Clean Factories, Cleanrooms And Clean Activities
by Wayne Tustin and Alvin Lieberman

Introduction
In this article we focus on a serious problem caused by building vibration: that is the subtle problem of particulate contamination of clean factories, cleanrooms, cleanroom products, and cleanroom processes. When someone walks past or rolls a heavy object past your cleanroom, do you feel vibration? Do you feel the vibration when those activities occur on an overhead level? If so, then product yield and process tool performance are in danger of being degraded. Remedial measures should be taken as soon as possible to reduce the vibration and its effects.

A potentially very expensive example: building vibration can degrade photolithography operations, especially serious in microcircuit production. A recent paper¹ presented at the Institute of Environmental Sciences and Technology 47th annual technical meeting discusses that problem from the viewpoint of possible problems that may affect the newest generation of photolithography tools. That paper discusses current lithography scanner support criteria in terms of receptance (displacement/force) spectra, and compares them with receptance measurements carried out in several fabs (microcircuit fabrication facilities). Design philosophies are discussed for both floor structures and tool support pedestals.

Natural Frequencies? Resonances?
What are these terms? Did you ever push a child on a swing? You timed your forcing frequency to match the swing's natural frequency and thus used resonance to give your child a much higher ride than would have been possible (applying the same dynamic force) at any other frequency.

Responses To Vibration
We normally give little thought to structural natural frequencies at which a building or other structure may respond strongly with relatively large motions to a seemingly-small input force. In 1940, the Tacoma Narrows bridge was destroyed by wind-induced motion. More recently, Bostonians are became? aware of twisting of the John Hancock Building caused by certain wind conditions. Nearing completion, severe twisting of the structure broke nearly all of the windows in that office building, leaving it uninhabitable for more than a year. It took that long to design and install dynamic vibration absorbers on an upper floor. These reduced twisting to about 25% the earlier amount. Then windows could be replaced and the building could be tested to ensure that the problems were indeed solved correctly.

Your present-day cleanrooms are more apt to be adversely affected by higher frequency excitation in which the frequency of some stimulus (external or internal) matches one or more of the natural frequencies of your facility and its support structure.

Vibration Stimuli
External building stimuli include high winds and wind gusts, aircraft flyovers, passage of railroad trains on nearby tracks, large motor trucks on nearby roads, and of course, "seismic events" or earthquakes and volcanic effects. Fortunately the latter are few and far between. Internal building stimuli include building machinery such as cooling fans, refrigerators, air conditioning systems, hoists, elevators, production machinery such as conveyor systems, nearby office machines such as printers and copiers (never allow these to be installed in a cleanroom), internal vehicles such as fork-lift trucks and, finally, pedestrian footfalls. All of those devices can create building vibration as well as generating and releasing some particles during their operation.

Contamination Mechanism 1 - Agglomeration
Aside from the problems due to movement of lithography components, building vibrations can also cause airborne particulate contamination that is difficult to detect with normal cleanroom particle measurement. Even low-level vibration can cause release into the air of particles that have settled upon surfaces. These particles can be in the nanometer and larger size ranges and in small quantities. If the settled particles are retained on a surface only by the effects of gravity, then vibration can cause the original particles to agglomerate, forming particles that can be much larger than the particles originally deposited on the surface. Even so, many of the discrete particle counters used to classify cleanrooms in the low concentration levels of the present US Federal Standard 209 are close to the limits of their particle detection sensitivity levels. Statistically valid data for the ISO 14644-1 cleanroom standard² are difficult to procure for ISO level 1. The particles that must be sized and counted for both documents at their lowest classification size level are defined as being equal to and larger than 0.1 µm in diameter. The US Federal Standard 209 specifies that the cleanest classification level shall have no more than 350 particles per cubic meter or 9.9 particles per cubic foot of air equal to and larger than 0.1 µm in diameter. ISO 14644-1 specifies that the cleanest classification level shall have no more than 10 particles equal to and larger than 0.1 µm in diameter per cubic meter of air. Statistical requirements result in the need to count at least 20 particles per measurement in order to provide an acceptable confidence level to in? reported data. These requirements result in excessive instrument operating time for to acquire? acceptable data.

Unfortunately, many state-of-the-art high technology devices are extremely sensitive to small particles that are generated at locations that may be "upwind" of the device processing tool. The quantity of small particles may be very low with regard to the air volume near that tool, yet sufficient to reduce yield to an unacceptably low level. The small particles that are generated may deposit on a surface that vibrates. Vibration results in agglomeration of the small particles to produce agglomerate particles that can be as large as ten to one hundred times the size of the initial small particles. This phenomenon was noted in a recent paper³ that examined the behavior of fine particles on a surface under ultrasonic vibration.

Contamination Mechanism 2 - Relative Motion Between Walls, Floor And Ceiling
How was your cleanroom constructed? Quite possibly plasterboard panels were screwed or nailed or stapled to a wood or metal framework. Joints between plasterboard panels were taped and the room was painted. One of the unfortunate results of building vibration, however caused, is relative motion along the edges of those panels, releasing paper and plaster particles into the air.

Contamination Mechanism 3 - Relative Motion Between Clean Room And Equipment
What equipment? Let us commence with air conditioners on the roof and work our way downward between rooms, noting ductwork for heating and cooling, a furnace, blowers, pumps. Inside the various cleanrooms are wall-mounted and floor supported cabinets, wall-mounted and ceiling-mounted light fixtures.

Another unfortunate result of building vibration, however caused, is relative motion between equipments, releasing particles into the air.

Contamination Mechanism 4 - Loss Of Particulate Uniformity
In the pharmaceutical industry, many medical processes involve blending of particulate materials, each with specific properties needed for a particular product to satisfy a medical need. Quite often, it is necessary to blend particulate materials in different size ranges to produce a product with the desired properties. The mixing process is designed to blend these materials so that the batch of powder contains the correct quantities of each of the initial particulate materials. Vibration can cause smaller particles to settle downward in into? spaces between the larger particles. Al - is this agglomeration? What had been uniformly blended powder is no longer uniform and the batch may require additional mixing with a possible decrease in uniformity. Maybe delete final 3 sentences that don't seem to relate to vibration.

A Specific Stimulus, An Earthquake
Earthquakes can shake particles loose. Consider the lower graph of Figure 1. A momentary acceleration peak of earthquake motion, measured at your building's foundation, might approach 0.25 g (one "g" is 32.2 ft/sec2 or 9.8 m/s2). The "random" vibration in the lower figure is a summation of vibrations that exist over a wide range of frequencies, perhaps 1 to 50 Hz. A few of these forcing frequencies will probably match up with structural natural frequencies, and a poured-concrete floor slab may resonate and behave like the central graph of Figure 1, exceeding 0.5 g. Think of your cleanroom as a piece of equipment resting on your floor slab. Imagine its natural frequencies matching up with the floor slab resonances. Vibration will be further magnified to several G, as in the upper graph. There is now sufficient acceleration to shake particles off walls, ceiling, and wall structures, as well as from horizontal work surfaces and from components mounted on tool surfaces onto product surfaces. As indicated in the discussion on "MECHANISMS", the effects of such resonances may also create new and larger particles by one or more of the mechanisms mentioned above.

Immediately after an earthquake or other major stimulus, check for particles that may have been generated and/or moved into the cleanroom. Inspect the cleanroom area and the tools operating in and about that area. For large tools, inspect product component entrance areas in particular, but do not ignore any of the tool interiors where in-process products may be exposed to "fallout" from particles generated by tool motion during the earthquake. Inspect all cleanroom filters, especially those in ceiling ducts. During that inspection, examine the filter seals to ensure that the earthquake has not caused filter movement sufficient to cause an opening in the seal area. An opening can permit significant quantities of contaminated air to bypass the filter medium and act as a local area contamination source. During the cleanroom inspection, make sure that small local contaminant sources have been produced as an effect of the movement caused by the motion of the tool or storage where in-process product components are stored.

Remember that many of the less-dramatic vibration sources listed under "STIMULI" may intermittently or continuously shake your cleanroom structure. At certain of the structure's natural frequencies, the resulting magnified vibration can shake particles loose into cleanroom air.

Cleanup Procedures; Effect On Production
Depending on the severity of vibration that has previously occurred, some areas in the cleanroom may no longer be adequately clean to meet either the process cleanliness requirements or the regulatory agency requirements as the pharmaceutical industry is always required to do. If a requirement is present that specifies a cleanroom classification level that must always be maintained, then procedures are written out in detail in the ISO 14644-1 and ISO 14644-2 documents. ISO 14644-1 specifies procedures to verify the cleanroom classification level. These procedures involve verifying the integrity of the final filter system, ensuring that the air flow rate is acceptable and that the integrity of the cleanroom enclosure is verified by leak testing.

For a large ballroom type cleanroom, this operation may require that production be stopped for as long as one or two weeks. Profitability of a 300 mm wafer fab can be severely affected if one or two weeks of production is not carried on. ISO 14644-2 specifies procedures for continuous and/or frequent monitoring in the cleanroom and for ensuring compliance with ISO 14644-1 If continuous or frequent monitoring is carried out, then reverification of the cleanroom class can be extended from a period of 6 months to a period of 24 months. In a two year time, availability of three "extra" work weeks production is very important to the profit of the facility.

Consider monitoring your entire cleanroom for atypically high particle concentrations. If production rates must remain high, monitor several locations in your processing area, especially where anomalous airflow patterns may have been recorded in the past. If you find excessive contamination in those areas, trace the particle "plume" to its generation source and take the necessary actions to remedy the problem. As cleanroom manager, you might also consider setting up a monitoring program for particle count and for air flow patterns at critical locations. Procedures for such activities can be seen in the Institute of Environmental Science and Technology's Recommended Practices. IEST-RP 006.2, "Testing Cleanrooms", is recommended as a guide for characterizing the cleanroom. IEST-RP-CC018.2, "Cleanroom Housekeeping-Operating and Monitoring Procedures" recommends procedures for monitoring cleanrooms. IEST can also supply the ISO cleanroom documents referenced herein.

Cocoons
When cleanrooms were smaller, their natural frequencies were higher and the movements were smaller. At this time some cleanrooms are ballroom-sized, with relatively low natural frequencies and relatively large motions, leading to an increase in particle contamination. One solution is a "cocoon" inside the large not-so-clean room. Not only does the cocoon have relatively high natural frequencies, but it can "float" on soft springs or air cushions. There are a number of companies that produce such devices. These can be used to protect the product from effluents generated by the environment or by the operating personnel. This is particularly true in areas where viable bacterial contamination is a potential problem.

References

• ¹ Amick, H., & Bayat, A., "Meeting the Vibration Challenges of Next-Generation Photolithography Tools" Proc. 47th Annual Technical Meeting and Exposition of the Institute of Environmental Sciences and Technology, Phoenix, AZ, April 22-25, (2001)
• ² ISO14644-1 "Classification of Airborne Particulate Cleanliness for Clean Rooms and Clean Air Devices"
• ³ Matsusaka, S., Nakamura, S. and Masuda, H. "Behavior of Fine Particles on a Plate Under Ultrasonic Vibration" KONA, 18, 213-219, (2000)
• ISO14644-2 "Test Schedule to Demonstrate Compliance with ISO14644-1"
• ISO14644-3 "Metrology and Test Methods"