Cleanroom Design: by Douglas H Erickson
Cleanroom Design: by Douglas H Erickson
Cleanroom Design: by Douglas H Erickson
by
Douglas H Erickson
Bachelor of Architecture
Bachelor of Science in Architectural Studies
Washington State University
Pullman, Washington
1975
June, 1987
Signature of Author
Douglas H Erickson
Department of Architecture
May 8, 1987
Certified by
Accepted by
t bJulian Beinart
Chairman
MAzSSAOsmTS INSr~iDpartment Committee for Graduate Students
F TECHNOLOGY
tlr 1
UmgAES
Room 14-0551
77 Massachusetts Avenue
Cambridge, MA 02139
M ITLib
Document Services
Ph: 617.253.2800
Email: docs@mit.edu
http://Ilibraries.mitedu/docs
DISCLAIMER OF QUALITY
Due to the condition of the original material, there are unavoidable
flaws in this reproduction, We have made every effort possible to
provide you with the best copy available. If you are dissatisfied with
this product and find it unusable, please contact Document Services as
soon as possible.
Thank you.
Pages 96 - 131
CLEANROOM DESIGN
by
Douglas H Erickson
ABSTRACT
The development of the integrated circuit which replaced the vacuum tube, started the
size reduction process for computer components. These integrated circuits are made
from silicon (chips) and are comprised of electronic switches, or gates. The gates are
measured in size of microns. The diameter of a human hair is approximately 60
microns across.
Facilities that develop, and manufacture these integrated circuits require the strictest
quidelines for environmental controls and prevention of potential health hazzards that
personnel may encounter while working in these facilities. The major environmental
controls are particle size and number, temperature, relative humidity, air flow velocity,
and pressure. Providing this and other forms of control are used to develop what are
called cleanrooms. Cleanrooms are used for the manufacture of a number of different
kinds of products. The focus of this research will be on the microelectronics industry.
This industry leads all other industries in developing systems, standards, and
monitoring technologies, to control microcontamination which is the essence of what a
cleanroom does.
This thesis will be divided into two parts. The first part defines what a cleanroom is
and what it is comprised of. Next, there will be methods presented to design this type
of space in a more energy and cost efficient manner.
The second part involves the research in the vertical laminar flow aspect of operating a
cleanroom. The vertical laminar flow offers a stru ctured method for controlling air
flow and provides an effective means for discharging particulates out of the
cleanroom. By comparison, the conventional air flow system throws the particulates
in a random fashion. The vertical laminar flow has its limitations. By itself, the
vertical flow operates well, but people, and equipment cause turbulence which
disrupts its effectiveness. Working with these variables through research, an alternate
method of working with this vertical laminar flow was developed. The results,
recorded by photographs show an alternative for dealing with the turbulence and
eddys caused by the operations in the cleanroom. There will be a discussion followed
by a number of questions, and responses which will be the basis for this research on
vertical laminar flow.
2
ACKNOWLEDGEMENT
The development of this thesis could not have taken place without the input, and
knowledge I was able to receive from a number of people.
I would first like to thank Dean John de Monchaux, head of the School of
Architecture and Planning at the Massachusetts Institute of Technology for his initial
guidance, and granting me financial help from the Cabot Foundation.
My advisor, Dr. Leon Glicksman, who steered me on course, with his experience and
knowledge and gave me the advice I needed. Mr. Tom Yule, A doctorate student in
the Mechanical Engineering department with his technical experience and his good
nature.
Two firms whom deal with cleanrooms, MC2 with vice president Leo-PierreRoy, and
also the architecture-engineering firm of Symmes Maini and MckeeAssociates. Mr.
Michael Powers who is a principle there along with Mr. Leo-Pierre Roy from MC2,
graciously gave me much literature and their time for my research.
Mr. Tony Collozi from MIT who designed the new cleanrooms at MIT, and who is on
staff there gave me good contacts for my research.
My research model which played such a vital role in my vertical air flow research,
would not have been such a success without the input from Mr. Harry Hadley, who
was instrumental in the design, and Mr. Anthony Ingemi, who was instrumental in its
construction. There firm, B.G. Wickberg and Company, located in North Quincy
Massachusetts was kind enough to allow the research model to be built there.
The last part of this acknowledgement is for my wife Leslie, who I give my sincere
love and respect, and whom without her support, none of this could have taken place.
Hi Mark.
3
CLEANROOM DESIGN (Energy Efficient Design)
Abstract 2
Acknowledgement 3
Table Of Contents 4
PART 1
Chapter 1 What Is A Cleanroom
1.1 Introduction 6
1.2 Terminology 7
1.3 Factors and Fundamental Design Concepts That Are
Important In Cleanroom Design 10
Chapter 2 History
2.1 How Cleanrooms Originated 31
2.2 Future Trends In Cleanroom Design 35
4.1 Introduction 46
- 4.2 Methods For Energy Efficiency And Cost Savings 47
4
PART II VERTICAL LAMINAR FLOW (RESEARCH)
5.1 Introduction 65
5.2 Laminar Flow (Vertical) 67
5.3 Research Model (Apparatus)
5.31 Design 69
5.32 Model 70
5.4 Research
5.41 Research Agenda 72
References 147
5
1.1 Introduction
Recent growth in high technology industries, especially electronics, has increased the
demand for what are called cleanrooms. Simply stated, a cleanroom is a controlled
environment in which temperature, humidity and airborne particulates are all carefully
regulated the predetermined standards. Generally speaking, the cleanroom controls
three major types of contamination. The first is airborne contaminations. This type of
contamination can be of any shape, size, or nature and is transferred by either air
convection currents or trajectory off of other particles. In either case, the medium of
transfer is air. The second type of contamination is control of particulate matter
transferred from one object to another by direct contact. The third type of
contamination is the transfer of particulates by means of fluids. This can occur with
cleaning solution as well as with lubricants.
Cleanrooms are classified by the number of particles of certain sizes that are present in
a cubic foot of air. A class 1,000 cleanroom has, on average, no more than 1,000
particles of 0.5 microns or larger per cubic foot. And a class 10 cleanroom has just 10
or fewer particles of .10 microns in a cubic foot of air. (see figure 1) In contrast to
this, a normal office space will have millions of particles 0.5 microns or larger in a
cubic foot of air.
10.000- -
10 -
DIFFERE\NCE
01 02 05 to0 20 5.0 10 0
PARTcLESZ (m
figure 1
6
1.2 TERMINOLOGY
7
DOP: (Dioctyl Phthalate) Smoke test that determines the amount of smoke penetration
through a filter.
EDDY: Air running contrary to to the main current.
GATES: A circuit that has an electrical output dependent on its input. Smallest
subclass of current on a chip.
HIGH EFFICIENCY PARTICULATE AIR FILTER (HEPA): A filter with an
efficiency in excess of 99.97% for 0.3 micron particles as determined by what is called
a Dioctyle Phithalate (DOP) test.
HIGH EFFICIENCY PARTICULATE AIR FILTER (VLSI): A filter with an
efficiency in excess of 99.9995% for 0.12 micron particles.
INTERGRATED CIRCUIT (IC) : An electronic device that performs the functions of
many thousands of transistors.
LAMINAR FLOW: Air flow in which the entire body of air within a confined area
moves within a uniform vertical air pattern along parrallel flow lines.
LAMINAR FLOW ROOM: A cleanroom with a laminar flow requirement.
a
UNIFORM AIRFLOW: Unidirectional airflow pattern in which the point-to point
readings are within plus or minus 20 % of the average air flow velocity for the total
area of the laminar flow work zone.
VERY LARGE SCALE INTERGRATION (VLSI) : Microelectronics manufacturing
technology for intergrated circuits with line geometrics as small as 0.1-micron widths.
WAFER: A thin, flat circular disk of primarily silicon that is masked, oxidecoated,
doped and processed for separation into numerous electronic devices or for packaging
as an intergrated circuit.
9
1.3 FACTORS AND FUNDAMENTAL DESIGN CONCEPTS
THAT ARE IMPORTANT IN CLEANROOM DESIGN
Following the pattern development steps, the wafer is cut into discrete chips. Each
chip is then mounted on a suitable holder, and electrical connections are made for
subsequent use.
At any step, the wafer or chip may be exposed to the ambient environment, and
particulate contaminants can be either ingested or deposited on critical surfaces. The
prevention of these contaminants is curtailed by good fundamental design factors in
cleanroom design. There are many factors that play an important part in the design and
total expense of a cleanroom installation. Along with these factors is the recent
concern about the health risks involved with personnel in the cleanroom environment.
In the initial planning stage, two basic considerations should be answered first:
1. What will be the purpose of the room? 2. What degree of cleanliness will be
required? After this has been established, emphasis should next be placed on the
allocation of space; position and type of equipment to be installed within the room; the
number and placement of personnel and their internal traffic pattern.
10
Also, the compatibilty of processes, the equipment operating processes and
maintenance techniques to economically achieve the project manufacturing
requirements should be established.
After these factors have been layed out and the design is set, the fundamental criteria
for good design and control should be considered.
11
1. Airborne Particles
Airborne particles range in size from 0.01 microns to several hundred microns.
Conditions for clean space vary widely with industrial and research requirements.
A reference on the control of airborne particulates is the Federal Standard No. 209B.
(see appendix)
2. Particulate Control
At the same time, consideration must be given to the quanity of the particles of a given
size that might be present at a particular point with the area.
It has been proven in studies that there is a definite relationship between the size of a
particle and the time in which it may be airborne, ( 1. Austin, 1970.) so it is
beneficial to discuss airborne particles by quanity of a given size. Federal
Standard2O9B discusses this relationship see appendix to further analyze the level of
contamination to be considered.
12
This can be divided into external sources and internal sources.
External Sources: For any given space, there exists the external influence of
gross atmospheric contamination of air pollution which tends to find its way into all
areas of a working environment.
Skin is constantly flaking off and generating particles in the 1 micron range, and
exaled breath contains large quantities of particles ranging in size from sub-microns to
several hundred microns. (2. Cambridge Filters, 1986.)
Within any working environment, a certain dynamic situation exists in the air.
Movement results from people working, activity of service equipment within the
processing area, fans blowing, motors rotating, etc.
13
All of these motions impart kinetic energy to the air and cause it to move at random
velocities within the space.
Fine particles are caught up in a random current within a room and are easily moved
from one location in the room, to another. This transfer of contamination through
random air currents from one part of the room to another is known as
cross-contamination and is a significant contribution to the contamination level at the
worksite. A resulting contamination buildup occurs within a space and reaches a
plateau or steady state condition. A plateau count of 0.5 microns and larger would
range anywhere from several thousand to several million particles in a typical
manufacturing environment. During off hours or lunch breaks a noticeable reduction
in the contamination level will occur,in the location of processing equipment.
(4. Cambridge Filters, 1986.)
Once the required cleanliness level is established for a facility or room the location of
each process to be performed within this area should be decided. This location of each
process should include the position of each operator in the facility. While preparing
this layout, consideration should be required for the area for each process, each
operator, and necessary service equipment.
The position of the operative relative to the process, and the location of each process
relative to another will indicate the need for laminar flow.
14
Other factors, in conjunction with the size and quanity of airborne particles, can also
influence the degree and method of contamination control. Some typical considerations
are:
Reverse isolation is used when the product generates contamination; the service
equipment or personnel, or both must be protected from this contamination. In most
cases, -the air requires special care after it has been exhausted from the area, and before
it is reintroduced to ambient conditions.
15
isolation is also necessary when two adjacent products or processes must be protected
from contaminating each other. This type of isolation is not only the most difficult to
control but the most required. Mutual isolation normally requires the use of laminar
control.
16
3. Air Pattern Control
Air should be directed to obtain the cleanest air at the most critcal work areas. As
contaminants are entrained, they are easier to be removed from the room. Two options
are available:
1. The introduction of large quanities of air at low velocities in the area of the most
critical work areas.
2. An unidirectional movement, usually downward or across the room, prior to
removal from the space. The choice of specific air arrangement should be based on the
criticality, of the conditions to be maintained in the space, size, and the ratio of space
occupied by the nature of these operations to be performed in the room.
Laminar Flow: In a laminar flow system, air is introduced evenly from one entire
surface of the room, such as the ceiling or a wall, with flow at constant velocity across
the room and removal through the entire area of an opposite surface. Laminar flow
provides a direct, predictable path which a micron size particle will follow through the
clean room, with the minimum opportunity for contaminating any room component.
It also provides a mechanism for directly removing the particles constantly generated
within the room introduced into the air stream. Ideally, the flow sreamlined would be
uninterrupted and, although personnel and equipment in the air stream do distort the
streamlines, a state of constant velocity and static pressure is approximated. Most
particles that encounter obstructions in a laminar air flow like personnel and
equipment, will strike and continue around it as the laminar air stream re-establishes
itself downward. There are, among other conditions two possible conditions that may
arise from laminar flow striking obstructions. These are called eddys, and the area of
virtually zero velocity at surfaces parallel to the air stream (This action will be further
discussed in Part II) to provide good air movement to prevent settling of particles, air
flow velocities of approximately 100 f.p.m. are recommended as standard design of
laminar flow cleanroom systems.
17
These average air velocities can be used in laminar flow rooms because of local high
velocity are minimized. Although it has been stated by sources that air velocities
below 70 f.p.m., will cause prevention of particles to settle. Normal body movement
has little effect above 70 f.p.m., while air velocities above 110-120 f.p.m. contribute
little contamination control advantage. (5. Cambridge Filter, 1986.)
Laminar flow rooms can usually be operated without the need for air locks, air
showers, and restrictions on occupant movement. These rooms fall into three
categories; vertical flow, horizontal flow, and diagonal flow with diagonal
flow used very seldom except on certain kinds of equipment processings.
Vertical flow consists of an entire ceiling of HEPA filters. Ideally, a grated floor
serves the air exhaust. As it moves through the ceiling, it is filtered essentially free of
all particles 0.3 microns and larger. This type of air flow provides a uniform shower
of air which bathes the entire room in a downward flow of clean air. Contamination
generated in the space will not move laterally against the downward flow of air and
will not contribute to a contamination level buidup in the room. This type of design
provides the cleanest working environment that is presently available.
Horizontal flow uses the same filtration air flow technique as the vertical flow
system, except that the air flows from one wall of the room to the opposite walls. The
supply wall consists entirely of HEPA filters supplying air at approximately 100
f.p.m. across the entire section of the room. This air then exits through the return
wall, or ceiling return, at the opposite end of the room and is recirculated within the
system. This type of flow has the same degree of filtration and the same air flow
characteristics as the vertical air flow sytem.
As with the vertical air flow, this design removes contamination generated in the space
at a rate equal to the air velocity and does not allow cross contamination from one side
of the room to another perpendicular to the air flow. However, a major limitation to
this design is that vertical contamination in the direction of the air flow will occur.
18
In this design, the first air, or air first coming out of the filter wall, is as clean as air in
the vertical flow room. The process activities can be oriented mostly to have the
cleanest, most crical operations in the first air, or at the clean end of the room, with
progressively less crical operations located toward the return air, or dirty end of the
room.
Point of use: All laminar flow arrangements are applicable to point of use stations
as well as entire rooms. In combination with conventional flow systems, they can
provide small areas with very high degrees of contamination control. Where critical
control is required, the laminar work station is often the only acceptable means for
meeting the operational criteria.
19
figure 2. Conventional Air Flow
4. Pressurization
There are two primary considerations to recognize when designing for pressurization.
Pressure between functional spaces will be influenced by both recirculation air quantity
and make-up airflow, while building pressurization is ultimately affected only by
make-up air flow quanity. To provide control over room pressures, air flow variations
should be minimized.
5. Lighting
21
6. Temperature Control
In cleanrooms of cleanliness class 1000 or better where full protective clothing (bunny
suit) is required, temperatures of 68-72 degrees fahrenheit (20-22, centigrade) are
necessary to satisfy human comfort. Generally, these temperatures will also satisfy
process requirements. For less stringent functions such as assembly or packaging,
where cleanliness classifications of 100,000 are common and protective clothing is
less restrictive.Temperatures of 70 -74 degrees fahreheit ; ( 21-23, centigrade) is
adequate.
7. Humidity Control
22
Humidity control is necessary to, among other things:
1. prevent corrosion
2. prevent condensation on work surfaces.
3. eliminate static electricity.
4. provide personnel comfort.
8. Cleanliness
I. I-nfiltation
H. Internal generation
III. Introduction of make-up air
23
I. Infiltration: High Efficiency Particulate Air (HEPA) filters are used in both
vertical and horizontal types of design in the control of infiltration into the cleanroom.
Air flows into the work area, and exits in either one of two ways; through perforated
floor sections into an air plenum; or through air grills at the floor level or side walls.
1. screening effect: not only particles bigger than the distance between fibres are
retained, but also particles whose streamline passes so near to a fibre that they collide.
2. inertia effect: Particles above 1 micron, where they are thrown out of their
streamline when it deviates sharply to pass a fibre of the filter median.
3. diffuse effect: Particles below 1 micron, where constant collision with the
molecules of the surrounding gas lead to an erratic path (Brownian movement.)"
(6. Schicht, 1985. )
Due to the low penetration velocity of the air through the filter paper, pleating is
necessary in order to arrive at compact filter units.
An important item to emphasis is the location of the HEPA filters. "The HEPA filters
should always be located so that they are the last items of the mechanical equipment.
This insures that all the air entering the cleanroom is filtered and that there is no
possibility of contamination with unfiltered air." ( 8. Kilpatrick, 1984.) Careful
installation of the HEPA filters is important as to not cause leakage around the edges.
The ratings of a filter is based on the resistance to air flow, its aerosol removal
capabilities, and its dust holding capabilities.
24
By definition, the HEPA filter must demonstrate a minimum collective efficiency of
99.97% for 0.3 microns (DOP) test, and a maximum filter pressure drop of 1.0" w.g.
"The 0.3" micron challenge aersol was selected since both theory (OSPD) reports
3460) and experiments indicate that a particle of this size is the most difficult type to
capture by a fibrous filter." (9. Kilpatrick, 1984.) This form of resistance has
strongly emerged within the last ten years.
1. viscous impingement.
2. dry type.
Impingement type filters are constructed of coarse fibers arranged in a porous bed.
Dry types differ from impingement types in that they have smaller sized fibers of
more densely spaced media. In both types, the pressure drop across the filter
increases with dust loading to a point where they should be replaced. This is normally
done when the air resistance reaches 2" w.g.
25
9. Controls
Controls are critical to the overall system design and required performance. However,
two things should be kept in mind:
1. A system, no matter how well designed, will operate no better than its controls.
A selection and design of the controls is an integral part of the entire design process.
Precision instruments should be specified where precision is important. This precision
is usually limited to:
as
10 Noise And Vibration
Noise is one of the most difficult variables to control within contamination control
equipment, due to the high volumes of air necessary to provide the cleanliness levels
required. For normal applications of laminar flow equipment, the noise level is
designed to be below 65 decibels. Air-handling systems, fume hoods, exhaust
systems and equipment areas (chillers, boilers, pumps, etc.) must be carefully located
and appropriate silencers and sound absorbing elements must be used to attain a sound
level that is acceptable to the personnel. Levels of vibration are controlled by utilizing
structural designs with stiffness and mass.
Pedestrian traffic within the cleanroom is one of the most important vibration sources.
The most reliable solution other than administrative controls are the structural
seperation of high traffic areas, or the use of heavier structural components to bring
footfall- induced vibrations within accepted limits.
27
11. Room Construction And Operations
Construction Finishes:
5 Lights: Flush mounted, sealed with tube removal preferrably from outside
of room.
IS
Personnel And Garments:
2. Lotions and soap containing lanolin are used to tighten skin particles.
3. Solvent contact with the skin is conducive to skin peeling and flaking and is
avoided.
6. Lint-free smocks, coveralls, and head and shoe covers are worn.
Entries:
30
Chapter 2 History
The orgins of cleanrooms today are a result of advancing technology in the field of
mechanical science. This advancement is a twentieth-century discovery. Cleanrooms
produce many present day products of sophisticated design. Control of the factors in
and about a product is a function of tecnological advancement in science. This is true
for many sciences. The first of these sciences to recognize this need was medicine.
Medical science controlled the environment in what are called "operating rooms". The
nature of work performed in an operating room on humans is similar in nature to work
performed in cleanrooms on equipment. Room decontamination functions are both
necessary and vital to continued, successful cleanroom operation. The events leading
to the development of operating rooms were initially paralleled by events leading up to
the development of cleanrooms. With greater emphasis being placed on
miniaturization of products by the electronic age, however cleanliness requirements for
some cleanrooms exceed those for operating rooms. Techniques of workers in
operating rooms and cleanrooms are similar in that they are controlled.
The major difference between operating rooms and cleanrooms is that in operating
rooms sterility of objects is of major importance. In cleanrooms, the quanity of
particulate matter is of major concern. Particulate matter is of little impotance in an
operating room as long as it is sterile. Cleanrooms have experienced little difficulty
31
1.Lack of bacteriological food.
2. Lack of bacterial colonies.
3. Low humidity.
Cleanrooms or rooms that were the predecessors of modern day clean -room can trace
their beginning back to World War I days. At that time there was no "cleanrooms" as
we know of today. But there were controlled areas that performed similar functions.
These controlled areas within factories and laboratories attempted to eliminate the gross
contamination associated with manufacturing areas. This contamination, consisting of
heavy dust- laden air, had caused seizure of small bearings and gears used in the first
aircraft instruments. As a result, controlled assembly areas were built.
By segrating the work areas from other manufacturing operations, and by providing a
filtered air supply, contamination control was first effected by cleaning practices. With
the beginning of the World War II period, better filtration systems were developed: air
conditioning and room pressurization were considered essential. Personnel protective
clothing were added later, as were air showers and personnel cleaning equipment. In
this period there were attempts to improve the quality and reliabilty of such areas as
metrology; precision optics such as bombsights, gyroscopes, and timing devices, met
limitations which appeared to relate to surface cleanliness. ( 10. Austin, 1970.)
Further study revealed that surface cleanliness was a function of how clean the
environment could be made and controlled. Up to this point there was little known
about size and quantities of particles in the manufacturing environment especially those
carried by air. When the realization of harmful effects caused by these particles on the
production of products was established, many organizations like the military started to
write their own independent and operating specifications and standards for
cleanrooms.
Many theories were generated and practices were based upon incomplete data.
Because of this the field of contamination control lost a lot of credibilty in the 1950's.
The first document accepted as a workable document for cleanrooms was one done by
the United states Air Force entitled "Standard Functional Criteria For The Design and
Operation Of Clean Rooms". (Air Force Technical Order 00-25-203) It was prepared
because of the confusion to the various contamination, temperature, humidity, and
pressure requirements for cleanrooms. The Technical Order 00-25-203 was published
in 1961. Also, in 1961 the Sandia Laboratories located in Albuquerque, New Mexico
started work on a new type of cleanroom design. This facility is a part of the United
States Atomic Energy Commission. Sandia has a vital interest in the field of
contamination control. Absolute cleanliness of contact points, freedom from
corrosion, no shorts from metallic particles were vital for the complexity of the
systems at Sandia. (11. King,1986.)
What resulted from such strict specifications was a design team from Sandia who
developed what is called "laminar flow" cleanroom design. (1962) This design
concept developed many of the designs currently in use; and then developed what is
now called "Federal Standard 209B".
During the time of the development of Federal Standard 209B this team had the
assistance, among others, all military and government agencies, and NASA. This
Federal Standard 209b was first published in December of 1963.
From its inception in 1962, laminar flow technology has been applied to the design
and modification of new and existing cleanrooms. Laminar flow systems have taken
the form of both rooms and individual work stations.
The years from 1962 to 1965 saw the laminar flow design used for mostly aerospace
industry. From the mid to late 1960's the laminar flow design was starting to be used
in the pharmaceutical industries, and the application in hospital operating rooms.
In addition to this new laminar design, other advances also occured at this time to
make the over all design of the cleanroom more efficient. Among some of these
advances were:
The microelectronic manufacturing industry is the largest user for cleanrooms these
days. Other fields include:
1. Medical
2. Pharmaceutical
3. Military
4. Aerospace
5. Food processing
6. Optics
7. Nuclear
8. Precision Manufacturing
9. Research facilities
34
2.2 Future Trends In Cleanroom Design
Revisions of the Federal Standard 209B are currently being proposed to control
particles 0.1 microns in diameter and larger to less than 10 particles per cubic foot
(Class 10). Also, for particles 0.1 microns in diameter and larger to less than 1
particle per cubic foot (Class 1). Because of this advancement, laser (light) scattering
particle counters are now being used for the counting of these particles.
(12. King, 1986.)
Class 100 is now being tested for lesser needs with class 10, and classl being
specified for the current, high quality production.
Along with these higher standards, the fundamental criteria mentioned in chapter 1 are
becoming more precise. New areas of concern are evolving. One of these areas to be
looked at is the eletrostatic changes, and electromagnetic interference which is more of
a problem with the micro-computer.
With class 10 and class 1, cleanrooms now being used, there will need to be a revision
in the Federal Standard 209B to accomodate the proposed revisions.
Chapter 3 MANAGEMENT IN CLEANROOM
OPERATION
3.1 Introduction
1. Technical consideration.
2. Attitudinal factors.
3. Clothing.
36
3.2 Technical Considerations
The management starts with the proper layout of the facility, specifically the
management of the circulation. Present cleanroom technology uses either full laminar
flow or clean tunnel-bay systems. In the full laminar flow design, the entire facility is
divided into clean modules, each module combines HEPA filtered ceilings wih grated
floors. This concept is for full flexibility in the equipment size and layout. An
exception to this would be the wall location for the return air. But you pay for this
flexibility, the tunnel approach handles traffic flow in a highly efficient manner. The
main artery, or tunnel, connecting the modules is the location of the densest traffic
flow. This system does not have the open space for flexibilty, but by not using the
entire space there is a cost savings. (see chapter 4, section 3, "Flexibility")
In either case, the overall cleanroom plan must manage the movement of personnel. At
the point of entry into a intergrated circuit manufacturing facility the person should
immediatly start mentally and physically work to create a clean atmosphere. Street
shoes should immediately be removed and or be covered by special slippers. As one
proceeds through the facility, further levels of cleanliness should immediately become
evident.
Floors, even in the reception type areas should be exceptionally clean. The level of
cleanliness should actually increase as one approaches the actual fabrication area.
These environments should take on a hospital-like appearance. Even locations
relatively far from the fabrication area, like the floors and walls should be finished
with a cleanroom type appearance as stated in chapter one. (1.3, section 11)
Besides always being constantly cleaned, the main circulation floors should be covered
with what is called a dust collecting rug or mat. The cleaning of these areas should be
done with the use of lint free brooms, and vacuum cleaners with a good filtering
system, and plastic bags to contain swept up debris.
The manner in which personnel enter and exit the fabrication areas should provide a
further measure of cleanliness. Essentially, all the people who enter into the
fabrication area should pass through several staging areas.
There will be four stages used in this example. These areas are the different levels that
should be achieved to meet a level of utmost cleanliness. The first area should be a
location where employees and visitors would remove their jackets along with the
slippers that were issued upon arrival to the facility.
The next stage would develop through a pass through air -locked set of doors into a
second area where cleanroom clothes are put on (slippers would remain on.) A third
area would be a relatively large laminar flow area, equipped with a metal grating floor
and HEPA filter ceiling. In this area, personnel would put on their second pair of
slippers and wash their hands. These areas could be equipped with what are called
electronic eyes which would allow the people to wash and dry their hands without
touching any surface. . A fourth and final stage would consist of a air wash tunnel.
This will last approximately 30 seconds. A general consensous about air showers is
that their function is more for pyschological reasons than actual performance.
At this point people would enter the fabrication area. What is important here was the
thought process people would go through to mentally concentrate on reducing the
contamination spread.
The next point to manage would be the traffic movement. Movement of workers
within the cleanroom create air currents that interfere with proper laminar flow.
Cleanrooms that are short of a perfect laminar flow are going to have pockets of less
clean air and locations of settled particulates.
Unnecessary movement in the cleanroom will stir up these problem areas and could
allow contamination to be placed upon the product being manufactured. Traffic
should be limited as much as possible for peak performance.
38
Visqueen or plexiglass shields can help in isolating critical areas from air movement
caused by the movement of personnel through critical area. Intercoms, and also
pass-throughs are helpful in maintaining low traffic levels. Any pass-through should
be accompanied by a intercom system to prevent the use of the pass- through for
purposes of conversation.
Another item to manage is the proper handling of the clean room product. Operators in
the cleanroom must realize the importance of protecting the product at all times from
contamination, and must know how to do this. Cleanroom guidelines must be
established and followed. Current personnel should be retrained; incoming personnel
and visitors should not be admitted until they have a clear understanding of these
guidelines. Everyone connected with the cleanroom facility should know and follow
the guidelines.
A final note on some specific issues that should be implemented in the management of
cleanroom operation:
1. Ordinary paper should not be used in a clean room of less than class 10,000
rating.
2. No food, gum, candy, or smoking material should be allowed in the
cleanroom.
3. The use of cosmetics, rubber bands, and plastic gloves should be prohibited in
the cleanroom.
4. If a vacuum cleaner is to be used in maintenance, make sure it is properl
filtered.
39
7. Do not allow employees to sit on work surfaces or stand on any furniture.
8. Segregate coats and sweaters from cleanroom garments.
9. Leave all personnel items outside the cleanroom area.
10. Vent all locker storage with laminar flow.
In the environment of a cleanroom facility with everyone wearing white, and working
in such a sterile atmosphere there is a sense of anonymity resulting in this necessity of
wearing cleanroom garments. In many cleanroom facilities there is a assembly line
atmosphere which after a while creates a task identity and significance problem.
This sameness of environment and task can result in an attitude problem. This is even
compromised by training programs of personnel that is primarily prescriptive (do this;
don't do that) instead of educating operators. A final point in this area, is that
"Cleanrooms, perhaps more than any other manufacturing setting ever, provide little
opportunity for operators to guage the quality or progress of their work. Both the
product and the problems that affect it are well below the visual threshold. Operators
are left to follow procedures designed to ensure proper manufacturing processing, but
this situation means that the work is more ritualized than it is product-focused. In
effect, operators are working in an infomational void." (15. Northcraft, 1986.) This
sense of anonymity also can develop into a sense of frustration, or a non-commitment
to work attitude in this working environment.
49
In conclusion, to promote a positive attitude in the cleanroom environment to increase
the performance of personnel, a number of steps should be taken. Among those:
That goes for operators, technicians, and to a point engineers knowing each others
job. This will in affect develop a repect for each others task as well.
2. Develop ways of communicating with each other. One way is by the use of an
inter-coin system. Another is to develop predictable work schedules for personnel so
that everyone knows each others schedule. This would help in not disrupting
workers.
3. Room finishes: Cleanrooms have the same problem that is apparent in operating
rooms and similar rooms in hospitals. What hospitals are doing now, and what
cleanroom facilities should do now is to create more of an exciting space visually.
This is done in hospitals now by the use of color, and wall murals. Another area that
needs to be looked in cleanrooms is the floors. For vertical laminar flow, the use of
floor grating is effective for laminar flow. But walking on this grating all day can be
very uncomfortable on the feet. What some Japanese facility designers have done was
benefiscial. "Japanese facility designers recognized that walking on grating all day
long would most likely become fatiguing. Thus, they replaced gratings with solid
flooring on major traffic paths in sections of the fabrication area where ultimate
cleanliness was not mandated.
41
It soon became evident however, that flooring made up entirely of steel grating would
have psychological as well as functional drawbacks. One Japanese manufacturer
solved this problem by placing prefilters directly under gratings."
(16. Edmark, 1984.)
As cleanrooms become more sophiscated the products will become smaller, and the
equipment will be doing more of the work load. This will make the personnel even
more alienated. And unless the personnel can develop an attitude of say, a craftsman,
or someone that feels he or she is really contributing by being managed as such,
maximum productivety will not be achieved.
42
3.4 Clothing
As mentioned in section 3.1, personnel in the cleanroom are the biggest potential
contamination contributor in the clean room. Because of this factor the managing of
the proper attire is mandatory in a cleanroom. "Full smocks (i.e., bunny suits and
jump suits) are standard issue in all fabrication areas, and considerable attention is
devoted to the tailoring of these smocks." (17. Edmark,1984.) Clean room clothing is
not a requirement to protect the personnel. It's only function is to protect the
environment of the work piece.
Personnel shed large amounts of skin flakes all the time. " The skin and hair
constantly shed tiny flakes, composed mostly of the protein keratin surrounded by
what is left of the membrane of the dead cell. Moisture droplets are ejected with every
exhalation, blink of the eye, sneeze, cough, and even every word spoken. The clothes
we wear also shed tiny particles. " (18. Waring, 1984.)
"To make matters worse, people use such products as body powders, cosmetics,
powdery deodorant sprays, and hair sprays, all of which eventually ends up as air
borne contamination. These products contain an amazing variety of compounds and
metals, including iron zinc, and titanium compounds, mica dust, carbon black, and
numerous organics. Even the best cleanroom garment available is not capable of
containing all of the human generated contamination." (19. Northgate, 1986.)
The cleanroom clothing must be made of a fabric and designed into a series of
garments which in combination will restrict the passage of particulates either through
the garment itself or by -passing closures. Cleanroom clothing must:
1. Be low-linting in design.
2. Inhibit the build-up of static charges. "A human when wearing cleanroom clothing
will generate static triboelectric means by which the rubbing of, say , the sleeves
against the body, the body against a plastic chair or bench may build a charge of
up tox 25,000 volts." (20. Waring, 1984.)
43
3. Withstand the processing for bacterial destruction when being cleaned.
4. Fully tailored.
With personnel spending so much time in these "bunny suits", it is important to design
theses outfits as to not feel too confined while wearing them. Having the proper attire,
and feeling comfortable in them is an important issue. But also, managing strict
guidelines by always being properly dresses, properly cleaned, and not having too
many people at one time walking through the clean room will help in the contamination
control. (See Figure 3.)
44
*A
K
(
vp
is,
45
Chapter 4 Energy Efficiency And Cost Savings In
Cleanroom Design
4.1 Introduction
During the research process, the need for energy and cost saving measures was always
present. With the costs of cleanrooms being "$350-750 dollars per net square foot,t
(21. Maini, 1986.) methods were being compiled whether it was by reading about
cleanrooms, talking to people working in cleanrooms, or walking through cleanrooms
themselves on ways of savings in energy or costs. This list of methods will always
be imcomplete for there will always be something that was either missed, or has not
been discovered yet. The intention here is to compile all the information gathered
and list them in this chapter. These energy, and cost saving methods will be primarily
involved in the design aspects as it pertains to intergrated circuit facilities.
4,
4.2 Methods For Energy Efficiency And Cost Savings
This section which is structured to discuss ways of energy and cost savings in
cleanroom design for microelectronics facilities is divided into twelve seperate areas
for savings in energy and costs.
These are:
1. Feasibility Studies
2. Site Considerations
3. Flexibility
4. Vibration and Noise
5. Water (Purity and Waste)
6. Protective Transfer Systems
7. Clean Room Monitors and Sensors
8. Energy Conservation in Minimizing Cooling Load
9. Fans
10. Humidity
11. Lighting
12. Testing
47
1. Feasibilty Studies
This is an important cost consideration. Knowing this important item will result in
not over designing the mechanical systems which will add high high costs which are
not needed as well as other increases in the overall costs of the facility. Along with
this, a well defined feasibility study will help in developing a more clear construction
schedule which will also help in generating a more accurate preliminary cost estimate.
"Microelectronics facility spaces can vary in cost from $125 per square foot for class
100,0000 to $750 per square foot for class 10." (22. Maini, 1986.)
2. Site Consideration
This is a very important aspect because of its large implications on the overall facility
with the location of the various building services needed to operate the facility.
Among these services which need to have easy access to the site are the accessibility of
water, and the required amounts of electrical power and clean air. The on-site water is
used to rinse the process components. "The initial purity of this local water source
can affect the ultimate cost of the sophisticated water polishing systems."
(23. Maini, 1986.)
Electrical power is a necessary source and the need for back-up emergency power for
possible power outage is mandatory. Large quantities of fresh air are needed to
supply the extensive exhausting systems that are used in the cleanroom operation.
Having a site in an open area can be important especially if the facility is a production
facility for integrated circuits which may use high concentrations of particular
chemicals. If these facilities are located next to an area of heavy concentrations of
people, than odor absorption equipment may be needed which in turn can be costly.
Possible site locations in areas of wide climatic variation will require more costly
heating and ventilating systems because of the strict controls for humidity, and
temperature that are required at the wafer fabrication areas.
Other factors to consider on the site plan for semiconductor facilities is providing
enough area for bulk storage of the many gases needed in the manufacturing process.
A site location that does not accomodate for all of these items will prove to be
potentially very costly.
49
3. Flexibility
The cleanrooms in the microelectronics facility can vary in function depending on the
process being done. This involves the use of different kinds of equipment. Also,
room for possible expansion should also be planned. With these considerations, it is
important to develop a facility which is flexible for change. Flexible in the movability
of its wall components and room layouts, and flexible for expansion of the
mechanical, and electrical services, as well as other services that are used in
microelectronic facilities like increased water capacity for increasing processes.
The use of modular building systems constructed using the materials, and finishes as
discussed in chapter one will provide for the most efficient change, cost-wise in the
microelectronic facility.
The cleanroom design solution must respond to the need for non-disruptive change
(flexibility), easy maintenance or cleaning on a periodic basis, non-shedding
construction materials, and ease in providing services for equipment through walls and
contamination control. Ceiling space costs can be substantially reduced if HEPA
filters are not used in the entire ceiling space. Only the areas that need cleanroom
control require HEPA filters.
Having the flexibility in the ceiling for HEPA filters only where needed for laminar
flow (spot laminar flow) saves on cost by minimizing fan horsepower too.
50
4. Vibration and Noise
A vibration free facility is unrealistic, but special structural design can help in the
vibration control. This is done by utilizing structural systems with stiffness and
mass. The most common system for these requirements is the concrete waffle slab.
This system maximizes the stiffness with its ribbed construction while also using more
concrete in volume than concrete slabs. This system also allows a suffienct depth to
dampen the vibrations caused by pedestrian traffic (Footfall). This is one of the most
important vibration sources to consider.
Another form of vibration is caused by the mechanical equipment. These items should
have spring isolators, inertia blocks, and flexible connections on the ductwork.
Vibration and sound are related. Vibration can cause sound, so placement of
equipment (boilers, chillers, etc.) must be strategically located. There is also some
equipment used in the fabrication of semiconductor microchips that have very stringent
environmental vibration limits. The final product may depend on the reliability of the
wafer manufacturing equipments vibration isolation system. One way is to provide a
concrete slab on which the machine is mounted on, and than which is physically
seperated from the remainder of the building's concrete floor.
51
Pneumatic isolators which are devices that rely on compressed air to support a load can
also be used on the equipment, and also on the column support for the building.
(See Figure 5.)
1. Source
2. Path
3. Receiver
2. Path control typically takes two forms. First, the distance between vibration
sources (mechanical equipment, etc.) and any equipment that might be adversely
affected by the vibration sources should be maximized. Increasing the path length
will reduce the level of ground-transmitted vibration that reaches the vibration-sesitive
equipment. The second important aspect of path control measure is to make sure that
no direct transmission paths (from source to receiver) are present in the foundation or
structure of the building that houses the complete system.
52
fIgure 4 Vibration Isolators
concrete slab
1
coignu
footing
53
5. Water
An important part of the intergrated circuit manufactufing process is the use of ultra-
clean water to wash the product at various stages of development. A major source of
contamination in the integrated circuit fabrication is using process water that is not of the
high purity level.
Future demands for pure water will only increase. "recent forecasts indicate that
production demands for pure water will increase at a 20-30% annual growth rate."
(25. Clifford, 1986.) With this aspect, available water supplies may decrease in both
quanity and quality.
54
6. Protective Transfer Systems
When examining ways to cut costs in cleanroom design, a major area to do this is the
reduction in the size of the laminar cleanroom. This can be done effectively by
leaving out the laminar portion in areas where transport of the wafers is done. Instead
of providing laminar flow, applying protective handling tunnels and box systems will
cut down on the costs of the cleanroom.
Comparing the costs of the laminar cleanroom (class 1, and class 10) at $201 per
square foot with the cost of a service aisles ($19.30 per square foot, one can easily see
the savings involved. (26. Burnett,1986.)
The laminar flow cleanrooms and the protective wafer transfer systems, called tunnel
clean concept are used to cut down on the amount of microcontamination that reaches
the wafer surface, thereby increasing the overall production yield. A protective wafer
transfer should be measured in order to determine the systems performance in
improving control of the defects per square centimeter. This measurement should be
done under conditions that are realistic.
55
7. Clean Room Monitors and Sensors
Table 1 briefly outlines the monitoring systems and technology which are most
common in intergrated circuit clean rooms. This table describes what method is used,
and by what type of monitor and technology that is used to count the particulate
present, and its minimum size in microns.
Table 1
Monitoring Technology (27. Tolliver, 1984.)
Type of Monitor Technology Used Lowest Sensitivity Level
(Microns)
56
Airborne particle monitoring systems are the most common type of particle monitors
found in cleanrooms.
The energy that is required for the environmental control in a clean -room is high.
This energy consumption is related to cooling, fans, humidification, and reheat.
Cooling energy is significant and 85% (28. Brown,1986.) of the total cooling capacity
will be required to:
57
1. Heat recovery in make-up air.
2. Highest possible water temperature.
3. Lowest cooling tower temperature.
The temperature for the make-up air that is required to provide for dehumidification in
most cleanrooms will need to provide water temperature in the 38-42 degrees
fahrenheit range.
when evaporative cooling is used for make-up air, there are many hours per year
when the outside air, while cooled to the required wet-bulb, is cooled to only 50-60
degrees fahrenheit. Since the chilled water for the recirculation air cooling and
precooling of make-up air can be accomplished with 50-60 degrees fahrenheit water,
the use of a dual temperature chilled water system is possible.
Energy consumption can be reduced for the make-up air cooling by using outside air
and pre-cool it by indirect evaporation. This can be accompolished by using a "finned
-tube heat exchanger piped either to a cooling tower or some other independent
evaporative heat exchanger." (29. Brown, 1986.) Another method of make-up air
precooling is accompolished by heat recovery, utilizing exhaust air. If a high
percentage of exhaust air is scrubbed , a heat exchanger downstream of the scrubber
can offer a good potential for additional energy recovery." (30. Brown, 1986.)
By operating the chiller with the lowest possible condenser water temperature will save
in the power that would have been needed to run the fan to lower the temperature.
This can be done by using low input power chillers by increasing the heat exchange
surface area over long operating hours.
58
9. Fans
Laminar flow clean air systems operate by moving a large quanity of air through High
Efficiency Particulate Air Filters (HEPA). The idea here is to reduce the electrical
consumption but not compromise on performance standards of fans. This power is
reduced by using less fans but larger ones. Fan energy becomes more significant as
the cleanliness class improves. Laminar flow clean air systems are large volume,
low static pressure systems. "The most efficient fan for this requirement is the low
speed, forward curved blade, squirrel cage type." (31. Marsh, 1981.) Larger fans
are more efficient than smaller ones of the same type. Motors for fans have different
efficiency factors. This is formulated by output horsepower divided by the input
wattage. The efficiency range can range anywhere from 60 to 85%. For an energy
efficient system, the efficiency factor should be at least 75%. (32. Marsh, 1981.)
Because of this, a high voltage three phase motor is not necessarily as efficient as a
lower one.
The location of the central fan system is more efficient the closer it is located to the
source. The increase in distance causes friction losses in the connecting distance.
In laminar systems the use of different fan speeds is benefiscial, especially during
non-production hours. By being able to cut down on fan speed when the facility is not
in production cuts down on the energy being used.
The laws of fluid dynamics state that whenever eddying or turbulant flow is present,
there is a greater loss in overall pressure than in a smooth flow. An abrupt change
requires more energy to conform than a gradual change. Design wise , having a
laminar flow of minimum air change can be more energy efficient. (See Figure 6.)
figure 6
5'9
10. Humidity
For proper design, the selection of monitors, and sensors for measuring and
controlling humidity require careful thought so that unnecessary costs can be avoided.
First, the humidity level and tolerance for the particular process should be determined.
Second, the number of sensors needed and then placement in the room should be
established. As mentioned before in the section of this chapter entitled "Cleanroom
monitors and Sensors," the manufacturers literature and specifications should be
carefully reviewed. Also, the cost and inconvenience of sensor recalibration and
replacement for maintaining the required tolerance must also be considered.
In summary the thorough planning before purchasing a humidity sensor, and the
placement of these in the cleanroom facility will be a savings both in money and time.
6-0
11. Lighting
Having the right amount of footcandles in the clean room is important, as was stated in
chapter one. But leaving the lights on all the time cuts down on the overall efficiency.
Keeping the lights on converts the energy from the lights into heat which adds to the
airconditioning load. Keeping the lights on when not needed also increases the
maintenance cost by replacement cost, and the maintenance cost of replacing there
lamps.
An energy efficient method to use would be a programmable control system which can
be readily applied to lighting loads with a payable period of less than one year.
(See Table 2)
Table 2
Lighting Costs (33. Schneider, 1984.)
61
basis: (1) Analysis for 1000 sq. ft.
(2) 4 watts/sq. ft. lighting level
(3) cost of power is $.10 per kwh.
(4) Airconditioning EER = 10 BTU/watt
62
12. Testing
Over the years there has been two basic types of tests for cleanroom systems to
properly evaluate a cleanroom facility:
The initial performance tests may result in corrections of problems within the
cleanroom facility before personnel occupy the space
A cleanroom facility cannot be fully evaluated until it has performed while the facility
is in operation. In laminar flow cleanrooms it is important to sample the work areas.
This sampling should be done with a light scattering photometer, providing a high rate
of samples (34. Cambridge Filter, 1986.)
Average velocity readings throughout the cleanrooms should be taken with what is
called a hot wire anemometer to determine velocity gradients and air patterns. Several
readings should be taken to give an average of each filter in the bank of filters of a
laminar flow cleanroom. With the needs of class 10, and class 1 rooms to be
certified, revisions are still being made.
63
These test areas are: (RP-006-84T)
64
Chapter 5 Vertical Laminar Flow (Research)
5.1 Introduction
Part II of this thesis is to develop a better insight with vertical laminar flow as it
pertains to cleanroom design. This insight will answer performance questions on its
design, and offer an improvement in the efficiency of the cleanroom design in regards
to the vertical laminar flow. There are cleanroom designs which also function using
horizontal laminar flow, and diagonal flow design. The majority of cleanrooms use
either vertical, or horizontal flow design.
The concern here is with the vertical flow aspect because of its heavy use and also
this design offers the most flexibility (See Figure 7.) An advantage of the vertical
laminar flow is the isolation of individual operations using laminar flow. Also, this
flow generates the shortest distance from possible contamination areas to removal from
the room. The biggest disadvantage of this form is the cost.
65
HEPA
filter
wwwN
1MAMIV~L~LkLMAA/ ~AAk
I 1~ I
I I
I I
I I wal
airi fiowj
I I I
I I
I I
I I
I I floor
A grates
iii ill liii I I I I~t] I II.III1IIIIIiII~~1
I I I I I I I
I I - I I I 1 I
figure 7 Vertical Laminar Air Flow
86
5.2 Laminar Flow (Vertical)
Laminar flow is the flow of a fluid (air) at very low velocities, moving along smooth
paths in layers with one layer moving smoothly over an adjacent layer (sometimes
called "streamline", or viscous flow) with no eddies or swirls. A viscous flow is
characterized by viscosity, which is the property of a fluid (air) that resists relative
motion of adjacent fluid layers. In true laminar flow, the action of viscosity cancels
out any turbulent tendencies.
Laminar flow cleanroom air systems operate by moving a large volume of air through
High Efficiency Particulate Air Filters (HEPA), and then quickly through the entire
area that is to be protected. This large volume of air is designed to make a single pass
through the clean area to remove all the airborne contamination and is then recirculated
back through the HEPA filters before making another pass. As velocity increases,
eddies or swirls form and this type of flow is called "turbulent". Turbulent flow is a
condition of flow which exists when the velocity exceeds a critical value. Laminar
and turbulent flow each have their own laws of resistance to motion which applies to
all fluids. This nature of flow whether it is laminar or turbulent, and its relative
position along a scale indicating its turbulent to laminar tendencies are indicated by
what is called, the Reynolds number. The nature of a given flow can be characterized
by this Reynolds number. C.Reynolds, in 1833 analyzed basic fluid flow
charateristics through experiments using the principle of "Dynamic Similarity"
( 36. Leary, 1951. page 28.)
67
This Reynolds number is defined as:
Rd = p ud/ul
3 4 5. 78 10 I 20 30
0604 B8 0 O0 200 300 400 80 -"M 4000 700
90 0 40100 000
010 ~1I 111
J- -
H1L'
I
e
i''II~
I-M
coo tIII)Il I I )
004
"E
n13i 004
003
00
002
004 0040
('I00
0013
00000 00 1
0 6 t I41-- 17 4
I06 1S0000
A64N 40tU6
figure 8
Friction factor vs. Reynolds number
showing laminar flow
68
5.3 Research Model
5.31 Design
This research model is designed to assimilate a vertical laminar air flow cleanroom.
(explained in chapter one.) As mentioned before, The vertical laminar air flow
cleanroom as presented here represents a large percentage of the type used in
intergrated circuit facilities.
This model operates in a laminar flow by using air supplied through the top and
streamlines through the ceiling perforated panel. The air moves through the cleanroom
space vertically and exhausts through the perforated flooring panels and exits out the
adjusted sliding doors which are located on all four sides of the model.
This cleanroom model is designed to move the air through the space in a vertical
manner. It was felt that by using two perforated flooring screens, 4 inches on center,
and a prefilter placed on top of the floor grating, this would structure the air movement
in a more linear pattern. Cleanroom air exhaust locations are either at wall locations,or
through floor gratings in the typical intergrated circuit facility. For a more laminar
flow, it is recognized that exhausting through the floor is more effective for laminar
flow. The design of this model is patterned after this. The scale is one quarter of full
scale, so the floor to ceiling height is 8'-O".
The research being done in this model will deal with laminar flow and not with
particulate control. Because of this, the use of High Efficiency particulate Air Filters
(HEPA) will not be needed as is required in all cleanrooms. (See Figure 9)
69
5.32 Model
R
0
-S
a
0
I
C" it
a
O
figure 9
Research Model
70
T
r
Pflotograph I
Reswearch M0,t
71
5.4 Research
This research is intended to show the relationship of people and other objects in a
cleanroom environment that is operated by using a vertical air flow to clean the room
of contaminates that are generated from the people, and equipment, etc. The reason
for this is to point out the limitations of this type of system.
A. Research Questions:
72
3. When personnel are working in a cleanroom, they are almost always
working with some form of equipment. For an effective use of verical air
flow, how should personnel best position themselves in relation to their
equipment ?
4. How are turbulence and eddys formed in cleanrooms? What are some of
the methods that may be used to prevent this from happening?
B. Alternate method for more effective vertical laminar flow in the cleanroom
environment.
73
5.5 Discussion On Research Process
The research was developed by first testing the model. The model was tested to
determine how uniform the air velocity was coming through the ceiling. This was
performed by using an IBM PC-AT computer with a "Based Data Acquistion System".
Attached to this unit was a anemometer and a hot film sensor probe which was placed
inside the model while the system was running. An advantage of using a hot film
sensor probe is that it has the ability to measure low velocities like those in
cleanrooms. A hot film sensor probe is essentially a conducting film on a ceramic
substrate. The program for this was entitled "Uncle Scope". (developed at MIT)
With this system running, and with the film sensor at a fixed position, the "Uncle
Scope' program was able to give out 1,024 readings, every 20 seconds on the
monitor. Every 20 seconds there were 3 velocity figures given.
1. minimum velocity.
2. maximum velocity.
3. average velocity.
These field measurements gave an overall series of numbers. These numbers were the
velocity readings. The closer the numbers were to each other the more uniform the air
flow was. The intent of this model is to provide a uniform flow. This is what is
achieved in a typical cleanroom. The sensor was placed randomly all through the
model at a working height of 42". (See chapter 1.3 , fundamental issue number 5
entitled "Lighting".
The difference of reading the voltage point output of the anemometer between the
lowest point and the highest point was .13 volts. For this research the velocity across
the ceiling of the model was judged uniform.
74
As mentioned earlier the model was designed at 1/4 full scale. This cleanroom
space has a floor to floor height of 8'-0". The human figure in this room is 6'-4", and
the table is 42" above the finished floor.
1. video tape: The video tape was used to observe the air movement. This tape was
attached to a thread that was placed across the ceiling of the model. The video tape
hung down individually next to the figure and table in different lengths to conform to
the shapes. This video tape stopped just short of touching the floor. The placement of
this tape hung freely and did not touch the figure or the table that would have
consequently influenced the movement of the video tape.
The vertical air flow was turned on, and the movement of the air was depicted by the
movement of the tape. (the more 'streamline' the air flow, the less the tape would
move.)
The tape when placed in the room with the system on , visually moved slightly. After
running the system several times, over a period that involved several days it was
observed that
A. The video tape moved slightly. This movement was in the form of a "wiggle".
B. The video tape moved the most by the floor. This movement started to appear
approximately 1-2" from the floor.
75
C. The video tape seemed to curve towards the figures face, especially when the
figures head was crouching over the table, similar in nature to an operator
working on processing equipment in a cleanroom. The video tape was a
preliminary study for the smoke sticks.
2.* smoke sticks: The smoke sticks were used to detect air currents for visual study.
Each smoke stick provided a continuous 10 minute smoke with each one giving the
effect of 2 ciggerates smoking simultaneously.
The smoke sticks were attached to the ceiling above the figure. When the air system
was turned on, the smoke moved in a 'streamline' fashion down the front of the figure
and working area. This action shows how the vertical air flow "washes" the area of
contamination and, in this case, exits through the floor. without this process,
contaminates would otherwise settle on the chip being processed, and cause a defect
to occur on the chip.
As indicated by the pictures the laminar air current in the room was affected by the
precence of furniture, and people. (See Photographs 2-11.)
76
5.51 Response to Research Questions. ( 5.41 Research)
1. Does the vertical 'streamline' air flow (laminar flow) as defined by Federal
Government Standard 209B (see appendix) adequately eliminate the particulates that
cause defects in chips being produced in a cleanroom ?
The problem with this is that particulates that are in this location will
be recirculated causing more possible defects on the chips being
manufactured. There is not total vertical laminar flow when people,
and equipment are placed in a cleanroom.
77
The design concept of the vertical laminar flow cleanroom is to reduce
the small scale turbulence to the point where it is acceptable and to
prevent the development of large scale turbulence resulting from objects
placed in the flow field, movement of people, thermal sources, etc.
Users of the vertical laminar air flow, and horizontal laminar air flow
must recognize these principles and seek to achieve the best possible
arrangements of equipment, personnel and movement to maximize the
effectiveness of these systems.
2. Question: Is the location of personnel in the cleanroom effective for the vertical
laminar flow design?
3. Question: When personnel are working in a cleanroom, they are almost always
working with some form of equipment. For an effective use of vertical air flow,
how should personnel best position themselves in relation to their equipment?
78
With the air flow coming down from the ceiling, this flow will then
either strike the equipment than the people, in that order or vice versa.
When the people are struck first by the vertical air flow, the potential
for defects on the chip are more likely. The better relationship is when
the equipment is affected by the vertical flow before the people.
4. Question: How are turbulence and eddy's formed in the cleanroom? What are
ways that can be used to prevent this from happening in the cleanroom .
Turbulent flow is an irregular flow, and eddy's are air flow that are
running contrary to the main flow. These are caused by the presence of
personnel and equipment in the cleanroom that disrupt the streamline air
flow by blocking this continous flow.
79
B. Alternate solutions for more effective vertical laminar flow in the cleanroom
environments.
80
figure 10
Showing proposed method for more efficient operation
of vertical air flow
81
5.52 Research Photographs
82
h otgraph
s rtI *rntar fo w
83
i
84
hotoqraph 4
t aif fOiw moving towards operators f ace
Processing rea
85
Photograph 5
Note turbulence and eddy's forming by
operators chin, and edge of table.,
86
Photograph 6
note injection of additional air by copper tube
per lt sidet ection ust startinq
87
hotograph 7
Additional air being lniected by tube located
upper left, Notice air being washed by operators
88
P~otograph S
Close up showinq turbulance forming undqr
operators chln. (additional air mat injected at
this point.
89
Photograph 9
Close up showlnq turbuleace forming by
operators chin and On top of processing area.
(additional air not injected at this tim0)
90
Additten! alr 6014., lnl.9ted by epper tuba
91
Photograph t
Cs -up of air benq inet0 d t font of
Operator and processing are"
92
5.6 Related Studies / Future Considerations
There has been other research through the years on laminar air flow since its
conception in cleanroom design which occured in 1962.
With what is described as laminar flow has been the various velocities of air movement
to give an overall efficient operation in conjunction with dispersion of particulates.
This has been accepted to be between 90-120 feet per minute. A velocity in this range
appears to be a reasonable value for well designed unidirectional flow systems to
achieve minimum dispersion and rapid recover' downstream from flow disturbances.
Though the research here has been focused on vertical air flow, the relationship of air
flow and particular control is unseperable, among the many studies that have been
done in this area are:
In planning for the future, the use of robotics, and other forms of automation should
be included. Automated guided vehicles are currently being used today, though
largely for material transfer. ( 37. Zygmont, 1986.) By reducing human handling,
and presence, the contamination level in cleanrooms will decrease. The semiconductor
industry in the United States is at an important stage now because of pressures from
overseas manufacturers. This is not from lack of new technology, rather from less
than efficient use of existing technology.
93
Chapter 6 Conclusion To Thesis
Manufacturing efficiency has become a key success factor within the integrated circuit
industry worldwide. The on going drive to increase the number of circuit functions
per device by shrinking its size and expanding die sizes has made today's devices
more vulnerable to all types of submicron contamination. This has made it necessary
to reduce and maintain defect densities as close as possible to zero.
At this point, additional research could coincide with this study, which would focus in
on the effects of the actual particulates. This could be addressed by how these
particulates respond to this air movement.
94
APPENDIX: FEDERAL STANDARD NUMBER 209B
95
Room 14-0551
77 Massachusetts Avenue
Cambridge, MA 02139
M ITUi brari0
Document Services
Ph: 617.253.2800
Email: docs@mit.edu
http://libraries.mit.edu/docs
DISCLAIMER OF QUALITY
Due to the condition of the original material, there are unavoidable
flaws in this reproduction. We have made every effort possible to
provide you with the best copy available. If you are dissatisfied with
this product and find it unusable, please contact Document Services as
soon as possible.
Thank you.
Pages 96 - 131
Fed. std. so. 2091
AtENDMENT-1
pay 30, 1976,
IEDERAL STANDARD
PACE 1
Cortents, after paragraph 5.1.1 add: Paragraph 5.1.1.1 Class 1000 (35).
PACE 3
Pararraph 5.1: Delete sentences three and four and add as folleva:
Particle counts are to be taken at specified Intervals during work activity periods
at a location which will yield the particle count of the air as it approaches the work
lecation. The preferred location for the particle count is at wor* level height with
the sarplinr probe pointed into the airstreaw. Other classifications (i.e. 500, 1000,
8V,00C etc.) may be used for particle count levels where special conditions dictate
their use.
PACE A
TABLE I
*1'etric syster
no::: Counts below 10(0.35) particles per cuhic foot (liter) are mureliable except
when a larte number of samplinCs are. taken.
PACr &
5.1.2.1 Class 100 (35). Particle count set to exceed a total of 1,000 particles
per cubic foot (35 particles per liter) of a aise 0.5 micren and larger (see 5.2).
75C 6636
I
96
Fed. Std. No. 2093
FACE 5
Table II, Particle Ose distribution curves, add a straight line denoting Class
1,000 (35) as follows:
A line originatiag at an ordinate value of 1,000 (35), bearing parallel to the three
established class lines on Table II, and terminating at an abscissa value of 12. The
line shall be selid above the ordinate value of 10 (.35) and dotted below it. The
line shall be labelled Class 1000 (35).
PACE 20
Paragraph 60.1.1 Nonlaminer flow clean rooss: Delete paragraph and substitute:
60.1.1 Nonlaminsr flow clean moovs. Airborne particle counts should be taken at
specified intervals or representative locations throughout the work activity area. The
preferred location for the particle count is at work level height with the sampling
probe pointed into the airstreas.
PACE 25
PACE 26
Addendum to Table III, Facility Guidelines for Air Cleanlinesa Classes as follows:
TAIL! III
GSA DC 77.1016
97
Ted. ltd. No. 209B
FEIERAL STANDARD
FOREWDRD
Withir. recent years, the application of clean ro technology has spread fro
camputing, and food processing. This expansion of the technology brought increased
experience and knowledge not only in the diversified requirments for air cleanliness,
but also in the methods whict wust be employed to meet these requirements.
Since the necessity for control of airborne particulates has been demonstrated
in maL.y govermental activities, the need for standardization has been recogr.ized.
Orders for this publication are to be placed with General Services Adinistration, act-
ing as ar. agent for the Superintendent of Doments. Single copies of this standard are
availatle at the GSA Business Service Centers in Boston, Nev York, Atlanta, Chicage, Kans
City, M, Fort Worth, Denver, San Francisco, Los Angeles, and Seattle, WA. AddItional
copies may be purchased for 45 cents each from the General Services Adrinistration.
Specifications Activity, Printed Materials Supply Division, Building 197, Naval Weapons
Plant, Washington, DC 20107.
98
-~ nssaw Mms
99
I
Fed. Std. No. 20%
%'V
CONTENTS
Page
100
Fed. Std. No. 209z
CONTENTS (cont.)
Page
section 6. CHANGrS 6
Section 7. CONFLICT WITH REFERENCED SPECIFICATIONS 7
TABLES:
I Air cleanliness classes 4
I Particle size distribution curves 5
101
Fed. Std. No. 209B
APPENDIX
Page
102
Fed. Std. No. 20%
APPENDIX (con%.)
Page
f03
Fed. Std. No. 209E
APPENDIX (cont.)
Page
104
Fed. Std. No. 209
APPENDIX (cont.)
Page
105
Fed. W14. No. 209B
1.1 Scope. This document establishes standard classes of environmental air con-
trol within clean rooms and clean work stations.
2. REFERENCED DOCUMENTS
2.1 Documents. The following documents, of the issues in effect on the data of
invitation for bids, form a part of this standard to the extent specified herein.
Nongovernmental:
(Application for copies should be addressed to the American Society for Testing and
Materials, 1916 Race St.. Philadelphia, Pa.. 19103.)
3. DEFINITIONS
3.1 Clean room. A clean room is an enclosed area employing control over the pa:
ticulate matter in air with temperature. humidity, and pressure control, as require
106
Fed. Std. No. 209B
To meet the requirements of a "clean room" as defined by this standard, all clean
rooms must not exceed a particulate count as specified in the air cleanliness class
of 5.1. 3.
3. 2 Clea-. work soa-:cn. A clean work station is a work bench or similar working
enclosure characterized by having its own filtered air or gas supply. -
3.3 Pa-acle size. Particle size is expressed as the apparent maximum linear
dirernsion or diameter of the particle. (Note: While measurements from light scat-
terir.; equ prent may vary some frorr. maximum linear dimension, this variation is
n.t us-..ally considered significant for normal clean room monitoring requirements .)
4. GENERAL REQUIREMENTS
4.14 Clear roem or clean work station area. These areas are to be operated with
em;..asis or mirnimizing airborne particle contamination to levels within the limita-
ticnrs nr_ icate4 in 5.1.
4.2 Ec.:i-mer.o cal:tratic.. All equipment used to control, monitor, and record
c:ear. room and wcrk station conditions shall be calibrated as specified.
4.4 Clear. roorr. air pressure. All clean rooms shall maintain a pressure above
that of surrounding areas to assure that all leakage shall be outward.
4. 5 Air change rate or airflow. Either the air change rate or the airflow velocity
shail be specified. 1
IWhz'n Lhe terms "as specified" or "shall be specified" are used without further
reference, the degree of control needed to meet product requirements will be
specified by the user or contracting agency.
107
Fed. Std. No. 209
4.7 Humidity range. The relative humidity range shall be 40 percent t5 percent
at the temperature/humidity control point, unless otherwise specified. Reasonable
allowance should be made for instrument and system tolerances when specifying
relative humidity requirements.
4.8 Audio noise level. The allowable audio level shall be specified and should
conform to local health codes and the Occupational Safety and Health Act of 1970 un-
less otherwise excepted.
4.9 Vibration. The amount and character of allowable vibration shall be spec-
ified.
4.11 Other environ.mental factors. Due consideration should be given to other en-
vironmental factors such as light level, electromagnetic radiation, ionizing radia-
tion, radioactive particles and particularly gases and vapors such as mercury and
cleaning solvent vapors. Adverse environmental conditions such as the presence of
hazardous materials are mentioned only to alert the user to the need for adequate
controls and may involve considerations beyond the scope of this Standard.
5. DETAILED REQUIREMENTS
5.1 Air cleanliness classes. The three classes defined by this standard are
shown in Table 1. Classifications are based on particle count with a maximum allow
able number of particles per unit volume 0. 5 micron and larger or 5. 0 microns and
larger. Particle counts are to be taken during work activity periods and at a locatio
- which will yield the particle count of the air as it approaches the work location.
Other classifications (i. e., 1000. 5000, S0, 000 etc.) may be used for particle count
levels where special conditions dictate their use. Such classes will be defined by t0
intercept point on the 0. 5 micron line in Table IDwith a curve parallel to the three
established curves.
108
Fed. Std. No. 209
TABLE I
Me:rnc svster-
5.1.1 Class 100 (3.5). Particle count not to exceed a total of 100 particles per
cubic foot (3. 5 particles per liter) of a size 0. 5 micron and larger. (See 5.2.)
5.1.2 Class 10, 000 (350). Particle count not to exceed a total of 10, 000 particles
per cubjc foot (350 particles per liter) of a size 0.5 micron and larger or 65 par-
ticles per cubic foot (2. 3 particles per liter) of a size 5.0 microns and larger.
(See 5.2. )
5.1.3 Class 100, 000 (3500). Particle count not to exceed a total of 100, 000 par-
ticles per cubic foot (3500 particles per liter) of a size 0. 5 micron and larger, or
700 particles per cubic foot (25 particles per liter) of a size 5. 0 microns and larger.
(See 5.2.)
5.1.4 Statistical average particle size distribution. In Table 1l the three classes
of clean rooms and work stations are depicted by the statistical average particle size
distribution curves. While the definitions for these classes are taken from these
curves, it should be recognized that single sample distributions may deviate from
these curves because of local or temporary conditions.
169
Fed. 4td.No. 209B
10, 000
050)
1,cM:O ~2
(35)
11
100
ii
(3.5) ES
B-
100
(.351
05 L0 5 10 100
Particle Size (microns)
Counts below 10 (0 35) particles per cubic foot
(liter) are unreliable excepi when a large number
of samplings is taken.
110
Fed. Std. No. 209
5. 2.1 Particle counting methods. For proof of meeting the requirements of the
class of clean room or clean work station, one or more of the following particle
couti;rL methods shall be employed on the site of use:
(a) For particle sizes 0.5 micron and larger, equipment employing light scattering
principles shall be used, as specified in ASTM F 50.
(b) For particle sizes 5. 0 microns and larger, microscopic counting of particles
collected on a membrane filter, through which a sample of air has been drawn, may
be used, as specified in ASTM F 25 and SAE-ARP-743.
(c) Other monitoring methods and equipment may be used only if demonstrated to be
of accuracy and repeatability equal to those methods listed in 5.2. 1(a) or 5.2.1(b).
Manua: microscopic methods are adequate for monitoring air in the classes 10, 000
to ICC, 000 range. When monitoring air below the class 10, 000 level, so few 5.0-
rr.:crons and larger particles will be present that the manual method may not detect
er:_-g- of these size particles to produce a statisticaUy valid determination; there-
fcre, air monitoring in the range below 10, 000 can be done only with light scattering
e: ;.: e nt.
6. CHANGES
When. a Federal agency considers that this standard does not provide for its essen-
tial needs, written request for changing or adding to the standard, supported by
ade;uate justification, shall be sent to the administration. This justification shall
explain wherein the standard does not provide for essential needs. The request shall
be sent in duplicate to the General Services Administration, Federal Supply Service,
Standardization Division. Washington. D.C.. 20406. The administration will deter-
mine the appropriate action to be taken and will notify the agency.
111
Fed. Std. No. 209B
Where the requirements stated in this standard conflict with any rquirement in a
referenced specification, the requirements of this standard shall apply. The nature
of conflict between the standard and the referenced specification shall be submitted
in duplicate to the General Services Administration. Federal Supply Service. Stan-
dardization Division, Washington. D.C.. 20406.
Army
Navy
Air Force
Department of Commerce
112
Fed. Std. No. 20E
APPENDIX
FOREWORD
Tw: dis:ir.:t approaches to the design and operation of such clean rooms are pres-
en*&%y utihzed, both of which are covered in this document. One approach, some-
t:mes caled a conventional clean room, but referred to here as nonlaminar flow,
makes use of highly filtered and conditioned air brought into the area through indi-
vid..a: d.:ffsers located in the ceiling, and exhausted through return air ducts lo-
cated near the floor around the periphery of the room. Emphasis is placed on
l.iting the amount of contamination introduced into the air in the room by closely
c r.:ro:hng the personnel, operations, and materials inside the facility. Personnel
are req.ired to wear low particle shedding type garments and all material must be
clear.ed before being introduced into the area. Accumulated contamination should be
rernoved through adequate maintenance and janitorial service. Normally, critical
operations would not be performed in a nonlaminar flow room.
The other approach utilizes highly filtered and conditioned air brought into the room
towards the work area through a filter bank comprising an entire wall or ceiling of
the room, and exhausted through the entire opposite surface facing the air inlet filter
bank. The air is moved through the room in a laminar flow fashion, thus making
only a single pass through any given area of the room. This laminar flow air move-
ment carries out of the room any airborne contamination released by personnel and
equipment, or generated by operations in the room. Contamination generated in
localized areas of the room may be isolated from other areas by the striations of the
113
Ted. Std. go. 209a
114
Fed. Std. No. 209
Military Specification:
10.2 O-ter p.:blications. The following documents form a part of this standard to
the extert specified herein. Unless otherwise indicated, the issue in effect on date
of invitation for bids or request for proposal shall apply.
Gcvern.-ental:
(A p,1cation for copies should be addressed to the U.S. Atomic Energy Commission,
Technical Infrrmation Center, P.O. Box 62, Oak Ridge, Tennessee 37830.)
20. GLOSSARY
20.1 Nonlarninar flow clean room. A room supplied with filtered air with no
116
Fed. Std. go. 209B
20.2 Nonlaminar flow clean work station. A work station characterized by nonuni-
form air patterns and velocities. This includes work stations which have constricted
air exhaust or ports.
20.3 Laminar airflow. For the purpose of this standard, laminar airflow is de-
fined as airflow in which the entire body of air within a confined area essentially
moves with uniform velocity along parallel flow lines.
20.4 Laminar airflow clean room. A room in which the laminar airflow charac-
teristics predominate throughout the entire airspace, with a minimum of eddies.
20. 5 Laminar airflow clean work station. A work station in which the laminar air-
flow characteristics predominate throughout the entire airspace, with a minimum of
eddies.
20.6 High efficiency particulate air (HEPA) filter. A filter as specified in MIL-F-
51068 with a minimum efficiency of 99. 97 percent as determined by test. The test
can be by the homogeneous dioctylphthalate (DOP) method or other equally sensitive
method at an airflow of 100 percent of the rated flow capacity for all size filters and
at 20 percent of the rated airflow for sizes 4, 5 and 6.
20.7 Prefilters. A filter to trap gross particulates located upstream from and wi!
lower collection efficiency than the HEPA filter. The efficiency of initial prefilters
is usually in the 20-30 percent range by the NBS Discoloration (Dust Spot) Test. whi
intermediate prefilters usually have a collection efficiency of 80-90 percent by the
same test.
20.8 First air. The air which issues directly from the HEPA filter.
20.9 First work location. The work location nearest the downstream side of the
HEPA filters in a laminar airflow device.
116
Fed. Std. No. 209
fabricated or tested, in which case the nominal temperature and range should be spec-
ified. For example, the temperature variation at the control point may range from
plus or minus 0. 25*F (0 0.140 C) in the most critical operations to as much as plus
or minus 5.0*F (. 2.8*C).
30.3 Relative humidity range. Relative humidity should be controlled in the clean
roorn area. Choice of range should be based primarily upon product requirements.
Two general problem areas should be recognized: (a) Rusting of parts can occur and
become a serious problem at relative humidities above 50 percent. Hygroscopic par-
ticles that collect on surfaces can adsorb enough moisture from the air to become
starting points for corrosion pits and will adhere to surfaces more tenaciously than
particles in low humidity environments. (b) Electrical static charges or. dielectric
rraterials or parts can cause problems due to particle attraction at low relative
humi.ies.
3C.4 Make-up air. Make-up air volume supplied to the clean room area should
meet American Society of Heating. Refrigeration and Air Conditioning Engineers
Guide, using agency or local building code requirements. Make-up air should be
pre!iltered to at least the same degree as recirculated air. In rooms utilizing a rel-
atively lwc volume of recirculated air, make-up should normally be specified as a
percen'age of the incoming air into the room; e. g., 5-20 percent fresh air make-up.
Ir rooms utilizing a high volume of recirculated air, make-up air should normally
be specified as a given volume per minute; e.g.. 30 cubic feet per minute per person.
Wre. vented hoods are used for vapor control, make-up air volume should be greater
than that of the exhausted air in order to maintain a positive pressure in the room.
Wher a large percentage of make-up air is required, air-to-air rotary heat exchang-
ers or run-around systems should be considered to reduce the air conditioning load
(i. e. , heating or cooling).
30. 5 Pressure. The minimum positive pressure differential between the room and
any-adjacent area of less clean requirements should be 0.05 inch of water, with all
entryways closed. When the entryways are open, the blower capacity should be ade-
quate to maintain an outward flow of air to minimize contamination migrating into the
room.
30.6 se. Clean room noise should be controlled to levels which permit neces-
sary communication, meet operational or product requirements and which are within
117
led. 514. NO. 20B
the range of personnel comfort and safety. Sound level meters which comply With
ANSI Standard S1.4 set on the A scale slow response are generally acceptable for
measuring broad band noise. Excessive low frequency or pure tone noise may
necessitate the use of other scale readings or an octave band analysis.
40.1.1 Clean room shell. Clean room shell, floor, walls, and ceiling should be
designed and constructed with materials and in such a manner as to eliminate air
leaks into or from the room. The walls and ceiling material should be low shedding
and easy to clean. The floor covering should be low particle shedding and sufficient-
durable to withstand wear imposed by traffic and clean room operations.
40.1.3 Anterooms. Anterooms should be provided as needed for clean room per-
sonnel clothing change area, clothing storage, wash-up facilities, air showers. toile
facilities, and other equipment for personnel clean room entry requirements. Ante
rooms may also be provided, as needed, to house parts cleaning and other room
support equipment.
40.1.4 Air supply and filtration equipment. Air supply and filtration equipment
should have adequate capacity to filter all recirculated and make-up air entering tht
room. A three-minute room air change is usually considered to be minimum for
rooms 8 to 12 feet (2.438 to 3.658 meters) high. Equipment should be provided to
supply fresh, or make-up air, as required.
Access should be provided for changing and testing the filters and for cleaning the
ducts leading to and from the filters.
118
Fed. St. No, 20%
40.1. 6 Clean room furniture. Clean room furniture should meet clean room opera-
tional needs and should be constructed of low particle shedding materials.
40.1.7 Clean room lighting equipment. Clean room lighting equipment should be
provided to meet work requirements within the clean room. Shadowless, uniform
lighngr at intensity levels of 100 to 150 foot-candles (1076 to 1615 lux) at work pos-
ition is satisfactory for most clean rooms. Ceiling light fixtures should be flush
mounted and sealed to prevent air leaks.
40.1.8 Fire protection. Clean room fire protection should be consistent with prod-
uct and local requirements and with National Fire Protection Association standards
fcr anticipated occupancy and use. Smoke detection units should be installed in the
downst ream end of the clean room to automatically shut off blower motors in case
of fire. A conveniently located remote blower switch should also be provided to shut
off blower motors in emergencies.
40. 2.1 Clear. room shell. For economy of operation, laminar airflow rooms should
have floor, walls, and ceiling assembled in such a manner so as to inhibit the leaking
of ary%air into or from the room. The materials for walls and ceiling should be low
sked:.ng. and the finish should be readily cleanable. Floor covering should have low
sherjg characteristics and should be able to withstand wear imposed by personnel
andt cperations within the room.
119
Fed. Std. No. 209B
40.2. 5 Clean room furnishings. The furniture and equipment which will be used in
the clean room should be constructed of low particle shedding and low static gener-
ating materials.
40. 2. 7 Laminar flow rooms. Several configurations are feasible with laminar air-
flow, some of which are:
40.2. 7. 2 Ceiling-to-floor airflow. This vertical downflow design usually has the
capability of providing the greatest control for the entire clean working environment
because airborne contamination generated by personnal or a specific operation is in
mediately carried down and out of the room. This design (Figure 2) may be varied
such that the blowers and HEPA filters are remotely located and the filtered air is
ducted to the top of the room and enters the room through a diffusion ceiling arrang
ment such that the airflow meets the provisions of 20.3. This variation may reduc
cost, the area of filter bank required, noise, and vibration, but it may also requirt
greater motor/blower capacity and greater attention to leak proof duct joints.
10
Fed. St.d. No. 209%
4' . 2. 8 EPA filter bank. The HEPA filter bank of laminar flcw rooms should
ccver eit*her one entire wall or the entire ceiling (Figures 2. 3, 4, and 5), except
whe. diff.sion ceiling or wall systems are used or when built-in benches are in-
cl.ded at the incoming air end of the room (Figure 1). In the latter case, the wall
filter bank may cover only the area extending from the bench work surface to the
ceiling. Class 100 conditions should exist in the zone immediately downstream from
the filter bank or the diffusion ceilhng or wall.
40. 2. 9 Final filters. All final filters should be of the HEPA type. Specify HEPA
filters free from pinhole or other leaks. DOP smoke penetration test ratings should
include penetration, if any, through the filter gaskets.
40.2.10 Pref:lters. Prefilters should be used to prolong the life of the HEPA
filters. The anticipated airborne contamination level to which the prefilters might
be sub,,ected should govern the decision to use one or two sets of prefilters and the
selection of prefilter efficiencies (see 20. 7). Initial prefilters remove the gross
i21
Fe~d. Std. SO. 209B
particulates and fibers whereas the intermediate prefilters are reasonably effective
in removing smaller airborne particles. Prefilters should be checked periodically
to determine whether they should be cleaned or replaced.
40.2.11 A ir exit. The air exit from vertical and horizontal flow rooms (excluding
vertical curtain units and tunnel rooms) should normally consist of an entire wall
area or grating flow surface. While other air exit designs are used, any decrease
in air exit area will reduce useable, highly controlled space. In these rooms. pre-
filters located behind the exit wall grills or grating floor will provide sufficient pres-
sure drop across the exit area to help assure a uniform room airflow. Manually
operated dampers in the air exhaust area may be required to maintain laminar flow
characteristics.
40.2. 12 Airflow velocity. Airflow velocity through the cross section of the room
normally is maintained at 90 feet (f-27. 5 meters) per minute with a uniformity within
plus or minus 20 percent throughout the undisturbed room. In certain applications
where user requirements permit, airflow velocity in vertical laminar flow rooms
may be reduced below the 90 fpm level. Any significant reduction in velocity can in-
crease both the clean down time and the possibility of cross contamination between
work locations.
40.3 Laminar flow clean work stations. Clean work stations defined by this docu-
ment as the laminar flow type should meet the following: (See Figures 7. 8. and 9).
40.3. 1 Containment surfaces. As applicable, side panels, tops and work surfaces
of laminar flow work stations should be constructed in such a manner that they will
provide a cross sectional area equal to and in the same configuration as the effectivf
filter area. The effective filter area is defined as that area within the sealant of the
external filter frame. There should be an air-tight seal at the juncture of the con-
tainment surfaces and the filter/bench frame to prevent ambient air from being
drawn into the bench at this periphery. These containment surfaces are normally
perpendicular to the filter face.
122
Fed. Std. No. 209-
40.3.2 Final filters. Final filters should be of the HEPA type and should occupy
the entire area of the.rear or top surface of the enclosure.
40.3.4 Air exit area. The area of air exit from the enclosure should be equal to
the effective filter area in clean work stations.
40. 3. 5 Airflow velocity. Airflow velocity out of the air exit of an unobstructed
wo-rk stat:on should be maintained at 90 feet (27. 5 meters) per minute average with
a unifcrrnity within plus or minus 20 percent across the entire area of the exit. Air-
fIcw velocity should not be measured closer than one inch to the internal contain-
rner.t surfaces.
50. 1 La--inar flow rocns. The following tests should be made on laminar flow
roc-s ater installation and thereafter as specified:
50.1 (a) An in-place filter test should be made to determine whether the HEPA
f.):er bank has any significant leaks. Tests should be made to determine leaks (1)
ir the fiter media itself, (2) in the bond between the filter media and the interior of
the filler frame, (3) between the filter frame gasket and the filter bank supporting
fra-es, and (4) between the supporting frames and the walls or ceiling. Leak tests
2
sho.ul be made by introducing a high concentration of smoke or fog into the plenum
4
upstream of the HEPA filters (concentration should be of the order of 10 particles
above the minimum sensitivity of the photometer used as the detector; this normally
2For exarr.ple, cold generated DOP fog. Ref: NRL 5929 "Studies of Portable Air-
Operated Aerosol Generators, " Echols and Young. National Technical Information
Service, Springfield, Va., 22151.
Fed. Std. No. 209B
0
will be on the order of 80 to 100 micrograms per liter). The entire downstream sur-
face of the HEPA filter installation is then scanned with an aerosol photometer probe
at a sampling rate of at least I cubic foot per minute. The probe should be sized to
provide approximate iso-kinetic sampling and should be held I to 2 inches from the
filter media and frame. (Thus for measuring laminar flow equipment having air ve-
locities of from 70 fpm to 110 fpm, the probe opening should be sized from I to 1. 5
square inches.) An aerosol photometer reading equivalent to 0. 01 percent of the up-
stream smoke concentration is considered a significant leak and should be sealed.
50. 1 (b) The performance test specified in 50. 1 (a) may be modified provided the
product or area requirement is such that a 99. 97 percent filter efficiency is adequate
and if it can be determined that the entire filter bank meets the requirements for
HEPA filters. This provision applies only for vertical or horizontal laminar flow
rooms where the air is recirculated.
50. 1 (c) Airflow velocity should be measured through the cross section of the room
and should conform to 40. 2.12.
50. 2 (a) The HEPA filter used in the incoming air ducts of nonlaminar flow rooms
should be scanned per 50. 1 (a).
50. 2 (b) Tests should be made at the point(s) of air inlet to the room if the HE PA
filter is not accessible and to check for aspiration of contamination from leaks in the
incoming air ducts between the filters and air inlets.
50.3. (a) Tests for laminar flow clean work stations should be conducted as pres-
cribed in 50. 1 (a).
50.3 (b) Airflow velocity out of the air exit of an unobstructed clean work station
should conform to 40. 3. 5.
50.4 Bonlaminar flow clean work station. The HEPA filter in the incoming air
duct should be scanned per 50. 1 (a) if it is readily accessible or by a method simila
to that described in 50. 2 (b).
124
Fed. Std. No. 209%
60. MONITORING
C. 1.1 N:n.laminar flow clear rooms. Airborne particle counts should be taken at
represetative locations throughout the work activity area, at work level height, and
a- specified intervals.
6.1. 2 Larinar flow clean rooms. Normally it can be determined that the contain-
ination level of the first air in a laminar flow device is acceptable if the facihty will
pass the filter leak test and air velocity test described in 50. 1.
Ir. a vertical laminar flow (VLF) room, the above facility performance check is
usually more meaningful than an attempt to make particle counts where extremely
few particles exist. If particle counts are desired for highly critical or unusual
operations, care should be exercised that the particle counting instruments do not
interfere with either the approaching clean air or the activity.
In a horizontal laminar flow CHLF) room, the contamination level of a work location
will be innfuen- by the contamination produced at work locations preceding it in the
airstream. A facility performance check will normally determine the acceptability
of air cleanliness at the first work location downstream from the NEPA filters. The
125
Fd. Std. NO. 209B
60. 1.3 Nonlaminar flow work stations. All work stations should be monitored at
representative work locations at specified intervals.
60.1. 4 Larrinar flow clean work stations. All new work stations and work stations
in which new HEPA filters have been installed should be monitored per 50.3 after
being put into operation, following any relocation, and thereafter as stipulated in a
specification. Periodic particle counts should be made at representative locations.
60. 1. 5 Changing HEPA filters. The methods used to determine when HEPA filters
should be changed are:
(a) When the airflow drops below the minimum rate, and installing new prefilters
does not increase the flow.
(b) When the static pressure drop across the filters exceeds the static capability of
the blowers.
(c) When the filters develop uncorrectable leakage leading to consistently abnormal
particle counts within the area.
Whenever HEPA filters are changed, a thorough leak check should be made (50.1).
70.1 General.
70.1 (a) All equipment should be cleaned by dusting, vacuuming, washing or other
means best suited to the equipment involved before being brought into the area.
70. 1 (b) Neither smoking nor eating should be permitted in the area.
70.1 (c) All personnel should wear lint free smocks or coveralls in the area.
70.1 (d) Head covers, covering all areas of head and facial hair, and other appar
should be used as needed to contain loose bits of hair and skin flakes.
70.1 (e) Limit paper entering area. There are special nonshedding papers for
necessary records or paperwork.
70.1 () Use only ballpoint pens for writing. Lead pencils and erasers should not
permitted.
126
Fed. Std. No. 209
70. 1 (g) Hand lotions, creams or soap containing lanolin to tighten skin particles
should be used, as appropriate.
70.1 (h) Solvent contact with hinds should be avoided. as many solvents remove
natural skin oils causing excessive "skin peeling" or flaking.
70.1 (j) Cosmetics and medication which may produce contamination should not be
worn by any personnel. In particular, eye make-up, rouge, face powder, and hair
spray shoLild be avoided.
70. 1 (k) Gloves, finger cots, tweezers, or other clean handling methods and equip-
men- sh..ld be used while working with or handling sensitive parts, to avoid contam-
ina::c of those parts by loose skin or natural skin oils.
70. 1 (1) Exhaust systems for grinding, welding, soldering. machining, or other re-
lated operations should be installed in accordance with the Industrial Ventilation
ar;.. putshed by the American Conference of Governmental Industrial Hygienists
ar.d the ANSI Standard Z9. 2. "The Design and Operation of Local Exhaust Systems."
70. 2 Parts clea.:nr and handling. Before entry into the clean area, all parts.
inr.rimens, materials, etc., should be cleaned, as required, to prevent contamina-
ticn of the roc... To prevent direct transfer of contamination, constant surveillance
cf the estat'shed procedures for handling clean parts and assemblies is recom-
rr. e et
127
Ted. StW. No. 209
70. 3 (a) Limit personnel access to the area to only those persons necessary for the
area operation.
70.3 (b) Shoes should be covered, cleaned or changed before entering area.
70. 3 (c) Use of compressed air or other high velocity gases for blow off or cleaning
operations, except under exhaust hoods capable of carrying residue to exterior of
clean room, should be avoided.
70.3 (d) Room maintenance should be restricted during normal operational periods
to avoid generation of airborne particles. Ifa contamination producing emergency
occurs or routine maintenance is required, normal work should be stopped until the
room is cleaned.
70.3 (e) Janitorial service for a room should consist of a properly supervised,
regAlarly scheduled cleaning program. Periodic or even continuous cleaning service
may be needed to maintain an acceptable airborne particle count level.
70. 3 (f) The use of laminar flow work stations within a nonlaminar flow room is
recommended where feasible. This combination of facilities not only offers areas
with a high degree of cleanliness at reasonable cost, but also upgrades the entire
room by continuous refiltration of air through the BEPA filters in the work station(s)
70.3 (g) The airborne particulate level of a nonlaminar flow clean room is highly
dependent on the amount of effort expended to control contamination. For example,
a properly constructed room housing low particle generating operations in which
strict personnel and janitorial controls are imposed may be able to meet Class
10. 000 requirements. There is usually a point, however, beyond which the appLcI
tion of additional cleanliness measures will not produce significant improvement ir
the cleanliness level.
1NP
Fed. Std. No. 20%
70.4 Laminar flow devices. Periodic checks should be made for proper operation
of the device including proper airflow patterns, velocity, and uniformity; adequate
lighting; properly sealed filters; and properly functioning prefilters. The blower sys-
te.-m shculd be turned on at least 15 minutes before normal operations are started.
During this clean-down period, fixtures and work surfaces should be cleaned.
70.4.1 (a) The protective grille should be lightly vacuumed or wiped during the clean-
down period noted in 70.4.
70.4.1 (b) Obstruction to the airflow should be kept to a minimum in the station, par -
ticulyary upstream from the critical work.
70.4. 1 (c) Care should be taken when locating work stations in an area to avoid air
currerns that could oppose work station airflow and could carry dust into the work
st a: on.
70.4. 2 (a) Operatcns should be graded according to cleanliness levels required for
critica: work. The most critical operations should be nearest the filter bank.
80. 1 General. There are usually several facility options available for achieving
.- %e of the standard air cleanliness classes. The selection of one or some corn-
bination of facilities should be based on the degree of air cleanliness needed, the
amount and type of work to be accomplished, cost, and anticipated future needs.
Table III furnishes guidelines for selecting facilities for the various classes of air
cleanliness. The information provided in Table 11 assumes that the HEPA filters
used in the facilities listed have been properly installed and are free from significant
129
Tod. Std. No. 20%
S0.2 Types of facilities. Figures 1 through 9 represent basic types of clean room
and clean work station designs. Many variations tothese designs which may satisfy
specific needs are commercially available. No attempt has been made to depict all
possible variations.
(3) American Society for Testing and Materials, 1916 Race Street.
Philadelphia. Pennsylvania 19103
130
TA FilE il
FACIlITY Gill1i-:.INI-S 1-11 Alit CLEAN L.lNISS CiASSI S
Type of Facility Clams 100 clanss 10. 000 Class 100. 000
laminar Airflow
Vi rtical flow room Entire work area usually Entire area normally meets Entire area meets require-
Vertical flow curtain unit meets requirements at requirements. ments.
V' cal flow bench normal working height
locations.
Crossflow room Entire work area normally
Tunnel room First work locations nor- meets requirements if par- Entire area normally meets
Wall-to-floor room mally meet requirements. ticle generation, work loca- requiremerits.
Croasflow bench tions and personnel are
reasonably controlled.
Nonlaminar Airflow
In some cases, can be up-
graded to meet requirements
by placing laminar airflow Will usually meet require-
devices (benches, modular ments with strict observa-
units, tunnel rooms or down- tion of rules governing per-
Conventional clean room Will not meet requirements flow curtain units) within the sonnel, operations, gar-
0
under operation conditions. room and continuously filter- menting, and janitorial .
ing the recirculated air. procedures.
Personnel and operation
restrictions and janitorial
maintenance are also re-
quired.
awtIN
a&
AllROWU
NMu suff~cr
RomWARt
olool CRAlI
4- MAKE UP
AIR INIfT
IPR[F II[RtDI
'U,
6
In
Figure 2. Ceiling-to-Floor Airflow Clean Room
z
0
til
- SUN- PtUst#A
SUPPlY P1INIM
0
406
A IR XHAUST
HiPA FIlfR BAN
:r
-51's-pt,"
lIPPLY PIEMIM
=A
AIR INL[T
SCEN
sLOWER
SULP POR T- - - -
CASTOR
All UnAL
136
Figure 6. Nonlaminar Airflow Clean Room
to
tN PAWl,
=a
AIR FtC-R
I
WORK SURF
PR-FltW!R
AIR DIFFUS OR
IEPA FItLTER
PROTEtIvE SCREEN
AIR FLOW
END PANE-
GRATNU
WORK
SUACE -
139
red. ftd. Do. 0sg
140
141
The Proposed Revision of Federal Standard
209B
By: Robert D. Peck
Technical Editor
142
exclusively to cleanrooms and contamination control. Now it is This section also makes clear that the standard does not encom-
used for all sorts of controls and all sorts of environmental pass the physical, chemical, radiological, or viable nature of contami-
factors. nation. While FS209 has not been used as a means of classifying such
contaminations (other than particulate). there were strong arguments
b) The word "cleanroom" has been standardized as a single word made for broadening the revisions to include all types and manner of
rather than as two separate words, thus reflecting that a clean- contaminants. The final decision was not to do so because of the great
room is not just a clean room. difference in contaminants, their effect and control, classification, test
c) The term "Work Station" in 1963 encompassed the only config- for compliance, etc. This is particularly true of classifications and
uration available other than a cleanroom. Now there are many, methods of test for the viable nature of contaminants. Standards are
types of facilities, spaces, etc., which are designated generically needed for these other types of contamination, and IES, among other
as "clean zones". organizations, has groups working on such documents-but not as
d) Most important is the clarification of what FS209 has always Federal Standards. A Federal Standard for classification of viable air
been; that is, a definition of "airborne particulate cleanliness cleanliness (or sterility) is unquestionably needed.
classes": not a compilation of all the requirements for a con- DEFINITIONS
trolled environment.
Perhaps nowhere is the recognition of the advance of contamina-
It is important to realize that these changes would be carried
tion control technology so evident as in the proposed revision of the
throughout the proposed document. No longer would there be
Definitions section of 209B.
separate mandatory and non-mandatory sections; no longer would
there be specific requirements for temperature, relative humidity, Of twelve definitions in the "Definitions" and "Glossary" sections of
airflow velocity, etc. While these, and other properties would be the present Standard, eight have been discarded completely: the
required to be stated for full description of a facility or for testing other four were substantially changed in wording.
procedures, the actual values would be left to the very broad areas of New definitions were added, among them:
application and technology to which contamination control now Airborne particulate cleanliness class
applies. As-built cleanroom
At-rest cleanroom
SCOPE CHANGE Operational cleanroom
Scope and Objective of Current 209B: Unidirectional airflow
"This document establishes standard classes of environmental air Nonunidirectional airflow
control within clean rooms and clean work stations. The objective of Condensation nucleus counter
this standard is to prescribe air cleanliness classes and certain other Optical particulate counter
environmental air conditions required for achieving and maintaining Particle concentration
the levels of environmental cleanliness specified in the product specif-
ications." (Italics by this article's author.)
Scope of Proposed Revision: 1~
"This document establishes standard classes of air cleanliness for 1loom
143
DETERMINATION OF CLASSES VERIFICATION AND MONITORING
The principal purpose of FS209 has been, and is, to "establish The procedures for determining if a cleanroom or clean zone meets
classes for airborne particulate levels in cleanrooms and clean a particular class are much clearer and far more extensive in the
zones." The proposed revision accomplishes this through significant proposed revision than is true of the present 2096. Particularly, there
changes in the classes and in their determination. are provisions for both "Verification" and "Monitoring". While the draft
does not specifically define these terms, the intent is as follows:
Three of these changes are evident in Table I and Figure 1.
a) Verification: to initially and periodically determine the airborne
First, Class 10 and Class 1 have been added to the listed classes
particulate cleanliness class by measurement of airborne parti-
now in 2098. cle concentration.
Second, the particle sizes listed for measurement are increased by
b) Monitoring: to determine if the airborne particulate cleanliness
adding 0.1, 0.2, and 0.3 to the present 0.5 and 5.0
class is being maintained by measurement of airborne particle
Third, an important statement has been added to both table and concentration after verification and during operations.
graph to emphasize that they are presented only for the purpose of
While Verification and Monitoring are separate sections in the pro-
defining classes and are not a representative size distribution relation-
posed revision, the latter is treated only briefly, and the following
ship. This statement is a compromise between those committee mem-
details are from the section on Verification.
bers who wished to eliminate the log/log graph and those who felt
strongly that it was a valuable part of the standard. SAMPLING
In understanding the proposed new provisions for sampling, it is
Table 1. Class limits in particles per cubic foot of size equal to or important first to recognize two new definitions, namely:
greater than particle sizes shown.* a) Unidirectional airflow (commonly known as laminar flow) Air
Measured particle size (micrometers) flowing in a single pass through a room or a clean zone in a
parallel direction."
Class 0.1 0.2 0.3 0.5 5.0
b) 'Non-unidirectional airflow (commonly known as turbulent
10 35 7.5 3 1 NA flow) Airflow which does not meet the definition of unidirec-
100 350 75 30 10 NA tional airflow."
100 NA 750 300 100 NA
NA NA 7 These definitions resulted from the realization that while in 1963
NA
10000 NA NA NA 10000 70 there were only the new "laminar flow" type of cleanroom, and the
100000 NA NA NA 100000 700 previous "conventional cleanroom", we now have "mixed flow", "vec-
tor flow", "nonlaminar flow", "flow control", "dilution control", etc. For
(NA-not applicable) sampling purposes first, and then for general categorization of
*The class limit particle concentrations shown in Table I and Figure 1 airflows, it seemed clearest to simply distinguish flow in a single pass
are defined for class purposes only and do not necessarily repre- in a parallel direction from all other designations of airflow.
sent the size distribution to be found in any particular situation. The sampling provisions of the proposed revision are based, there-
fore, upon these definitions. While this section of the draft is lengthy, it
In addition to these changes in the table and graph, there are other provides these points (realizing that a cleanroom is a specific instance
important provisions. of a clean zone):
a) The requirements for meeting a tabulated class would be a) Not less than two sample locations shall be taken for a clean
changed in two very significant and new ways. zone.
First, while the designation for a class is the number of particles b) At least one sample shall be taken at each sampling location;
listed for the class under the 0.5-micrometer size column, the more than one may be taken at each location, and different
class requirement would be met by measuring at any of the numbers of samples may be taken at different locations.
particle sizes listed in the table for that class. Thus, Class 100 is c) The sample locations shall be uniformly spaced throughout the
labeled from the limit of 100 particles permitted for the class (of clean zone, except as limited by equipment.
0.5 micrometer and greater size). However, Class 100 would be
d) For unidirectional airflow, the clean zone is identified by an
considered met if the particle concentration is measured for
entrance plane perpendicular to the airflow, and immediately
particles of 0.2, or 0.3 micrometer size, and the number of
upstream of the work activity area. The minimum number of
particles so measured falls below the number shown for the size
sample locations shall be the lesser of (a) the area of the
in Table I (750 and 300 respectively for 0.2 and 0.3 size), and
entrance plane (in square feet) divided by the square root of the
second, when a statistical analysis of the sampling locations is
airborne particulate cleanliness class designation, or (b) the
met. This provision is discussed later in this paper.
area of the entrance plane (in square feet) divided by 25.
b) Provision is made for measurement at particle sizes other than
those tabulated, provided that the particle count limit for the e) For non-unidirectional airflow, the number of sample locations
next larger particle size in Table I is not exceeded; counts may shall be uniformly spaced horizontally, and as specified verti-
be interpolated between points, but may not be extrapolated cally, throughout the clean zone, except as limited by equip-
beyond the end points of Table 1. ment within the clean zone. The number of sample locations
shall be equal to the square feet of floor area of the clean zone
c) Provision is also made for defining classes other than those divided by the square root of the airborne particulate cleanli-
tabulated in Table 1.As is true in FS2098, these classes are ness class designation.
defined by the intercept point on the 0.5-micrometer size line of
the log/log graph (Figure 1). f) A minimum volume of air per sample must be taken in accord-
ance with Table 11.
d) The slopes in the log/log graph (Figure 1) have been adjusted
to reflect the same particle size coverage for each class as is Table Ill.Minimum volume (in cubic feet) per sample for the
shown in Table 1. air cleanliness class and measured particle size shown.
Running throughout the proposed new revision is the use of the Measured particle size (micrometers)
phrase: "As Specified". The same phrase was incorporated into 2098, Class 0.1 0.2 0.3 0.5 5.0
but was much less used. In the proposed revision, it indicates that the
user or contracting agency shall specify the degree of control needed 0.6 3.0 7.0 20.0 NA
to meet requirements. Thus, requirements which vary from applica- 100 0.1 0.3 0.7 2.0 NA
tion to application (for example, temperature and relative humidity) 1000 NA 0.1 0.1 0.2 NA
are not specifically designated in the proposed revision, but the 10000 NA NA NA 0.1 3.0
responsibility of the user to designate them is required. 100000 NA NA NA 0.1 0.3
100000 NA NA NA 0.1 0.3
144
Sample Calculations
The data and calculations presented in the following paragraphs are intended to serve as a working example, illustrating the statistical
procedures involved in determination of acceptance criteria for cleanrooms and clean zones. The data and calculations are based upon
testing at 0.3 micrometer measured particle size; and for Class 10, with a class limit of 30 from Table 1.
Tabulation of Particle Count Data
Particle Counts Total Average
No. of Total Particle
Location 1 2 3 4 5 Samples Count Concentrations
A 15 NR NR NR NR 1 15 15.00
B 33 24 9 15 NR 4 81 20.25
C 18 3 12 24 NR 4 57 14.25
D 39 18 9 33 6 5 105 21.00
E 0 27 6 0 NR 4 33 8.25
NR = No reading taken.
Mean of the Averages (M)
M = (15.00 + 20.25 + 14.25 + 21.00 + 8.25 = 78.75) + 5 = 15.75
Standard Deviation (SD)
SD = V (15.00-15.75)2+(20.25-15.75)2+.(14.25-15.75)2+(21.00-15.75)2+(8.25-15.75)2
(5-1)
SD = 5.17
Standard Error (SE)
SE = 5.17 + = 2.31
Upper 95% Confidence Limit (UCL)
For 5 locations, 95% UCL Factor = 2.1
UCL = (2.31 x 2.1) + 15.75 = 20.6
Conclusion
Since the upper 95% confidence limit is less than 30 and all location average particle concentrations were less than 30, the above example
data meet the acceptance criteria for Class 10, although some of the individual particle counts were above 30.
Figure 2. Statistical Analysis
PARTICLE COUNTING METHODS classification of spaces now is added, questions arise as to how many
and where samples should be taken, and what results constitute
Counting particles 5 micrometer and larger: The membrane filter
passing; that is, must every sample be below the class limit, only a
method is used, but the complete procedure is detailed in the pro-
certain percentage, the average of the samples taken, the mean, or
posed revision, rather than just referencing ASTM F25. It is based
what? The "statistical analysis" and "interpretation of data sections of
upon F25, but with elimination of those portions not current or not in
the proposed revision help resolve these problems.
conformance with other parts of FS209-such as relating to only 5
micrometers and larger, sampling locations, frequency, etc. Most sig- The first step would be to select: (a)the cleanliness class to be met;
nificant, the aerosol monitor method and the laboratory open filter (b) the minimum particle size which you want to count; (c) the number
holder method are given equal choice and prominence. of sampling locations required; (d)the number of samples to be taken
at each location. All of these selections must be in accordance with the
Counting particles 0.1 micrometer and larger: Again, the complete requirements under "Sampling", discussed above.
procedure is detailed in the revision rather than simply referencing
ASTM F50, but it essentially follows ASTM F50, F328, and F649, Sampling is then performed, the results collected, and the statistical
taking applicable sections from each; an optical counting instrument interpretation of the data completed (see below). Based upon the
shall be used, itmust count single particles, and it must be periodically results, the cleanliness class would be met -with a 95% confidence
calibrated and properly maintained. limit -if the number in Table 1,for the class and particle size selected,
is greater than:
In the "Definitions" section of the proposed revision, the definition of
"Optical Particle Counter" is worded broadly enough to include laser a) The average particle count at each sampling location (regard-
particle counters, and a definition is included for the "Condensation less of the individual counts at that location), and
Nucleus Counter". Footnotes to the paragraph on "Methods and b) The mean of the averages of all sampling locations.
Equipment for Measuring Airbome Particulate Concentration", com-
bined with these definitions, permit the use of laser particle counters, The specific procedure for the statistical interpretation of data is
condensation nucleus counters, and sedimentation methods, most easily understood by referring to Figure 2, which is similar to an
although there is no explanation, description, or requirements for example provided as an appendix in the proposed revision. With
doing so. reference to this example, the procedure is as follows:
STATISTICAL REQUIREMENT a) Determine:
Completely new in this proposed revision of the Federal Standard 1) The cleanliness class to be met
are the requirements for statistical considerations in the determination 2) The minimum particle size which would be counted
of class, and of meeting a class. 3) The number of sampling locations to be used
The current FS209B has only a definition for particulate air cleanli- 4) The number of samples to be taken for each location
ness classes, without an explanation of how to apply such definition to b) For each sampling location, total the counts for all samples at
a specific situation. Such an approach was reasonable since FS2098 that location; divide by the number of samples at that location;
has been a definition of the cleanliness only of a cubic foot of air. When and thus obtain an average count for each sampling location.
145
This number is termed the "Average Particle Concentration" state-of-the-art in particle management should provide sufficient
for the sampling location. To meet the class requirement, all of pressure to ensure that FS209 does not again languish for thirteen
these must be below the corresponding Table I figure. years as it did prior to now.
c) Add the averages for each location, divide by the number of
locations, and thus obtain what is termed the "Mean of the
Averages". REFERENCES
1. Federal Standard No. 2098, April 24. 1973 and Amendment No. 1. May 30. 197&
d) For each sampling location, subtract the overall "Mean of the 2. Peck, R D . History and Present Status of Contamination Control Standards Pro-
Averages" from the "Average Particle Concentration" for that ceedings of 27th Annual Technical Meeting of the Institute of Environmental Sciences,
sampling location. Square the resulting difference. Los Angeles, California, May 7, 1981.
3. 2098 Revision: Background and Update", Microcontamonation volume 1. No. 1;
e) Add the results from (d)for all the sampling locations and divide JuneJuly 1983.
by a number which is one less than the number of sampling 4. 'Federal Standard 2098 Revision: An Update": Pharmaceutical Manufacturing; vol-
locations. ume 1. No. 4: June. 1984.
f) Take the square root of the result obtained in (e). The result is 5. Proposed Federal Standard No. 209C, Proposed Draft No. 14 of April 1, 1986.
the "Standard Deviation".
g) Divide the Standard Deviation by the square root of the number
of sampling locations. The result is the "Standard Error". NOTE TO THE READER:
h) Multiply the Standard Error by the 95% UCL Factor. The 95% The Institute of Environmental Sciences emphasizes that the mate-
UCL Factor is a number related to the number of sampling rial presented herein has only been proposed for adoption by the
locations as follows: General Services Administration (GSA). The final version will come
No. of Locations 2 3 4 5-6 7-9 10-16 17-29 >29 from GSA; thus, the material in this article cannot be used in place of
the current Federal Standard 2098.
95% UCL Factor 6.3 2.9 2.4 2.1 1.9 1.8 1.7 1.65
i) Add the product of the multiplication in (h) to the Mean of the
Averages obtained in (c).
The Institute of Environmental Sciences wishes to thank all mem-
j) Compare the result to the number listed in Table I for the bers of the RP50 Working Group for the outstanding job done on the
particle size measured and the class desired. If the result is 2098 Revision. The members of the committee include:
less than that number, the test area is below the class limit and
the class meets the statistical requirement. John A.McDonough, Mine Safety Appliances Company
Richard Beeson, Lockheed Missiles amd Space Company
William Brader, CUH2A
APPENDICES George H. Cadwel, Flanders Filters, Inc.
Frederick J. Carleton, Pfizer, Inc.
In the present FS209B, the appendix is an extensive compilation of Douglas W.Cooper, IBM Watson Research Center
all sorts of nonmandatory information regarding cleanrooms, how to Warren Dexter, Environmental Air Control Inc.
operate and maintain them, and how to obtain the cleanliness Gabriel Danch, Environmental General Company
required. When FS209 was written in 1963, much of the information E. 0. Dixon, Air Control, Inc.
was new and of value to persons just becoming acquainted with Harold Fitch, IBM Corporation
contamination control. Now, however, much of the present appendix James Flannery, ENV Services
has been outdated by the advances which have been, and continue to Clifford Frith, Anatel Instrument Corporation
be, made in our industry. In addition to the resulting confusion is the Fred Gerbig, Gerbig Engineering Company
fact that in many cases, inspectors have not honored the "nonmandat- Humphrey Gilbert, Consultant
ory" status of the appendix, even in cases where the provisions are Roger T. Goulet, Cambridge Filter Corporation
obviously not practical or appropriate. Robert 1.Gross, Medical Repair Labs.
In the proposed revision, the appendices are limited to one detailing Patricia Keady, TSI, Inc.
the requirements of the membrane filter method of sampling, one J. Gordon King, Henningson, Durham and Richardson, Inc.
describing the same for the particle counter method, one providing an Alvin Lieberman, HIAC/ROYCO
example of the use of the statistical interpretation of data, and one Wendell J. Lough, Weber Technical Systems
listing publications which are sources of information about clean. R. Claude Marsh, CMCO Intemational
rooms and contamination control. Richard Matthews, Filtration Technology
Charles F.Mattina, Kimberly-Clark Corporation
Robert L. Mielke, Monsanto Research Corporation
WORK REMAINING Susan Morris, Merick, Sharp and Dohme
From the foregoing discussion, it might appear that the proposed Stacey Nakamura, Johnson Space Center
Robert D. Peck, Controlled Environment Equipment Corporation
revision of 209B has resolved every outstanding issue to come before
the RP50 working group. However, as with every standards docu-
Stephen J. Pijar, HHS PHS FDA Center Radiological Health
ment, the finished product is the result of numerous compromises, Ludwig Pulaski, IBM Corporation
recognized omissions and still unresolved technical questions which
R.D. Rivers. American Air Filter
John H. Robertson, Upjohn Company
need continuing study.
Ron Skowronski, Mine Safety Appliances Company
It can also be anticipated that the use of the proposed new docu- Robert L. Sorensen, U.S. FDA.
ment in the field will highlight some new and some possibly still Brian Stembal, Travenol Laboratories
existing technical deficiencies as well as both grammatical and inter- Konrad H. Stokes, University of Arizona
pretive weaknesses which should be addressed by the RP50 group in David C. Swinehart, Delco Electronics Division
continuing studies leading to future revisions of the document. Marvin Tillery, Los Alamos National Laboratories
The importance of contamination control as a major component of Jeannie West, Climet Instrument Company
high technology industry and medicine, and, the rapidly changing Richard A.Yeich, AT&T Technology Systems
146
References
147
Djordjevic, Walter. "Unwanted Physical Vibrations: A Source Of Contamination In
Semiconductor Fabrication Facilities." MICROCONTAMINATIONS,
February 1985. pp. 43-74.
Dooley, John. "Interfacing Between Robots, Transport Systems And Semiconductor
Processing Equipment". 1986 PROCEEDINGS-INSTITUTE OF
ENVIRONMENTAL SCIENCES. pp.395-396.
Edmark, Karl. Quackenbos, George. "An American Assessment of Japanese
Contamination-Control Technology". MICROCONTAMINATION, October 1984.
pp. 47-53.
Fujii, Shuji; Hayakawa, Ichiya; Iwase, Kazuo. "Environmental Integration
Technology Today For a Quality Tomorrow: Studies On Design Theory Of Laminar
Flow Type Clean Rooms. I.E.S. 1984 Proceedings, 30th Annual Technical Meeting.
pp. 107-111.
Goodfellow, Howard D. "Advanced Design Of Ventilation Systems For Contaminant
Control". 1985. Elsevier Science Publishing Company Inc.
Gupta, Praveen. Sheng, David. "Environmental Characterization Of A VLSI
Cleanroom". MICROCONTAMINATION, October 1985. pp. 62-68.
Hamberg, 0.; Shon, E; "Particle Size Distribution In Clean Rooms".
PROCEEDINGS-INSTITUTE OF ENVIRONMENTAL SCIENCES, 1984.
pp. 14-19.
Hoenig, S.; Ph. D. "Clean Rooms and Air Cleaning: Better Operation At Lower Cost
Through Applied Science". ASHRAE TRANSACTIONS 1986. Part 1B. American
Society Of Heating, Refrigerating And Air-Conditioning Engineers. pp. 261-264.
Ichiya, Hayakawa; Fujii, Shuji; Kim, Kwang Young". "Clean Room Supervision
And Operation In Japan And Particle Technolgy". MICROCONTAMINATION
CONFERENCE PROCEEDINGS. 1984. pp. 63-67.
148
King, Gordon. " The History Of Cleanrooms". ASHRAE TRANSACTIONS 1986.
Part 1B. American Society Of Heating, Refrigerating And Air-Conditioning
Engineers. pp. 299-308.
Leary, Wm. A.; Tsai, D.H. "Metering Of Gases By Means Of The ASME -Edged
Orifice With Flange Taps". July 1, 1951. Sloan Laboratory For Automotive And
Aircraft Engines. Massachucetts Institute Of Technology.
Lieberman, Alvin. "Particle Control And Air Cleanliness For State Of The Art
Microelectronics Manufacturing". MICROCONTAMINATION, February 1986. pp.
29-38.
Little, James. "Clean-Up Characteristics And Double Filtration". HEATING,
PIPING & AIR CONDITIONING. April 1966. pp. 111-115.
149
Salvo, Nancy. "Practical Considerations In Contamination Source Control: How To
Make The Best Of A Retrofit Cleanroom". MOCROCONTAMINATION, August
1985. pp. 43-60.
150