How to decide which ISO class each area of a nanofab needs, why Dhaka's climate reshapes the air handling design, what gowning and construction materials actually require, and the running cost universities consistently leave out.
Two things go wrong in most Bangladeshi cleanroom proposals, and they go wrong together. The class is over-specified, because a higher number reads better to a review committee, and the air handling is under-specified, because it is a mechanical services line item nobody at the table owns. The result is a room certified once at handover, running two classes worse within a year, with an electricity bill arriving in a department that never agreed to pay it.
This article covers how to choose the class for each area, what the air handling has to do in Dhaka specifically, how gowning and construction decide whether the class survives, and the running costs that belong in a departmental budget from day one.
ISO 14644-1 classifies air by the number of particles above a given size in a cubic metre. Each step down the class number is a factor of ten fewer particles. The classes most relevant to a nanofab, with the older US Federal Standard names that vendors still quote:
| ISO class | Particles 0.5 micron and larger per cubic metre | Old name | What it is used for in a nanofab |
|---|---|---|---|
| ISO 5 | 3,520 | Class 100 | Lithography bay, resist spinning and exposure, critical device work |
| ISO 6 | 35,200 | Class 1,000 | Lithography and deposition where features are around a micron |
| ISO 7 | 352,000 | Class 10,000 | General process area: etch, deposition, wet benches, metrology |
| ISO 8 | 3,520,000 | Class 100,000 | Gowning room, sample preparation, packaging, service chase |
Two points get missed in tender documents. First, a class means nothing without an occupancy state. ISO 14644 defines as-built, at-rest and operational, and a room holding ISO 6 at rest with nobody inside can be ISO 8 with a class of students working in it. Specify the state, the sample points and the test method, or acceptance becomes an argument. Second, the class is set by the smallest feature you pattern and the yield you need, not by ambition. A facility patterning 2 micron features for teaching needs ISO 7 across the floor with an ISO 5 or ISO 6 lithography bay. Building an entire room to ISO 5 and running it as a teaching lab is the most expensive available way to obtain ISO 7 performance.
The honest design is mixed class: a large ISO 7 process area, a small ISO 5 or 6 island around lithography under unidirectional flow, an ISO 8 gowning suite, and an unclassified service chase behind the tools.
Cleanliness comes from filtered air changes: supply through HEPA or ULPA filters in the ceiling, return low on the walls or through a raised floor, and enough changes per hour to dilute what people and processes generate. ISO 8 needs modest air change rates, ISO 7 several times more, and ISO 5 generally needs unidirectional flow across much of the ceiling rather than a change rate at all. Those ratios drive fan power, duct sizes and plant space, so a class decision made in a meeting becomes structural and electrical consequences on the drawing.
What is different here is the latent load. Outside air in Dhaka through most of the year is warm and close to saturated. Every cubic metre of fresh air brought in has to be cooled well below its dew point so that water condenses out, then reheated to reach the room set point. That dehumidification duty, not the temperature difference, is where the electricity goes, and it is routinely sized against a European or a temperate design day.
People are the dominant particle source in any cleanroom. Excellent air handling with careless gowning will not hold a class, and extra filtration does not compensate. The gowning suite is where the class is maintained.
Then settle the question that sets the recurring cost: who launders the garments? Cleanroom laundering is a specialised service, and where it is not available locally the facility runs on disposable coveralls, hoods, boots and gloves. That is an imported consumable with a lead time, and it belongs in the annual budget at a realistic weekly rate. A facility that assumes reusable garments and then cannot have them cleaned properly ends up with garments that shed particles rather than contain them.
A cleanroom is not a room, it is a machine that runs continuously. Switch it off overnight and you lose hours of recovery time and expose tools to humidity swings. The recurring cost is predictable, and it belongs in the departmental operating budget, not in the project budget that ends with the grant.
Labs go dark after a grant closes far more often because nobody owned the electricity bill than because a tool failed.: Vvon engineering team, on research facility planning
One more Bangladesh specific point, and it is not a small one. Put the air handling and the exhaust fans on the standby generator, not just the instruments. The common pattern is a UPS on the electron microscope and nothing on the air handling, so load shedding leaves a room with no pressure cascade, no exhaust and chemistry on an open bench. Exhaust and pressurisation are safety systems, and they belong at the front of the generator schedule.
For a department building its first facility, a modest room that runs well beats an ambitious one that cannot be maintained. A practical shape: an ISO 7 process area sized for the tools you have funding for, one ISO 5 or ISO 6 lithography bay under unidirectional flow, an ISO 8 gowning suite with a proper bench and airlock, a service chase behind the tool wall, and a bay's worth of floor and ceiling capacity left deliberately empty for the next grant.
Softwall enclosures and laminar flow benches have a legitimate place in that plan, particularly for sample preparation and for optical work that needs cleanliness rather than a classified room. They cannot substitute for a pressurised suite once wet chemistry and process gases are involved. Use them to start, and design the shell so hardwall bays can be added later without demolition.