TOPIC SERIES

Cleaning & particle control

Poor pump-down is usually outgassing, not leakage; and particles often come from handling, not the wall.

1 article in this series

"It is clean but it still will not pump down" is the most common puzzle on a vacuum site. The answer usually lies in surface chemistry: water follows a clear decay law and hydrogen hides in the damaged layer, and neither treatment substitutes for the other. Particle control, meanwhile, is often limited not by the chamber wall but by wafer handling and the flow field.

What this series covers

  • Water outgassing falls as 3×10⁻⁹/t (hours): ten times the pumping time gives one order lower outgassing and pressure.
  • Baking removes four orders of water but only one of hydrogen; hydrogen needs vacuum annealing or long pumping.
  • Cleaning has three layers: grease and particles, oxide and contaminated films, and the damaged layer. Doing only the first is the most common shortfall.
  • A single elastomer seal takes the ultimate from the 10⁻¹⁰ range to the 10⁻⁸ range — the quantitative case for all-metal construction.
  • Criteria must be reproducible: pressure levelling, post-cooling recovery, and an RGA 18/28 peak comparison. All three, not two.

FAQ

The chamber has been cleaned and still will not pump down. What to check first?

Separate outgassing from leakage: close the valve, record the pressure-rise curve, and read the RGA. A dominant mass-18 (water) peak that decays over time indicates outgassing; peaks rising together without decay point to a leak. Getting the order wrong costs a needless teardown.

When is all-metal (CF) construction worth it?

When the target ultimate is below 10⁻⁹ mbar and repeated baking is required. Elastomer seals cap the ultimate around 10⁻⁸ and limit bake temperature to the material's ceiling. If the process only needs the 10⁻⁶ range, all-metal is over-engineering.

Request a QuoteBack to Articles

WhatsApp us