What is Gas Load?
When discussing pressures and pumping, we are really speaking of the effects of gas-phase molecules in the gas-phase, which are the only ones we can measure or pump. However, if we could remove all gas-phase molecules instantaneously from a vacuum vessel, the result would not be zero pressure. Molecules are continuously entering the gas phase from various sources which can be summarized as:
- Real leaks at welds, flanges, or porous construction materials
- Virtual leaks such as trapped volumes at welds, screw threads, or two mating surfaces
- Outgassing which includes gas or vapor
- Desorbing from the wall surfaces
- Diffusing from the wall matrix
- Evaporation of materials with high vapor pressure
- Permeation through elastomeric gaskets
- Permeation through the glass or metal walls
- Backstreaming gases from the pump
- Backstreaming oil vapor from an oil-sealed pump
- Backstreaming condensable vapors (e.g., solvents) coming out of solution in the pump oil
- Desorbing gas from a saturated trap
- Desorbing gas from a cryogenic trap with a falling cryogen level
- Deliberately injected gas required by the “process”
The rate at which molecules enter into the chamber's gas phase from all these sources is called the chamber's gas load.
In a tight vacuum system, a major contributor to gas-phase molecules is wall desorption. For example, a spherical chamber, one foot diameter at 1 x 10-6 torr with its inner wall covered with one monolayer of adsorbed water vapor, has ~7000 times more molecules on the surface than in the gas-phase. And it is a reasonable assumption that a real chamber at 1 x 10-6 torr has a surface coating thicker than one monolayer.
Often, the sources listed above are lumped together and the whole process is called outgassing. In one sense, this is justifiable since no matter what the source, all molecules — inside a vacuum chamber at low pressure — eventually hit a surface, stick, and later desorb. (The interval described by “later” encompasses nanoseconds to years.)
At reasonable vacuum pressures (say 1 x 10-4 torr and lower), a molecules desorbing from a particular spot on a surface will travel, without collision, across the chamber and hit another part of the surface where it will again stick. Multiply that event by billions and you get some idea of conditions at the molecular level inside a vacuum system.
From there, it is not hard to project that a fraction of those desorbing molecules, instead of hitting other walls, will head for the pumping mechanism and be “lost” from the system. Let's say the fraction is 1/1000 of all desorbing molecules. As the number of desorbing molecules increases (that is, as outgassing rate increases, say, during bakeout), the fraction remains constant. So, the absolute number of molecules heading for the pump increases but is still 1/1000th of the total.
The actual fraction is probably nothing like 1/1000. It depends, mostly, on the ratio of total surface area to the pump mechanism's area (as seen from the chamber). But is also depends on shape factors — a spherical chamber and a long, skinny, “parallel plate” chamber will have quite different fractions going to their pump mechanisms.
But in any particular chamber, at any (low) pressure, the fraction heading to the pumping mechanisms remains constant and is called the gas load. Wait a minute! Does this mean that molecules desorbing from “here” and zipping to (and sticking) “there” are not part of the gas load? Well, yes. If they don't enter the pumping mechanism or the gauge, you know nothing about them. Of course, you infer their presence by noting the fraction that reaches the gauge. The higher the number of “zippers” — the higher the absolute number that are detected by the gauge and the higher the absolute number that reaches the pump.
As a final concern, note that for practical vacuum chambers, gas load decreases slowly over time, but never reaches zero.
Gas Load Units
Gas load is a mass flow rate and is measured in units of volume × pressure per unit time, such as:
Torr.liters per second: T.L/s
mbar.liters per second: mbar.L/s
Pascal.cubic meters per hour: Pa.m3/h
Torr.liters per minute: T.L/m
std.cubic centimeters per minute: sccm