7.2.7 How to Make (and Maintain) a Vacuum

about creating a vacuum, until you remove the adsorbed (surface concentration) and absorbed (material penetration) gases and vapors, you do not have an "empty" system.

So, say that you take a glass, pour out the contents, dry the walls, and bake out the glass so it is truly empty. The question is, Will it now remain empty until water is poured back in? As far as vacuum science is concerned, no. As soon as the glass is exposed to the atmosphere at room temperature, the walls will resaturate with water vapor, and the glass will no longer be empty. To maintain a glass as "empty," it must be isolated from the atmosphere. Otherwise you must repeat the drying process.

Unfortunately, glass vacuum systems cannot be "baked out" (as is done with metal vacuum systems) to remove adsorbed water. Baking is likely to damage stopcocks or rotary valves; or if the baking temperature is sufficiently high, the glass walls of the system itself. Thus, glass vacuum systems are not practical if baking your system is required.

What typically happens with a glass vacuum system is that first a mechanical pump removes a great deal of the "loose," or "free," gas particles. Then, greater vacuum is achieved with the combination of a diffusion pump (or similarly fastpumping unit) and traps that remove or bind up the various vapors within the system (for example, oil, mercury, and water). The only way a system can achieve a vacuum lower than 10"6 to 10~7 torr is if the pump can remove water vapor faster than the water vapor can leave the walls. Most diffusion pumping systems cannot achieve this goal, but even if they could, there is such a substantial amount of water vapor within the glass that, unless the walls are baked, a better vacuum cannot be obtained.

Aside from adsorbed gases, there are potential leaks in any vacuum system that must be dealt with or a viable vacuum cannot be created. Leaks and leak detection are dealt with in Sec. 7.6.

7.2.8 Gas Flow

As one is trying to remove the various gases from a vacuum system, whether the system is a bell jar or a complex collection of tubing with valves or stopcocks all interconnected, the gases must pass through tubing of various types and sizes between the system, the pumps, and the outside world. Depending on the size of the connecting tubing, the length of the tubing, and the vacuum at any time, there can be significant performance changes throughout the system. This all has to do with how gases flow through tubing of different sizes and at different pressures.

Fig. 7.3 Turbulent flow is primarily a gas-gas interactions.

Fig. 7.4 Viscous gas flow.

There are three basic types of gas flow: turbulent, viscous, and molecular. The type of flow passing through a given system is dependent on both the mean free path (MFP) of the molecule(s) and the diameter of the container (tube) through which they are flowing. A useful formula when talking about MFP is the Knudsen number (Kn), defined in Eq. (7.6).

When a system is first brought into vacuum conditions from atmospheric pressure, the flow is turbulent (see Fig. 7.3) and is not unlike the chaos seen in whitewater rapids. At this time, the MFP is approximately 9 x 10~7 cm, which is considerably smaller than any tube the gas is likely to pass through. The great preponderance of interactions (that is, when a gas atom or molecule interacts with something else) are gas-gas, which means there is a greater likelihood that a molecule of gas will hit another molecule of gas than it will hit a wall.

As the pressure decreases, the next type of flow transition is called viscous flow. The nature of this flow is complex and is dependent on flow velocity, mass density, and the viscosity of the gas. Viscous flow is similar to the flow of water running down a slow, calm stream—the flow is fastest through the center of the tube, while the sides show a slow flow and there is zero flow rate at the walls (see Fig. 7.4). The gas interactions in viscous flow are gas-gas and gas-wall; in other words, a molecule is equally likely to hit a wall than another molecule.

When Kn of the system is <0.01, the flow is probably viscous. The transition between viscous and molecular flow is fairly straightforward: When 1 > Kn > 0.01, the flow is in transition to molecular flow; when 1 < Kn, there is a molecular flow.

Fig. 7.5 When the mean free path is fairly short (relative to the inside diameter of its container), a gas molecule to more likely hit another gas molecule than the walls of the container. This situation is known as a gas-gas interaction.

Fig. 7.6 When the mean free path is longer than the diameter of the container of the gas, gas-wall interactions will predominate and you will have molecular flow.

Typically, the first phase of molecular flow is gas-gas. That is, a molecule or atom is more likely to interact with another molecule or atom than the wall (see Fig. 7.5). As the pressure continues to drop and the mean free path increases, gaswall interactions become the predominant type of gas flow.

One of the more interesting characteristics of gas-wall interactions is that not only are the gas reflections not specular (mirror-like), but the molecule or atom may go back into the direction that it once came from (see Fig. 7.6). This bounceback is partly because, at the molecular level, wall surfaces are not smooth but very irregular. In addition, there is likely to be a time delay from the time a molecule hits a wall to the time it leaves the same wall. At the molecular level, when a molecule hits a wall surface, instead of reflection (like a billiard ball), the process is more likely to be adsorption (condensation). When the molecule leaves a wall surface, the process is desorption (evaporation), thus explaining why there is the random movement and time delay for molecular reflection.

The gross movement of molecules in a high-vacuum state is a statistical summation of the parts. For example, say you have two containers that are opened to each other through a small passageway (see Fig. 7.7). One of them (A) is at 10"3 torr and the other (B) is at 10"5 torr. The movement of molecules in both is completely random, but the net movement will be from A to B. There will be molecules from B that will find their way into A, even though the pressure is greater in A. But again, the net molecular movement will be from A to B. Once the net pressure of 10~4 torr is achieved, there will still be movement between the two containers. Eventually, according to the first principle of gases presented earlier, there will be the same number of molecules in B as in A. There will always be a greater number of A molecules than B molecules, but the number of molecules on either side of the system will eventually be the same.

The time necessary for the molecules to travel from A to B (or B to A) depends on the abilities of molecules to pass through a tube. This passage is dependent on three things: the pressure difference between A and B, the diameter of the connecting tube, and the length of the connecting tube.

A B

Fig. 7.7 The movement of molecules from one vacuum container to another (of greater vacuum) is statistically random.