7.6.11 Helium Leak Detection Techniques

There are four different approaches to helium leak detection. They are displayed in Table 7.17. A graphical representation of these techniques is shown in Fig. 7.59 through Fig. 7.61.

Some of the other leak detection methods explained earlier in this section are based on these techniques. Methods such as bagging parts of a vacuum system and filling the bagged parts with a probe gas are similar to the outside-in technique. However, it is not possible to quantify a leak with this approach. Alternatively, filling a container with gas and either submerging it in a liquid or covering it with a bubbling solution is similar to the detector-probe technique.

What makes the helium leak detector unique is that it can be calibrated against a standard leak. Then, any measurements made (using the inside-out or outside-in techniques) can provide a quantifiable number. This calibration feature is usually more important to the manufacturing industry than to the needs and demands of the laboratory. For example, when the top of a soda pop can is scored to provide its "pop-top" feature, the depth of the scoring is very important: If the scoring is not deep enough, the top will not work as advertised; if it is too deep, the gases within the container can diffuse out, leaving a flat, nonbubbly product. So, by periodically checking to see that the amount of helium diffusion past the scored top is within tolerable limits, the canning industry can ensure that a good product will be delivered to customers, and the customers will be able to obtain access to the beverage. The canning industry does not care where the leak is—it already knows. What it does not know is if the leak is within tolerable limits or not.

When examining a vacuum system for leaks in the laboratory, calculating the size of a leak is often unnecessary for two reasons:

1.The techniques used to find the size of a leak entail the use of the insideout or outside-in techniques, both of which do not allow indication of

To pump

Stopcocks

Toy balloon (filled with helium)

Detector-probe Technique

Fig. 7.59 In the detector-probe technique, the tested piece is filled with helium and the sniffer probe "sniffs" the areas in question to detect leaks. Always "sniff" from bottom to top.

Table 7.18 Permeability of Polymeric Materials to Various Gases"

std cm3 cm/cm2 sec bar x 108 (note bar = 104 dynes/cm2)

a Edited selection from Introduction to Helium Mass Spectrometer Leak Detection, Table 6.1, published by Varian Associates, Inc. ©1980, p. 49, reproduced with permission.

* Data not provided

ural rubber, Buna-N, and then neoprene. The permeability of several materials is included in Table 7.18.

One additional complication with elastomers has to do with the connection between a vacuum system and a helium leak detector. If it is possible, connect the two with a nonelastomeric connection such as a stainless steel bellows. If this connection is not possible, the best elastomeric connection that you can use is a high-vacuum, thick-walled (!/2 in I.D., 3/16 in. wall) vinyl tube of the shortest length possible. Do not use rubber or neoprene tubing. If you have several systems that may need leak detection, it is unlikely that one length of tube will accommodate all systems. It is best to have several connecting tubes, one for each system (or several systems) being tested, to allow you to select the shortest one that can be successfully used.

The actual connection between the vacuum system being tested and the helium leak detector can be made at a variety of points. One approach is to attach the connecting tube to the foreline of the diffusion pump where the fore pump would normally be connected. This attachment has the advantage of being physically at the end of the system, and therefore it is less likely to encounter confusing signals as you examine the system from one end to the other. The disadvantage is that you need to remove, and then later replace, the fore-pump hose. In addition, this attachment can only be made if the system being tested is small (i.e., less than 5 to

Sniffing probe

Leak

\

Glass tube with leak being checked

Control valve

Fig. 7.64 Using a plastic tube to increase the sensitivity of a sniffing probe.

the amount of maintenance required for the leak detector is at a minimum due to less potential contamination. (There is, of course, an amusing contradiction here: if you require a leak detector, you are not operating at the highest possible vacuum.)

Most helium leak detectors will not operate with pressures above 10"4 torr to 10" 5 torr. At these greater pressures, the main element to the mass spectrometer will burn out. Fortunately most, if not all, helium leak detectors have various safety check mechanisms that automatically shut off the current to the main filament if the pressure goes above a set limit. So, you must depend on alternate leak detection methods, or use the detector-probe technique to discover large leaks. Once large leaks have been discovered and closed, you can concentrate on the smaller leaks that can be found with the tracer-probe technique.

The detector-probe technique is, at best, about 10% as sensitive as the tracerprobe technique. It can be made more sensitive by placing a greater pressure of the probe gas within the system. This greater pressure cannot be used on glass systems for fear of blowing out stopcock plugs or standard taper joint pieces. One technique for introducing helium to your vacuum system (with a moderate pressure) is simply to fill a balloon with helium and attach it to the vacuum system. The pressure within a helium balloon is not great enough to do any damage to a glass vacuum system. However, be careful how you "sniff around the balloon to avoid misreading helium permeating from the balloon walls. Using a Mylar balloon will provide less permeation through the walls, but it is more difficult to attach to the vacuum system.

By placing a small vinyl tube over the end of the "sniffer probe" (see Fig. 7.64), it is possible to press the tip of the probe over the area of the possible leak. This method closes the leak off from ambient air, which both increases the sensitivity at this location and helps to precisely pinpoint the leak site. If you hold the vinyl tube too long near the helium, it will become saturated, and your "sniffer" will be reading the "leak" from the plastic on the nose of the probe. If this saturation occurs, have another plastic tube (that has been kept away from any helium) ready to slip on. A saturated tube will lose its helium in several days.

®Q)

Helium leak

detector

Fig. 7.65 A suggested pattern for spraying helium on a vacuum system when using the tracer-probe technique. The pattern is a compromise between proximity to the leak detector and spraying high before low areas.

If you ever wish to change from the detector-probe technique to the tracer-probe technique, you must remove all the helium already in the vacuum system by running the leak detector's vacuum system and/or the pumps of the vacuum system being tested for a period of time. Otherwise false and/or artificially high readings may affect any serious or useful leak testing. A dry nitrogen purge is recommended to help sweep the helium out of the system.

The spraying pattern used with the tracer-probe technique is important. Improper spraying can significantly affect the amount of time and quality of the leak detection. If there is too much ambient helium in the air (caused by sloppy helium spraying), false readings can result. You can blow helium from a pipette tip or from a blunt hypodermic needle inserted in the end of a tube connected directly to a regulator on a helium compressed gas tank. To control the flow rate, you will need to adjust the needle valve on the regulator. To see the amount of helium spraying out of the tip, place the pipette tip into a beaker of water. The flow rate can then be set to allow for a slow but steady stream of bubbles.

The location of where to spray is a combination of (1) spraying the system parts close to the helium leak detector and then moving further out and (2) spraying the highest parts of the system and then moving on down toward the floor. The reason for the former is to keep cleanup time and response time to a minimum. Otherwise, you must wait for all the helium to be removed before you can check for leaks closer to the detector. The reason for the latter is because helium rises in air. If you spray down low first, the helium that drifts up may get sucked into an unknown leak in a higher section of the vacuum system, thus creating confusion as to whether the leak is in the area you are studying or anyplace above the area you are studying. See Fig. 7.65 for a suggested spraying pattern.

Response time can be easily explained with a hot water analogy: When you turn on the hot water in your shower, water immediately comes out of the shower fixture, but there is a time lag before hot water arrives. This lag time is simply a function of water pressure, the amount you have the valve open, the pipe diameter, and the distance between the hot water tank and the shower. It is important to keep in mind this delay factor, or response time, when operating a helium leak detector.

For example, say your setup takes one minute from the time you introduce helium to a leak until it is acknowledged by the helium leak detector. In addition, say your spray probe is being moved at four inches per minute (which is pretty slow if you think about it). You will therefore be four inches away from a leak when it is first identified. Generally, moving the spray probe at a rate of one foot per minute provides an adequate time response, which of course can be made faster or slower as conditions warrant. However, if you know the response time, you will know about how much to back up to retest the area in question to allow pinpointing the specific leak spot.

The response time for a helium leak detector is "the time required for a known leak rate to be indicated on the leak detector from zero to 63% of its maximum equilibrium level." Of course the helium leak detector must be calibrated with a standard leak so that the maximum reading can be properly set on the leak detector. The response time is dependent on the quality of vacuum, the size of the leak, physical barriers (such as constrictions, bends, or traps) between the leak and the leak detector, and the proximity of the leak to the leak detector. Because of so many factors, there can be no set specific response time.

The cleanup time is the length of time for the detection signal to stop indicating that it is sensing a leak. More specifically, it is "the time required after removal of a tracer gas from a leak for the initial leak indication to decay 63% of the indication at the time the leak was removed."88

Cleanup time is almost always longer than response time because of the difficulty in desorption of helium from a vacuum system, compounded by helium permeation into porous materials. So, when using the tracer-probe technique do not overspray helium onto your system. The more helium that enters your system, the more that can permeate into porous materials and the longer the cleanup time.

Liquid-nitrogen cold traps stop condensable vapors* from traveling between mechanical pumps, diffusion pumps, and the rest of the system. They also protect the helium leak detector from possibly contaminating materials (such as siliconbased diffusion pump oil).

Any coating that develops on the ion source of the helium leak detector will hinder operation. This coating is not easy to prevent because during operation, the ion source behaves like a getter-ion pump, drawing materials to its surface. Therefore anything that can be done to protect the ion source without hindering normal operation is desirable.

During operation, maintain the liquid nitrogen at a constant level to prevent the evaporation of condensable vapors that could in turn collect on the ion source. Use the trap-filling method suggested in Sec. 7.4.3.

Helium is a noncondensable gas. If you are using an oxygen or halogen probe gas, you have a new complication because these gases are condensable. Thus, using a cold trap will likely prevent the leak detector from detecting the tracer gas. Because a liquid nitrogen cold trap would not be practical in this circumstance, consult the manufacturer for assistance on how to protect the electronics of your leak detector.

During operation, if a heavy buildup of vapors collects on the cold trap, an incidental backlog of probe gas can develop, thus causing a rise of background probe gas noise, which in turn slows down the speed of detection ability and decreases the sensitivity of the leak detector. If this situation develops, close the valves to the vacuum system being checked, turn off all gauges within the leak detector,* and let the cold trap heat up. Meanwhile, the pumps should remain on to draw the condensable vapors from the leak detector. Do not run the mechanical pump with the diffusion pump off under a no-gas load (nothing coming in) condition because mechanical pump oils will backstream into the diffusion pump.

The following are several extra general maintenance tips:

1.If you need (or desire) to run the mechanical pump of a helium leak detector with the diffusion pump off, the best means to protect the diffu-

sion pump oils is with a dry nitrogen gas purge. This purge will prevent mechanical pump oils from backstreaming into the diffusion pump.f The leaking dry nitrogen maintains a low-level laminar flow toward the

mechanical pump. The nitrogen (leak) pressure should be maintained between approximately 1 and 10~2 torr.

2.Never use anything but hydrocarbon oils in the diffusion pumps of helium leak detectors. Silicon oils can deposit insulating layers on the ion sources of the mass spectrometers and ion gauges. However, if hydrocarbon oils are allowed to migrate to the guts of the leak detector's mass spectrometer section, they can be cracked by ions and electrons. The cracked hydrocarbon remains can deposit on the filaments and other system parts, causing a loss of sensitivity and resolution. Admittedly the effects of hydrocarbons are not good, but they are better than the effects of silicon oils.

3.Within a helium leak detector, there are valves that isolate the mass spectrometer and gauge sections, the diffusion pump, and the trap from the rest of the system. These valves are used to prevent contamination or damage to these sections when cleaning, adjusting, or venting is required during use. Find and use them.

4.When shutting down a system, it is best to remove a cold trap while it is still full of liquid nitrogen. This removal procedure will help decrease the amount of material that could pass into the leak detector, or the leak detector's pumping system, after the system is turned off. The process for removing a trap is as follows:

(a)Be sure that all protective valves around the cold trap are closed

This procedure should always be done when running any vacuum system without the use of traps because of the damage some condensable vapors can do to vacuum gauges.

f On a regular vacuum system, this can be done by either (a) separating the diffusion pump from the rest of the system by stopcocks or valves or (b) the use of a nitrogen cold-trap between the mechanical and diffusion pump.