Leak Testing with Hydrogen Tracer Gas in Vacuum

INFICON is often approached by users of vacuum leak detection systems who want to switch from helium to forming gas (5% hydrogen, 95% nitrogen). Using a mass spectrometer and simply switching from helium to forming gas results in a loss of sensitivity in the order of 10,000 times. This article will explain the details of this limitation.

In vacuum leak detection, the smallest detectable leak rate is determined by two characteristics:

  • The smallest detectable leak rate of the leak detector used
  • The background concentration of the tracer gas in the vacuum chamber as measured by the detector

The most sensitive hydrogen detector for vacuum applications is the mass spectrometer. In a clean lab test, using very long pumping times there is only a marginal difference in sensitivity for helium and hydrogen respectively. The limiting factor for leak testing in vacuum with hydrogen is the background concentration.

Three factors influencing the hydrogen background concentrations:

  • Natural concentration in ambient air.
  • Rate of outgassing from surfaces
  • Rate of thermal decomposition (cracking) of hydrogen compounds on the spectrometer filament.

Because natural concentration of hydrogen in air is about 0.5 ppm compared to 5 ppm for helium, the natural background should not be a problem. However, when working in the mbar range of pressure, many other substances, that are normally present, will start to degas / outgas. In a mass spectrometer this will give an increased hydrogen signal.

How Outgassing Affects Sensitivity for Hydrogen

Hydrogen is a reactive gas and many compounds like water (humidity / moisture) or hydrocarbons (oil, grease, solvents etc) contain many hydrogen atoms. Most objects handled in an industrial environment will be "covered" with a thin invisible film of oil, grease and moisture, etc. When vacuum is pulled, many of these substances will evaporate and pass the spectrometer where some will decompose and produce hydrogen. This gives a highly increased background signal.

Helium, however, is an inert gas that does not react chemically with other substances and is not contained in any compounds. Therefore, no helium is generated from decomposition of compounds.
In an industrial environment at normal production speeds, we will therefore see hydrogen background that is roughly >1,000 times higher than the corresponding helium background. This makes leak testing with hydrogen in a standard vacuum chamber a very insensitive testing technology. The smallest detectable leak rate corresponding to the hydrogen background is in the 10-5 mbarl/s range if using 100% hydrogen.

For safety reasons, hydrogen is not used in pure form, but is always mixed with nitrogen (5% hydrogen in 95% nitrogen). Considering this dilution factor, the smallest detectable leak is about >20,000 times larger than what can be detected with helium.

The smallest practical leak rate for forming gas in a standard, mass spectrometer vacuum chamber test set-up is then around 2*10-4 mbarl/s for a small chamber. For larger chambers, the outgassing of surfaces will increase and the smallest detectable leak rate will be even bigger.

However, it is possible to improve the sensitivity if measures are taken in the design of the vacuum chamber system to limit the amount of outgassing substances reaching the mass spectrometer. A close cooperation of end-user, system integrator and INFICON is needed for implementation.

Important to note: Diluting the helium to 1% is still >100 times better than using forming gas in a standard vacuum chamber.

There are also other hydrogen detectors like INFICON's Sensistor ISH2000 that not produce hydrogen from compounds. This detector is, however, less sensitive and not suited for large chambers that require mbar pressures.


Vacuum chamber testing of automotive components intended for gaseous substances, such as airbags or car fuel tanks require the sensitivity only a mass spectrometer can deliver. Using hydrogen in such applications is normally not recommended.

Small objects and objects with milder leak specifications (>1x10-4 mbarl/s) such as parts intended for water, oil and other liquids can also be tested in atmospheric accumulation chambers with helium or hydrogen.

For more information on how leak testing with hydrogen tracer gas can improve your process, contact your local INFICON office.