*3.2. Diffusion Test—Principle of the Method*

This test has been newly designed in our laboratory in order to induce gas transport through a concentration difference at atmospheric pressure on both sample sides. As the use of a gas spectrometer has caused some calibration difficulties, we chose to use a simpler gas analyzer with a 500 ppm resolution. This apparatus was calibrated to detect helium into nitrogen, and it can work in a closed circuit. This means that helium will diffuse into a reservoir containing nitrogen (initially 100% nitrogen) and that the apparatus will analyze a small proportion of the mix (nitrogen + helium) and re-inject this proportion into the reservoir after analysis. The principle of this test is indicated in Figure 3. There is a continuous pure helium flow at the upstream side. This implies that the helium concentration is constant at this side despite nitrogen diffusion.

**Figure 3.** Schematic principle of the diffusion test; nu is the number of helium moles, cu and cd are helium concentrations at the upstream (u) or downstream (d) sides, respectively. Vd is the downstream reservoir volume.

The device designed for diffusion tests is presented in the picture in Figure 4. The balancing device at Patm is realized by dip tubes whose height (into oil to avoid evaporation) can be adjusted. This allows for a pressure regulation with an accuracy that is better than one millibar. The percentage of helium at the downstream side is periodically measured by the gas analyzer.

**Figure 4.** Picture of the device designed for diffusion.
