We constructed a new two-chamber device for gas diffusion measurements in the laboratory, where the diffusion process is followed by two counter-diffusing gases. One of the two chambers of the new apparatus is flushed with Ar, and the diffusion process is started by removing a metal slide separating the chamber from the soil sample. The partial pressure of Ar as well as O2 is timelapse monitored in the flush chamber as well as in a passive chamber coupled to the other end of the soil sample. We measured the gas diffusion coefficient for undisturbed and remolded soil when the samples were drained to a matric potential of -300 hPa. Cores of autoclaved, aerated concrete (AAC) were prepared in the sizes fitting the new device as well as an existing onechamber apparatus. Gas diffusion was measured through the AAC cores and through simple tubes for both methods. The relative gas diffusivity of Ar and O2 for soil samples in the two-chamber device closely followed a 1:1 relation, indicating no significant O2 consumption during measurements. The relative diffusivities for the cylindrical tubes were identical to their cross-sectional area, providing a quantitative assessment of the absolute accuracy. The intercalibration of the two methods based on the AAC cores showed no significant difference, and also indicates that the error from not taking the storage aspect into account for the one-chamber system is negligible for moist soil samples. We consider the new two-chamber, two-gas methodology as promising for detailed studies of gas diffusion in soil.
