The divided cell arrangement was used as the experimental apparatus because of its procedural simplicity. Since the ceramic specimens have a relatively large diffusion coefficient with respect to hydrated cement paste, usable data could be collected within a reasonable period of time. As mentioned previously, the ceramic material had a relatively small surface charge, surface adsorption effects could be neglected. The result is a material that should exhibit nearly ideal diffusive transport properties.
Iodide was chosen as the measured ionic species because commercial solid-state ion selective combination electrodes for iodide have experimental advantages over chloride electrodes. For each measurement, the electrode was standardized using reference iodide solutions.
Each ceramic frit was first mounted, using epoxy, into an acrylic annulus. After the epoxy had cured, the sample was saturated with a potassium iodide (KI) solution. The saturated frit was then clamped between two glass vessels, each having a capacity of approximately 250 mL, using rubber o-rings to make a water tight seal. One vessel was then filled with the same KI solution and the other vessel was filled with a test electrolyte. Both vessels were sealed to reduce evaporation, and then placed inside a walk-in environmental chamber maintained at 25 ºC. All sampling and concentration measurements were performed within the chamber in order to ensure stability in both the diffusion and the measurements; there is a nominal 2 % / ºC variation in the self-diffusion coefficient as a function of temperature [30], and the concentration measurement precision improved within the stable environment. Additional details of the experimental procedure are given elsewhere [7].