For this experimental program, the solutions chosen to represent pore solution are composed solely of potassium hydroxide and sodium hydroxide. The molar ratios of potassium to sodium studied were 4:1, 2:1, and 1:1, as these appear to represent the vast majority of possible pore solution compositions beyond 28 d [18,19,20,21]. The potassium hydroxide concentration was varied up to 1 mol/L, and the sodium hydroxide concentration was adjusted according to the predetermined molar ratio.
The solution conductivities were determined using a glass conductivity cell containing platinum electrodes. The conductivity cell was cylindrical, with an inside diameter of approximately 25 mm, and had an electrode separation of approximately 320 mm. The cell constant (the effective ratio between the apparatus length to area) was (5.0578 ± 0.0030) cm-1, and was determined from 0.01 mol/L and 0.10 mol/L standard potassium chloride solutions [22]. The uncertainty reported in the cell constant is the difference between the two calculated cell constants for the two standard solutions; the individual precisions were each less than this reported uncertainty.
All measurements were performed in a walk-in environmental chamber that was maintained at (25.0 ± 0.4) º C; the reported uncertainty is the standard deviation of the temperature control hysteresis. The pore solution was allowed to thermally equilibrate overnight in a volumetric flask. The conductivity was determined using a commercial impedance spectrometer. Measurements were repeated until the calculated conductivity changed by less than 0.2 % over 1 h; because of safety concerns due to the caustic nature of the solutions, the cell was filled outside the chamber, resulting in a small thermal measurement drift after returning the cell to the environmental chamber. Given the uncertainty in the cell constant, a 0.2 % uncertainty would characterize the uncertainty in the reported conductivity measurements.