Exchange rates of ethanol with water in water-saturated cement pastes probed by NMR

Diffusion of ethanol into water-saturated white cement pastes has been investigated by carbon and proton nuclear magnetic resonance (NMR). The diffusion of ethanol tons shown to be Fickian, assuming one-dimensional diffusion under perfect sink boundary conditions. Derived diffusion coefficients were found to increase with increasing water/cement (w/c) ratio from (2.7 +/- 0.5) 10(-8) cm(2)/s at w/c = 0.30 to (59 +/- 5) 10(-8) cm(2)/s at w/c = 1.0. At the end of the exchange process, only a fraction of the total volume of water is exchanged with ethanol, varying from 60% for samples containing mainly micro- and mesopores to about 80% for samples where additional capillary pores are present. Time needed to reach 90% and 95% exchange of the total intrudable amount of ethanol in cylindrical samples with diameter of 5.5 mm varied from 1 day to nearly 3 weeks. This has importance for exchange in larger samples with typical diameters of 10 mm or move (as used in mercury intrusion porosimetry), which may require on the order of months for 90% exchange to take place. The mole fraction of ethanol and tenter in the pure system was determined from sampled carbon and proton NMR spectra vs. exchange time by comparing H2O-saturated and D2O-saturated samples. At the end of the exchange process, water tons found to occupy the remaining volume not accessible to ethanol. In the tested w/c ratio range, the water content in all samples is below the value where damage to the pore structure normally occurs due to internal tension when exposed to drying. An empirical relationship between chemical shift of the CH3CH2OH/H2O peak and mole fraction of ethanol is derived, enabling the mole fraction of ethanol from the NMR peak to be estimated.. (C) 1998 Elsevier Science Ltd.