Pore structure analysis by water vapor adsorption. III. Analysis of hydrated calcium silicates and portland cements

Abstract

The pore volume and surface distributions of three hardened pastes of hydrated dicalcium silicate, prepared with water to solid weight ratios, w/s, of 0.45, 0.57, and 0.70, of two hardened pastes of hydrated tricalcium silicate, with w s of 0.45 and 0.57, and of three hardened pastes of a hydrated portland cement, with w s of 0.40, 0.45, and 0.70, were determined by water vapor adsorption-desorption isotherms. In the analyses, the two new methods developed by Brunauer and his coworkers were used: the “MP-method” for the micropores and the “corrected modelless” method for the wider pores. Seven of the eight pastes were previously analyzed by nitrogen adsorption-desorption isotherms, using the same methods. The pore systems of the pastes were only partly accessible to nitrogen; the micropores were not accessible at all, and even the wider pores were only partly accessible. The results indicate the presence of many “ink-bottle” pores, i.e., pores with wide bodies and constricted entrances. In the hydration of both calcium silicates, the products are a calcium silicate hydrate, called tobermorite gel, and calcium hydroxide. The former is a colloidal substance, and it is responsible for the pore properties of the pastes. The pore structures of the pastes of the two silicates were found to be somewhat different. An average portland cement contains about 50% tricalcium silicate and about 25% dicalcium silicate by weight. The tobermorite gel produced in the hydration of portland cement is responsible for the pore properties of cement pastes. These properties were found to be intermediate between those of the two calcium silicates, but closer to those of the tricalcium silicate pastes. The other hydration products of portland cement appear to have little effect, if any, on the pore properties of the tobermorite gel. Tobermorite gel is a layer crystal, similar to some of the clay minerals. Recently, the hypothesis has been advanced that the difference between water vapor and nitrogen adsorption is due to the fact that the former can enter between the layers of the gel, but the latter cannot. The pore structure results presented in the paper refute this hypothesis. The paper also offers a new and more illustrative method for the presentation of the pore structure data than the customary differential curves.