Hydration chemistry of slag cements

The scanty results in the literature for percentages of the slag that has reacted (Table 9.3) show wide discrepancies. The extents to which these result from differences between slags, experimental errors, proportions in the 

Determined by the method of Kondo and Ohsawa (K42). with correction for other phases undissolved except for Ref L27, for which an EDTA method was used. 

Harrisson ct til. (H4.H49) represented the results of X-ray microanalyses of individual spot analyses in all parts of the microstructure other than unreacted clinker grains on plots of Al/Ca ratio against Si/Ca ratio and of Mg/Ca ratio against Al/Ca ratio (Figs 9.1 and 9.2). If the analyses of the material formed in situ from the slag are excluded, the plot of Al/Ca against Si Ca is broadly similar to those obtained for pure Portland cement pastes, and may be interpreted in the same way (Section 7.2.5). 

Taken together, the results in the two figures show that the material formed from the slag has an approximately constant Si/Ca ratio of 0.62, and ratios of Mg Ca and of Al/Ca that vary from point to point on a micrometre scale but which are related to each other by the equation shown in Fig. 9.2. This was interpreted (H49) as indicating mixtures in varying proportions of 

C-S-H having Si/Ca 0.62, Al/Ca 0.09, with a hydrotalcite-type phase having Al/Mg 0.38. As with Portland cement pastes, the material described in Fig. 9.1 as C-S-H may really be an intimate mixture of the latter with AFm and. to a minor extent, hydrotalcite-type structures. 

---