Page 113 - Handbook of Thermal Analysis of Construction Materials

96 Chapter 3 - Formation and Hydration

In the determination of the constitutional water associated with
the C-S-H phase, it is important to differentiate the unreacted water
from that which is bound by the gel. Feldman and Ramachandran [37]
carried out TG/DTA of a bottle hydrated C S that was conditioned at several
3
humidities for lengthy periods, starting from the d-dry, 100% RH and 11%
RH. It was determined that the stoichiometry of the bottle hydrated C-S-H
gel at 11% RH (approached from 100% RH) is 3.28CaO•2SiO •3.92H O.
2
2
It was also found that the thermograms could delineate free and interlayer
water and the relative intensities of the peak effects of samples exposed to
different conditions showed a good relationship with adsorption isotherms.
A method called Dynamic Differential Calorimetry was adopted by Kurczyk
and Schweite [1] to determine CaO/SiO ratio of C S hydrated to different
2 3
periods. The ratios at 1, 2, and 3 months of hydration were 1.88, 1.87, and
1.92 respectively. The DTG method can also be applied with success to
estimate calcium hydroxide in the hydrated silicate system as the dehydra-
tion of calcium hydroxide is attended by a loss of water.
A mature hydrated silicate paste contains both the C-S-H and CH
phases. Any characterization study of the C-S-H phase is not easy because
of the possible interference due to the presence of calcium hydroxide. A
procedure has been developed by Ramachandran and Polomark for extract-
ing calcium hydroxide from the C S paste without affecting the C-S-H
3
phase. [38] It involves exposing a thin disc of the paste to an aqueous solution
of calcium hydroxide in such a way that the concentration of the solution

remains in the range 9–12 millimole CaO/L throughout the extraction
period. Calcium hydroxide remaining in the solid is monitored by DSC and
further extraction is stopped when the amount of free lime remaining in the
solid is almost nil. In Fig. 20, the amount of lime remaining in the C S pastes
3
at different times of extraction is shown.
It is an established fact that C S is much more reactive than C S.
3
2
The relative reactivities of Ca in various compounds of interest in cement
science were studied by the extent to which they reacted with silver
nitrate. [39] DTA was used in the cooling mode to estimate the unreacted
silver nitrate by an exothermal transition at about 190°C. (See Table 3.) The
reaction between CaO and AgNO is almost stoichiometric. Calcium
3
hydroxide also reacts stoichiometrically with silver nitrate. Only 0.81 mol
out of 3 mols in C S has reacted with silver nitrate. This suggests that about
3
27% of tricalcium silicate is more reactive than the rest. Isothermal
conduction calorimetric curves show that about the same amount of C S is
3
relatively more reactive. Thus, it seems that all Ca ions are not the same in
the silicate. Dicalcium silicate phase has reacted about 6 % only and this