Silicate anion structure

Because C-S-H gel is nearly amorphous. X-ray diffraction has given only very general indications of its structure. The nature of the silicate anions has been determined from the kinetics of the reaction with molybdate (S43), and, in greater detail, by trimethylsilylation (TMS) and Si NMR. In TMS methods, the sample is treated with a reagent that converts the silicate anions into the corresponding silicic acids, which then react further with replacement of SiOH by SiOSi(CH3)3. The resulting TMS derivatives can be identified and semiquantitatively determined by various procedures, of which the most widely used have been differential evaporation to isolate the 

NMR spectrum, the amount of CH given by TG and the cumulative heat evolution. 

Early eonelusions from TMS studies that the average size of the silieate anions in pastes inereases on drying (M46,B54) were not supported by later work (P21). An NMR study on the effeet of drying eondilions on the anion structures of C-S-H prepared in suspension (M47) (Seetion 5.4.4) has important implications for pastes. 

---

Figure 7.2 Silicate anion structures (o) orthosilicate, (6) pyrosilicate, (c) three-silicate ring, (d) six-silicate ring, (e) pyroxene, (/) amphibole, and (g) phyllosilicate.

Figure 7.3 Sheet silicate anion structure idealized.

Figure 24 Representation of possible silicate anion structures for the species identified by b-n-j connectivites deduced from the 2D 29Si INADEQUATE spectrum.The solid lines represent silicon-oxygen-silicon linkage and closed circles (A, B and C) are inequivalent site within each anion illustrating connectivities.

The silicate anion structures of C-S-H(I) preparations appear to be affected by how long the material remains in contact with its mother liquor and by how strongly it is subsequently dried. Experiments using the molybdate method showed that the anions in precipitates made from CaCl, and sodium silicate solutions at 0 "C were mainly those present in the silicate solution, and thus monomeric if the latter is sufficiently dilute (S45). By allowing such products to stand in contact with their mother liquors at O C. and drying at — 10°C, preparations with Ca/Si ratios of 1.2-1.5 were obtained that contained only dimeric silicate anions. 

Some other models for the structure of C-S-H gel that have been proposed are incompatible with the evidence. A proposal identifying it with natural tobermorite, based on IR and extraction results (S37), appears to ignore both composition and degree of crystallinity. Another, assuming it to be closely related to the CH structure, with incorporation of monomeric silicate ions (G38), is inconsistent with the observed silicate anion structure. As noted earlier, one assuming three-dimensional anionic clusters (C23) is inconsistent with the Si NMR evidence, and with the overwhelming proportion of the other evidence on silicate anion structure. 

Silicate anion structures in Portland cement pastes have been studied by the methods described in Section 5.3.2 for calcium silicate pastes. Trimethylsily- i lation (TMS) studies (L20,T12,S69,T36,L31,M43,M44) show that, as with C,S. the proportion of the silicon present as monomer decreases with age and that the hydration products contain dimer, which is later accompanied and eventually partly replaced by polymer (>5Si). Some results have i indicated that fully hydrated pastes of cement differ from those of CjS in that substantial proportions of the silicate occur as monomer (S69,L31), but the results of a study in which pastes of CjS, P-CjS and cement were compared (M44) suggest that the differences between the anion structures of cement and CjS pastes are probably within the considerable experimental errors inherent in the method. The recovery of monomer from unhydrated P-CjS was only 66% and results for cement pastes can only be considered semiquantitative. 

A Si NMR study (BlOO) showed that for cement, as for CjS pastes, the content of Q° silicate tetrahedra decreases with time and that those of Q and later of tetrahedra increase. After 180 days, the degree of hydration, estimated from the intensities of the NMR peaks, was approximately 90%. These results are consistent with those obtained by the TMS method, and suggest that the hydration products present after 180 days contain at most only a small proportion of monomer. The possible effects on the silicate anion structure of drying, whether during hydration as a result of localized water shortage or subsequently, were considered in Section 5.3.2. 

Table 7.2 Comparison of silicate anion structures in pastes of C S and Portland...

Fig. 11-4. Sheet silicate anion structure idealized. [For a real example see A. K. Pant Acta Cryst., B, 1968, 24, 1077.]...

Their role in the conductivity is also not large. The conductivity of waterglass is determined mainly by the motion of alkali cations and depends on the silicate anion structure and on the solution s viscosity. 

---