PHYSICAL CHEMISTRY

CHOU PIN-I, O. I. LUKYANOVA, and E. E. SEGALOVA

ON METASTABLE SOLUTIONS OF CALCIUM SILICATES

(Presented by Academician P. A. Rehbinder, 1 VI 1961)

Classical physicochemical studies of the system CaO—SiO₂—H₂O and of the hydration processes of anhydrous calcium silicates β-Ca₂SiO₄ (β-C₂S) and Ca₃SiO₅—(C₃S) (¹–⁵) leave open the question of the conditions of formation and the properties of metastable solutions that arise as an intermediate stage in the interaction of these compounds with water.

The works of P. A. Rehbinder and co-workers (⁶–⁹) showed that the physicochemical basis of the hardening processes of inorganic binders is the formation of dispersed crystallization structures that arise and develop in a medium supersaturated with respect to the hydration products of anhydrous binders. In this process the decisive role belongs to the magnitude and kinetics of the change in the concentration of supersaturated solutions. In the present work the formation of supersaturated solutions during the hydration of β-C₂S and C₃S was investigated, and the metastable solubility of these silicates was established as a function of the concentration of Ca(OH)₂ liberated as a result of their interaction with water. The change in the composition of the aqueous phase during the hydration of silicates was studied under conditions protected from CO₂ in the air, using dilute suspensions (0.04–2.0% silicates), by direct measurements of the concentrations of Ca and silicate ions (conventionally SiO₂) at a sufficiently high stirring rate. The anhydrous silicates were synthesized at 1500° from chemically pure oxides and contained: β-C₂S, 96%, and C₃S, 90% of the principal mineral, with an admixture of low-basicity silicates and with complete absence of free CaO. SiO₂ determinations were carried out by the colorimetric method using the blue silicomolybdate complex; Ca, by the complexometric method.

Determination of the composition of the aqueous phase during the hydration of calcium silicates at 20 and 30° (β-C₂S—Fig. 1a, C₃S—Fig. 1b) showed that the concentration of silicate ions passes through a maximum, the time required to reach it decreasing with increasing concentration of the suspensions and fineness of the anhydrous silicates, as well as in going from β-C₂S to C₃S under otherwise equal conditions. In the same order there occurs a gradual decrease in the SiO₂ concentration to practically 0. At the same time the concentration of Ca⁺⁺ continues to increase, amounting to 8–10 mmol/l with almost complete disappearance of SiO₂ from the solution.

The maximum concentrations of SiO₂ ($C_m$), or the equivalent amounts of silicates in solution, are invariant only in a first approximation and, unlike non-hydrolyzing binders, in particular gypsum, the period of their existence in the suspension is very short. The apparent absence of a prolonged period of high concentrations of supersaturated solution, especially noticeable with increasing concentration of the anhydrous silicate already within the range of dilute suspensions, may lead to the erroneous conclusion that supersaturation plays an insignificant role as a structure-forming factor in the case of calcium silicates. As will be seen below, all these features are explained by the dependence of the solubility of silicates on the concentration of Ca(OH)₂ liberated during hydrolysis. For the same reason, the solubility of β-C₂S and C₃S in water, determined by an indirect method (¹⁰), cannot be attained under the real conditions of silicate hardening. It is natural to assume that, like the equilibrium solubility of hydrosilicates,

it decreases as the concentration of Ca(OH)\(_2\) released in this case during the hydrolysis of anhydrous silicates increases.

Comparison of the concentrations of Ca and SiO\(_2\) in the aqueous phase of suspensions of β-C\(_2\)S and C\(_3\)S shows that, when \(C_{\max}\) is reached, an excess of Ca(OH)\(_2\) has already formed in the solution relative to the stoichiometric value of the Ca(OH)\(_2\) concentration. Thus, for example, at 30°, by the time \(C_{\max}\) is reached—1.05 mmol/l for β-C\(_2\)S and 1.07 mmol/l for C\(_3\)S—the CaO/SiO\(_2\) ratio in solution increases from the initial value of 2 to 3 for C\(_2\)S and from 3 to 3.8 for C\(_3\)S. Thus, the concentrations \(C_{\max}\) determined by us already correspond to solutions of Ca(OH)\(_2\) of definite concentration, which is not the same for β-C\(_2\)S and C\(_3\)S.

Fig. 1. Kinetics of formation of supersaturated solutions of calcium silicates. a—β-C\(_2\)S (30°): 1—0.08; 2—0.2; 3—0.4; 4—0.8% silicate. b—C\(_3\)S (20°): 1—0.04; 2—0.08; 3—0.2; 4—0.4% silicate.

In this connection it is of interest to consider the results obtained in the form SiO\(_2\) = \(f\)(CaO), used in studying the system CaO—SiO\(_2\)—H\(_2\)O. In Fig. 2, which shows this dependence for β-C\(_2\)S at 30°, three regions can be distinguished, corresponding to three stages in the formation of supersaturated solutions of calcium silicates. The left-hand part of the graph reflects the process of dissolution of anhydrous silicates, which, as the dissolution of an ionic crystal lattice, is the transfer of the constituent ions into solution in stoichiometric ratios and, since these compounds are unstable with respect to water, is accompanied by the formation of supersaturated solutions. The molar ratio in the solution corresponding to the composition of the anhydrous silicate is shown in the graph by the dashed straight line.

Fig. 2. Composition diagram of a supersaturated β-C\(_2\)S solution during its formation: a—suspension in water; b—in Ca(OH)\(_2\) solution, 2 mmol/l; c—the same, 4 mmol/l.

Further, crystallization of a low-basicity hydrosilicate is superimposed on the dissolution process; this is expressed by the deviation of the experimental points from the straight line as a result of a change in the CaO/SiO\(_2\) molar ratio in the solution. As a result of hydrolysis of SiO\(_4^{4-}\), after its transfer into solution, the CaO/SiO\(_2\) ratio in the hydrosilicate precipitate is smaller than that of the initial anhydrous silicate, which leads to accumulation in the solution of an excess of Ca(OH)\(_2\). The solid hydrate formed under these conditions, both in the case of β-C\(_2\)S and of C\(_3\)S, is monocalcium hydrosilicate, which agrees with the literature data.

The right-hand part of Fig. 2 shows the decrease in SiO\(_2\) (silicate) concentration with an increase in the total CaO content in the solution. This part of the curve does not depend on the concentration of the suspension. However, additional introduction of Ca(OH)\(_2\) into the suspension lowers \(C_{\max}\) and the subsequent SiO\(_2\) concentrations to a greater extent than is required by the dependence of the SiO\(_2\) concentration on the total calcium content in the solution.

It is more appropriate to express the dependence of the concentration of SiO₂ on the excess of Ca in comparison with the stoichiometric amount for the given silicate, or, what is the same, on the amount of Ca(OH)₂ released as a result of hydrolysis. The curves reflecting this dependence (Fig. 3) do not depend, within the limits of experimental error, on the concentration of β-C₂S or C₃S in the suspension, or on the initial content of Ca(OH)₂ in the solution (curve 1), and show the dependence of the concentration of Ca(OH)₂ on the metastable solubility of anhydrous silicates, which is reached after passing through \(C_{\max}\) in the course of silicate hydration. Hydrolysis of the silicates with liberation of monocalcium hydrosilicate and accumulation of Ca(OH)₂ begins before the metastable solubility has been reached. Therefore, under real conditions of hydration of calcium silicates, their metastable solubility is always lower than the solubility in pure water.

Fig. 3. Dependence of the metastable solubility of calcium silicates on the concentration of Ca(OH)₂ in the solution, released during hydrolysis or introduced as an additive (30°). 1—β-C₂S: a—Ca(OH)₂ released during hydrolysis; b—2 mmol/l; c—4 mmol/l Ca(OH)₂ introduced additionally into the suspension; 2—C₃S; 3—equilibrium solubility of hydrosilicate (11).

Comparison of the metastable solubility of silicates (\(C\)) with the equilibrium solubility (\(C_0\)) at identical concentrations of Ca(OH)₂ (Fig. 3, 3; data of Roller and Ervin (11)) reveals, to a first approximation, constancy of the ratios \(C/C_0\), especially evident for β-C₂S over the entire interval of Ca(OH)₂ concentrations where the SiO₂ concentration can be measured. The magnitude of this ratio (4.4 for β-C₂S) may conventionally be regarded as the relative supersaturation, the constancy of which is a consequence of the identical or very similar dependence of the metastable and equilibrium solubilities of SiO₂ (silicates) on the excess of Ca(OH)₂.

Thus, the experimentally observed decrease in the concentration of supersaturated solutions of β-C₂S and C₃S after passing through \(C_{\max}\) in the process of hydration still does not indicate a decrease in the relative supersaturation in the aqueous phase of the suspensions, but only reflects the dependence of the metastable solubility on the concentration of Ca(OH)₂. The period of high supersaturations during hydration of calcium silicates apparently continues even after the concentration of the saturated Ca(OH)₂ solution has been reached.

The invariance we have found of the concentration of supersaturated solutions with respect to the concentration of the suspension, as well as the fact that monocalcium hydrosilicate is isomorphous with the more highly basic hydrosilicates formed during hydration in concentrated suspensions (pastes), give grounds for believing that the regularities found in this work are also valid in the case of concentrated suspensions.

Moscow State University
named after M. V. Lomonosov

Received
20 V 1961

REFERENCES

1. H. Steinour, Chem. Rev., 40, 391 (1947).
2. E. Flint, L. Wells, J. Res. Nat. Bur. Stand., 12, 751 (1934).
3. K. G. Krasilnikov, Proceedings of the Conference on Cement Chemistry, Moscow, 1956.
4. T. Torvaldson, V. Vigfusson, Trans. Roy. Soc. Canad., 22, 3, 423 (1928).
5. H. F. W. Taylor, J. Chem. Soc., No. 12, 3682 (1950).
6. P. A. Rebinder, E. E. Segalova, V. N. Izmailova, Dokl. Akad. Nauk SSSR, 110, No. 5 (1956).
7. P. A. Rebinder, E. E. Segalova, Stroit. Mater., No. 1 (1960).
8. E. S. Solov’eva, E. E. Segalova, Kolloidn. Zh., 20, issue 5 (1958).
9. E. E. Segalova, Z. N. Markina, P. A. Rebinder, Dokl. Akad. Nauk SSSR, 133, No. 3 (1960).
10. V. B. Ratinov, Dokl. Akad. Nauk SSSR, 136, No. 4 (1961).
11. P. S. Roller, G. Ervin, J. Am. Chem. Soc., 62, 461 (1940).