U.S. DEPARTMENT OF COMMERCE NATIONAL BUREAU OF STANDARDS RESEARCH PAPER RP1590 Part of Journal of Research of the National Bureau of Standards, Volume 32, June 1944 STUDIES OF PORTIONS OF THE QUATERNARY SYSTEM ABSTRACT A study has been made of portions of the system soda-lime-silica-water at 25° C. The only solid phases found were Ca(OH), and a four-component gel of variable composition. The boundary, Ca(OH)2-soda-lime-silicate gels, was determined and the compositions of the gels along this boundary shown to vary from 0.003 Na0:2.0CaO:1.OSiO2:H2O (at 0.2 g of Na2O per liter) to about 0.25 Na,O:1.0ca0:1.OSiO2:H2O (at 20 g per liter and extending to 101 g per liter of Na2O). In regions off the boundary, at selected constant concentrations of Na,O but with increasing concentrations of SiO, in solution, the Na2O:SiO, molar ratio of the gels varied only slightly from 0.2 in most cases; the CaO: SiO2 molar ratio, however, decreased to values approaching 0.1 at the maximum concentrations of SiO, used. Interpretations pertaining to relations between the composition of the gels and solutions are given. (a) Compositions of soda-lime-silicate gels___ 289 (b) Relation between compositions of gels and solutions 293 3. Regions of higher silica concentrations 296 V. Applications of the results of the study to concrete. 298 VI. Summary 301 VII. References_ 302 I. INTRODUCTION The role of the alkalies, soda and potash, in hydrating cements is not fully understood. Information on how these constituents alter the chemical composition of the lime silicate hydrates and of other products in hydrating cements should be highly useful in explaining the mechanism of the various hydration processes involved. Also, a knowledge of the compositions of the alkali solutions in which the hydrated cement compounds are stable should prove helpful in elucidating the effects of the alkalies on hydrating cements in the presence of "reactive" aggregates.. The present investigation was accordingly designed for obtaining a portion of such data and was concerned with the effects of Na2O on the compositions of the lime silicate hydrates and of the aqueous phase in which they are stable. The investigation was essentially a study of portions of the quaternary system soda-lime-silica-water at 25° C. II. MATERIALS The starting materials were solutions of sodium silicate and sodium hydroxide of known compositions, crystals of Ca(OH), and distilled water. The sodium silicate solution was a reagent-quality product containing not more than 0.2 percent of CO2 and having a molar ratio of 1.00Na2O:3.34SiO2. Sodium hydroxide was freed of carbonate by preparing a saturated solution. The clear supernatant liquid was drawn off and diluted as required. Calcium hydroxide of reagent quality and having a particle size of less than 1 micron was used for most of the experiments. Also, Ca(OH)2 crystals of varying sizes, prepared by adding solutions of NaOH to saturated solutions of Ca(OH)2, were used in a number of exploratory preparations. III. PROCEDURE 1. PREPARATION AND FILTRATION OF MIXTURES A few tentative mixtures of the standard solutions and the Ca(OH)2 were made in order to establish the amounts required to produce the desired quantities (about 0.5 to 1.0 g) of the precipitates in mixtures having a volume of 200 ml. Most of the mixtures were made by adding a solution of sodium silicate to a well-agitated suspension of fine-grained Ca(OH)2 in a solution of NaOH. The resulting preparations in tightly stoppered ceresin-lined flasks were stored in an air bath maintained at 25±0.05° C until filtered at ages ranging from 2 to 18 weeks. During the first 5 to 8 hours after mixing the preparations, they were shaken about once each hour, and then once daily after the first day. Although this was the general procedure followed, it will be necessary, as required in certain sections of this report, to give more specific details. The time allowed the mixtures to "age" was selected arbitrarily. It may be mentioned, however, that neither the solid phase nor solution underwent any determinable changes in time periods of from several weeks to several months. It is recognized that this steady state may be displaced by eventual crystalization of the solid phase, a process which, however, appears to proceed exceedingly slowly at room temperature. The mixtures were filtered through a medium-textured sintered glass crucible by application of suction. The residue on the filter during the last stages of filtration was firmly compacted with a rubber plunger, and as much of the solution was withdrawn as permitted by the retentivity of the gelatinous precipitate. The precipitate was next redispersed in a small amount of wash solution (ethyl alcohol and water or water if the filtrate contained more than 70 g of SiO2 per liter) and the solution again filtered off. The most suitable ratio of alcohol to water for each wash solution was determined by testing a few milliliters of the original filtrate and selecting the maximum alcohol content which just failed to produce a suspension. Four to five successive washings were made with solutions of progressively increasing alcohol content, then four to five washings with 95-percent ethyl alcohol, and finally four to five washings with ethyl ether. In the course of each washing, the precipitate was redispersed and during filtration, compacted on the filter. After being washed the precipitate was left in the laboratory air for 1 day and then stored in a tightly stoppered vial. 2. METHODS OF CHEMICAL ANALYSES Standard procedures of chemical analyses were followed. All of the solutions and most of the precipitates were analyzed for soda, lime, and silica. The "hydrate" water of the precipitates was determined by difference. The amounts of Na2O in the filtrates containing relatively small amounts of SiO2 (1 percent or less of the total solids) were determined by titrating the solutions with HCl, using methyl red as indicator. Allowance was made for the Ca(OH), present, which was determined gravimetrically. The errors in the results were not greater than 1 percent of the amount of Na2O, as determined occasionally for check purposes by precipitation. In all the other determinations, the Na2O, after removal of SiO2, was precipitated and weighed as sodium zinc uranyl acetate hydrate. Calcium was precipitated as oxalate, ignited, and weighed as CaO. However, when the filtrates contained over 50 g of Na2O per liter the CaO was precipitated as phosphate, converted to sulfate, and reprecipitated as oxalate according to the method described by Hillebrand Lundell [1]. In the analysis for SiO, the solutions were dehydrated twice with HCl. The combined residues were ignited to constant weight and treated with HF. The loss in weight was reported as SiO2. IV. RESULTS AND DISCUSSION 1. PRELIMINARY OBSERVATIONS Exploratory mixtures showed that only gelatinous precipitates were formed. Some of these, in mixtures of high concentrations of SiO2, were highly solvated gels and others, in mixtures of high concentrations of NaŎ and low concentrations of SiO2, appeared as grains of gel. Analyses indicated that the CaO:SiO2 molar ratio of the gels decreased as the concentration of Na2O in solution was increased and the Na2O:SiO, molar ratios were found to increase up to a certain value. At constant concentrations of Na2O in solution, but with increasing amounts of SiO2, the CaO:SiO, ratios were found to decrease. The four-component gelatinous products are referred to as soda-lime-silicate gels, or, more often, simply as gels. The shifts in the compositions of the gels could not be followed microscopically as the changes in the indices were small and appeared to be affected by aging. The presence or absence of small amounts of fine-grained crystals of Ca(OH)2 could not be ascertained reliably by means of the microscope. Large crystals of Ca(OH), could not be Figures in brackets indicate the literature references at the end of this paper. used as they were converted but slowly to the siliceous products owing to the formation of gelatinous deposits around the crystals. 2. BOUNDARY, Ca(OH),-SODA-LIME-SILICATE GELS Considerations based on the preliminary results indicated that the best initial procedure to follow in studying the system would be to ascertain the compositions of the aqueous phase in contact with Ca(OH)2 and the soda-lime-silicate gels. Accordingly, the mixtures for establishing this boundary were prepared, only a moderate excess of crystalline Ca(OH), being used. The concentration of Na2O was varied from 0.2 to 152 g per liter. After 2 to 7 weeks, the mixtures were filtered and the filtrates and precipitates analyzed. All the data pertaining to the mixtures, filtrates, and precipitates are reported in table 1, the compositions of the solutions being given in grams per liter of Na2O, CaO, and SiO2. The table shows that the amount of CaO in solution decreased from 0.92 to about 0.02 g per liter as the concentration of Na2O was increased from 0.2 to about 20 g per liter. Simultaneously, the concentrations of SiO, increased from 0.002 to about 0.020 g per liter. In the range of concentrations of Na2O from 20 to 152 g per liter, the amounts of CaO in solution remained nearly constant at a value of about 0.010 g per liter, but those of SiO2 increased from 0.02 to over 1.1 g per liter. Figure 1 shows the g per liter of CaO and SiÓ, found in solution at the different concentrations of Na2O. FIGURE 1.-Amounts of Na2O, CaO, and SiO, in filtrates from mixtures containing solid Ca(OH)2 and soda-lime-silicate gels. TABLE 1.-Compositions of solutions in contact with crystalline Ca(OH), and sodalime-silicate gels and the compositions of the mixed solids (a) COMPOSITIONS OF SODA-LIME-SILICATE GELS The precipitates obtained in the mixtures used for locating the boundary, although containing Ca(OH)2, were suitable for ascertaining the Na2O:SiO2 molar ratios of the soda-lime-silicate gels. The compositions of the precipitates, assigning a value of unity to the mols of SiO2, are given in table 1 (last four columns) as the molar ratios of Na2O:CaO:SiO2:H2O. Thus, the column under Na2O gives the molar ratio of Na2O:SiO2 of the gels. The molar ratios of CaO to SiO2 of gels along the boundary could not be determined reliably by preparing mixtures, computed to be removed just slightly off the boundary so as to avoid the presence of Ca(OH)2, and analyzing the precipitates. Because of this difficulty, the following procedure for determining the composition of the gel coexisting with crystalline Ca(OH)2 at constant concentration of Na2O was selected as the most expedient and the one likely to give reliable results. Series of 5 to 10 mixtures each were prepared at Na2O concentrations (or levels) of 0.2, 2.4, 5.5, 9.2, 11.6, 22.3, 51, 75, and 101 g per liter. (See table 2.) The total quantities of Ca(OH)2 and SiO2, added as sodium silicate, were estimated from the amounts needed for the solutions as shown in figure 1, plus the approximate amounts required for the precipitates as indicated by results on exploratory |
