Scanning electron microscopy (SEM) using secondary electron imaging was used to observe the morphologies of the ettringite. Samples were ground and stored in a vacuum chamber for 1 day before the observation.
Results and Discussion 1. ettringite formation
X-ray patterns of the synthesized ettringite are shown in Figure 2. In the absence of superplasticizers, the ettringite had a highest peak at around 9.2° in 2θ, indicating the presence of ettringite. In the presence of NS, the peak intensities were increased, suggesting the stabilization or formation of ettringite (Figure 2(a)).
Similar behaviors were found in ettringite in the presence of PC (Figure 2(b)). The peak intensities were increased, except at 0.02%. It is possible that the ettringite was stabilized, not newly formed. Both NS and PC unlikely supply additional sulfates so as to increase the ettringite formation. The stabilization of ettringite in the presence of superplasticizers is further discussed in Section 3.2.
(a) (b)
Figure 2: X-ray diffraction patterns of the synthesized ettringite in the presence of (a) NS or (b) PC superplasticizer.
The numbers indicate the peak intensities.
It should be noted that a small peak appeared at around 12.4° at high superplasticizer dosages. Such peak was not found in ettringite in the presence of superplasticizers at low dosages (NS at 0.2% and 0.02% PC).
This peak indicates the presence of gypsum. It appears that the amount of gypsum formation was increased by both NS and PC. The mechanism is not clear. During the synthesis, the solutions mainly contain calcium ions from calcium oxide and aluminium ions and sulfates from aluminium sulfate. The aluminium ions have the highest positive charge and may adsorb more superplasticizers. If so, the aluminium ions are easily
repulsed and not able to participate in the formation of ettringite. Therefore, the calcium ions and sulfates in solution are likely to form gypsum.
The SEM images agree with the results of XRD patterns (Figure 3). In the plain pastes, the crystals are small.
In the presence of PC and NS, the crystals are cubic and large. The crystal size is around 1 μm. With 0.1%
PC, there are some lengthy crystals with a size about 10 μm. They are gypsums. These SEM images clearly show that the superplasticizers help to stabilize the ettringite.
(a) (b)
(c) (d)
Figure 3: SEM images for ettringite (a) in the absence of superplasticizers, or in the presence of (b) 0.5% NS, (c) 0.05%
PC, and (d) 0.1% PC.
2. Stability of ettringite
Synthesized ettringite was immersed in solutions with various pH and its weight loss was calculated after several immersion days. Results are shown in Figure 4. The ettringite was unstable in both acid and basic solution but likely stable in solution with pH 10. This result in general agrees with the findings in literature (Damidot and Glasser, 1993). In solutions with pH 9 or lower, the ettringite was quite unstable and the weight loss reached 16% just within one day. However, in solutions with pH 11, the ettringite appeared to be stable in the first two days and then the weight loss suddenly increased. These results suggest that ettringite is possibly unstable in pore solutions of cement pastes, in which the pH is generally around 13 (Collepardi, Corradi et al., 1981). It should be noted that the weight loss in solution with pH 11 was significantly reduced at immersion time of 8 days. We then measured the pH of the immersion solution and found that the pH was 10.17, suggesting that the decomposed ettringite could greatly change the alkalinity of the solution so the decomposition did not continue. This phenomenon may explain why ettringite was not fully decomposed in limited acid solutions (pH 4). The pH of the acid solution at immersion time of 8 days was 10.16, which again suggests that ettringite stabilizes in solution around pH 10.
The stability of ettringite was influenced by the superplasticizers. The PC stabilized the ettringite in solutions with pH 4 or pH 9 (Figure 5(a)). The weight loss was reduced when compared with those specimens prepared in the absence of superplasticizers. Similar results were found in ettringite in the presence of NS (Figure 5(b)). The weight loss was reduced. Both superplasticizers clearly stabilized the ettringite (Figure 6).
The stabilizing effects depend on types of superplasticizers and pH of the immersing solution. In both acid (pH 4) and basic solutions (pH 9 and pH 11), the PC had a greater stabilizing effect than the NS.
Mechanisms responsible for these ettringite stabilizations are not clear. The superplasticizers were present during the synthesis of ettringite. The molecules of PC and NS are large so it is not possible for them to enter the crystal structure of the ettringite, be part of it and strengthen the microstructure (Figure 1). On the other hand, it appears that the superplasticizers adsorb on the surface of the ettringite and act as a protective layer.
If the adsorbed superplasticizer does not decompose in those acid or basic solutions, the ettringite with the adsorbed superplasticizers on surface can then be stabilized.
The superplasticizers used in this study is around acid (pH 4.55 for PC) or close to the neutral (pH 6.40 for NS) in plain solutions, so they are assumed to be unstable in the base. The PC has a long backbone and side chains with neutral charges. It dispersed the cement particles by steric force more than the electrical charges (Uchikawa et al., 1997). It is likely that, in both acid and basic solutions, the PC stabilizes the ettringite more than the NS by its long side chains. However, in strong basic solutions (pH 11), the differences between the weight losses of the specimens in the presence of these two superplasticizers were not quite large.
The stabilization of ettringite in the presence of superplasticizer is important to the incompatibilities between cements and superplasticizers. The ettringite has been found to adsorb more superplasticizers than most of the other phases in hydrated cement pastes and also considered to be responsible for the gelation in the pastes with NS (Chen, 2007). This study further supports this argument. The pore solutions of the cement pastes have high alkalinity, around 11－12 in pH. In the presence of NS, the ettringite is stabilized and the amount of ettringite was much more than those in the plain pastes so the bridging effect can occur. In the presence of PC, although the ettringite can also be stabilized, the dispersing effect by PC remains. The dispersion by PC is induced by not only the electrostatic repulsion but also the steric force. The dispersion is likely more dominant than the bridging by ettringite.
Figure 4: Weight loss of ettringite synthesized in the absence of superplasticizers.
(a) (b)
Figure 5: Weight loss of ettringite in the presence of (a) 0.02% PC, or (b) 0.2% NS.
Figure 6: Weight loss of synthesized ettringite in solutions with various pH after 4-day immersion.
Conclusions
The stability of synthesized ettringite is influenced by both the alkalinity of the surrounding solution and the presence of superplasticizers during the synthesis. Ettringite was found stable in solution with pH 10. In the presence of superplasticizers, the resulting ettringite was stabilized in both acid and basic solutions. The stabilization induced by carboxylated superplasticizer was more than that by naphthalene-based superplasticizer. Results in this study may explain the cement-admixture compatibility problems due to the ettringite formation during the early hydration of cement pastes.
Acknowlegement
The authors thank the National Science Council, Taiwan, ROC, for their support through the grant 99-2815-C-011-024-E. We would also like to thank Professor Leslie Struble at University of Illinois for fruitful discussions on ettringite synthesis.
References
P.-C. Aïtcin, C. Jolicoeur and J.G. MacGregor, 1994. Superplasticizers: How They Work and Why They Occasionally Don't. Concrete International, 16(5), 45-52.
C.-T. Chen, 2007. Interactions between Portland Cements and Carboxylated and Naphthalene-Based Superplasticizers.
Thesis (PhD). University of Illinois.
M. Collepardi, M. Corradi and M. Valente, 1981. Influence of Polymerization of Sulfonated Naphthalene Condensate and Its Interaction with Cement. Detroit: American Concrete Institute, 485-498.
D. Damidot and F. Glasser, 1993. Thermodynamic Investigation of the CaO-Al2O3-CaSO4-H2O System at 25° C and the Influence of Na2O. Cement and Concrete Research, 23(1), 221-238.
C. Jolicoeur, J. Sharman, N. Otis, A. Lebel, M. Simard and M. Page, 1997. The Influence of Temperature on the Rheological Properties of Superplaticized Cement Pastes. Farmington Hills, Mich.: ACI International, 379-391.
S.M. Khalil and M.A. Ward, 1980. Effect of Sulfate Content of Cement upon Heat Evolution and Slump Loss of Concrete Containing High-range Water-reducers (superplasticizers). Magazine of Concrete Research, 32(110), 28-38.
B.-G. Kim, S. Jiang, C. Jolicoeur and P.-C. Aitcin, 2000. The Adsorption Behavior of PNS Superplasticizer and Its Relation to Fluidity of Cement Paste. Cement and Concrete Research, 30(6), 887-893.
A. Moore and H. Taylor, 1970. Crystal Structure of Ettringite. Acta Crystallographica Section B: Structural Crystallography and Crystal Chemistry, 26(4), 386-393.
W. Prince, M. Espagne and P.-C. Aitcin, 2003. Ettringite Formation: A Crucial Step in Cement Superplasticizer Compatibility. Cement and Concrete Research, 33(5), 635-641.
I. Schober and U. Mader, 2003. Compatibility of Polycarboxylate Superplasticizers with Cements and Cementitious Blends. 1st. Farmington Hills, MI: American Concrete Institute, 453-468.
L.J. Struble and P.W. Brown, 1984. An Evaluation of Ettringite and Related Compounds for Use in Solar Energy Storage. NBSIR-84/2942. Washington DC: National Bureau of Standards Department of Commerce.
H. Uchikawa, S. Hanehara and D. Sawaki, 1997. The Role of Steric Repulsive Force in the Dispersion of Cement Particles in Fresh Paste Prepared with Organic Admixture. Cement and Concrete Research, 27(1), 37-50.