Mechanical Performance and Microstructure of Alkaline Earth-Activated Slag Cements

Abstract

Mechanical performance and microstructural features of various types of alkaline earth-activated slag cements were investigated as a possible alternative to OPC using X-ray fluorescence spectroscopy, laser diffraction, compressive strength test, conventional and synchrotron X-ray diffraction, thermogravimetric analysis, mercury intrusion porosimetry, solid-state 27Al and 29Si magic-angle spinning nuclear magnetic resonance spectroscopy, and scanning electron microscopy with energy dispersive spectroscopy analysis. Calcium hydroxide-activated slag cements have been investigated as an alternative cement. However, in general, the compressive strength of calcium hydroxide-activated slag cements has not been competitive with that of OPC. In this study, four different auxiliary activators which are sodium hydroxide, sodium carbonate, sodium sulfate, and gypsum were used with calcium hydroxide which is a major activator for GGBFS. All auxiliary activators increased early age strength at 3 days but the strength at 28 days was comparable to or even below the sample which has no auxiliary activators. The reduction of water-to-slag ratio appeared to be more effective way to improve the strength than the use of the four auxiliary activators. Microstructures of calcium hydroxide-activated slag cements were highly affected by auxiliary activators where the cements produced different types of hydration products, revealed different reaction degree of slag, and formed C-S-H with different characteristics depending on the additional activators. Nevertheless, the level of compressive strength did not show big differences, which indicates that the compressive strength of calcium hydroxide-activated slag cements could be determined by physical reasons such as porosity, not by chemical factors (e.g., type of hydration products, or dissolution degree of slag). Based on the results of 29Si MAS-NMR, the substantial proportions of monomeric and dimeric silica (i.e., Q1 and Q2 silica) might not belong to the C-S-H, where the mean chain length of C-S-H had no relationship with the strength, which indicated that substantial portions of monomers and dimers of silica might be attributed to amorphous phases, not to C-S-H. Barium hydroxide was also used for slag activation as a main activator and the properties of barium hydroxide-activated slag cements were compared to those of calcium hydroxide-activated slag cements. The strength of barium hydroxide-activated slag cements was significantly higher than that of calcium hydroxide-activated slag cement except at 3 days. The lower strength of barium hydroxide-activated slag cement at 3 days might be attributed to the lack of ettringite formation because barite (BaSO4) was a more favorable phase than ettringite, which could play a role in producing early age strength. Nevertheless, the strength of barium hydroxide-activated slag cements was always higher than that of calcium hydroxide-activated slag cement after 3 days because the higher pH value was expected due to the higher solubility of barium hydroxide compared to calcium hydroxide and more hydration products were produced in barium hydroxide-activated slag cement than calcium hydroxide-activated slag cement. The hydration products formed in barium hydroxide-activated slag cements were barite, witherite, strätlingite, hydrotalcite-like phases, portlandite and C-S-H while those of calcium hydroxide-activated slag cements were ettringite, portlandite, and C-S-H. Although the C-S-H was formed with both barium hydroxide and calcium hydroxide, the characteristics of C-S-H were clearly different. The hardened matrix in calcium hydroxide-activated slag cements had Ca/Si ratios around ~ 1.4 while that of barium hydroxide-activated slag cements had a Ca/Si ratio below ~ 1.1, which was similar to the typical Ca/Si ratio of C-S-H(I). The highest dissolution degree of slag was achieved in barium hydroxide-activated slag cement due to the higher solubility of barium hydroxide, resulting in the highest amount of hydration products. Thus, the highest compressive strength and the smallest porosity were obtained in barium hydroxide-activated slag cements. Calcium oxide-activated slag cements from four different slag sources were investigated to verify the relationship between the intrinsic properties of slag and the properties of calcium oxide-activated slag cements. Despite the same mixture proportion, the strength development of calcium oxide-activated slag cements was significantly different, which indicated that the material characteristics of slag could highly influence the properties of calcium oxide-activated slag cements. However, the strength development was not governed by any single dominant material parameter of the raw slag, but rather by the combination of various favorable factors. Favorable characteristics of slag for high strength were (a) higher calcium sulfate content, (b) finer particle sizes, and (c) higher basicity and chemical modulus, while the influence of glass content was not notable. The main hydration products of calcium oxide-activated slag cements were C-S-H and portlandite as well as ettringite in cases containing calcium sulfate and various types of AFm with calcium sulfate sources. The slag which contained both calcium sulfate and calcium carbonate produced the highest compressive strength with CaO activation because of its high amount of ettringite formation. The slag collected from Dubai had the lowest strength due to the presence of bulk-sized porous calcium carbonate, indicating that, although finely ground calcium carbonate could contribute to the strength development of calcium oxide-activated slag cements, the presence of large sized calcium carbonate could be harmful to strength development because it could be a fragile point for crack initiations under loading. Finally, the influence of calcium carbonate powder on CaO-CaSO4-GGBFS composite cements was investigated. In this study, carbon sequestrated calcium carbonate powder was used as one of the raw materials. The compressive strength of CaO-CaSO4-GGBFS composite cements was increased with increasing calcium carbonate substitution up to 20 wt.% but decreased with further substitution. However, the total porosity of CaO-CaSO4-GGBFS composite cements measured by MIP gradually decreased with increasing calcium carbonate powder. All samples produced significant amounts of C-S-H and ettringite as main hydration products and formed a small amount of portlandite. The samples which contained 10 wt.% and 20 wt.% of calcium carbonate powder produced hemicarboaluminate but the samples which contained 50 wt.% of calcium carbonate powder did not, due to the lack of aluminum supplies from GGBFS. While the amount of GGBFS was decreased with the increasing amount of calcium carbonate powder, the amounts of hydration products did not show a large difference and even increased, slightly, with increasing calcium carbonate powder, indicating that the amounts of hydration products might be determined by the amount of activators or that calcium carbonate powder could increase the dissolution degree of GGBFS. EDS spot analysis clearly showed that the Ca/Si ratio in C-S-H increased with increasing calcium carbonate powder, which was the reason why the compressive strength of the sample which contained over 30 wt.% of calcium carbonate powder decreased while the porosity gradually decreased. The mechanical performance of C-S-H decreased with the Ca/Si ratio. Thus, the strength reduction in the samples which contained over 30 wt.% of calcium carbonate powder could be attributed to poor mechanical performance of C-S-H with high Ca/Si ratio.