Microstructure and Properties of Sulfoaluminate Cement-Based Grouting Materials: Effect of Calcium Sulfate Variety

Effect of calcium sulfate variety on the microstructure and properties of cement-based grouting materials made by sulfoaluminate cement clinker, calcium sulfate resource, and quicklime was investigated in this paper. Isothermal calorimetric, XRD, DTA-TG, and SEM-EDS were used to characterize the reaction process of cement-based grouting materials. +e mechanical property and expansion of the harden body were also tested.+e results showed that (1) the formation of ettringite can be completed during the first 2 h when gypsum was selected, causing nonexpansion of the harden body, and ettringite tends to be precipitated near the clinker grain, which has a negative influence on the “skeleton effect” of ettringite, resulting in lower strength of the hardened body. (2) Ettringite can be continually and slowly precipitated when an anhydride was chosen, and the hardened body shows significant expansion characteristic. Besides, ettringite in anhydride paste tends to be precipitated in the form of complex interleaving, which can generate a positive influence on mechanical property of the hardened paste.

Ettringite is one of the most important hydration products of sulfoaluminate cement [11][12][13][14][15][16][17]. Sulfoaluminate cement has shown characteristics of rapid hardening, high early strength, and microexpansion which is mainly attributed to large number of ettringite formation in the hydration process [18][19][20]. Ye'elimite has a decisive role for the formation of ettringite in the system of sulfoaluminate cementitious material. ere are two ways for the formation of the ettringite crystal in the hydration process of ye'elimite [21,22]. First, ye'elimite in presence of a sulfate source and in aqueous medium will form long columnar ettringite and amorphous gibbsite (AH 3 ) according to equation (1) in the early hydration stage, causing rapid hardening and high early strength of the sulfoaluminate cement paste. Second, calcium oxide or calcium hydroxide can also react with aluminum hydroxide and calcium sulfate to form acicular ettringite (equation (2)), causing rapid coagulation and expansion characteristic. (2) According to the basic hydration characteristic of ye'elimite, the authors have used the sulfoaluminate cement clinker, natural anhydride, and quicklime to prepare a novel grouting material in previous studies, which had a significant rapid hardening, early strength, and microexpansion characteristics compared with Portland cement-based grouting material [23].
Calcium sulfate source is an important auxiliary mineral material, which can provide the necessary sulfate ions for the formation of ettringite in the ye'elimite hydration process. e ye'elimite will be completely converted to ettringite when the calcium sulfate is sufficient. Once the sulfate source is depleted, monosulfate (AFm) is formed according to equation (3). e amount of added calcium sulfate strongly modifies the ettringite to a monosulfate mass ratio, which will have a significant impact on the macroscopic properties of the material system. e authors' previous work also proved that the amount of calcium sulfate could affect the hydration process of ye'elimite and strongly influence the macroscopic properties of the sulfoaluminate cement clinker-natural anhydride-quicklime ternary system [23]. In fact, the calcium sulfate variety (gypsum, basanite, or anhydride) is another key issue to affect the precipitation of ettringite in addition to the influence of the calcium sulfate content [24,25]. is is mainly because different types of calcium sulfate source have significantly different dissolution and release rate of sulfate ions, which is strongly bound to affect the ettringite formation in the ye'elimite hydration process. However, the relevant research about the effect of calcium sulfate variety on the hydration of the C 4 A 3 S-CaSO 4 -CaO ternary material system is quite limited. Based on this situation, this paper mainly studies the effect of calcium sulfate variety on the hydration of the sulfoaluminate cement clinker-calcium sulfate-quicklime ternary material system. Due to the rapid condensation rate of basanite, it is unfavorable to the fluidity of grouting slurry [24]. erefore, gypsum and anhydride are chosen. rough the research work of this paper, the influence mechanism of a calcium sulfate variety on the hydration of the ultrafine sulfoaluminate cement clinker-calcium sulfate-quicklime ternary grouting material system will be revealed, and it also can provide some theoretical direction for the selection of calcium sulfate resource on the sulfoaluminate cement clinker-calcium sulfate-quicklime ternary grouting material system.

Materials.
e manufacturer, component, and particle size distribution of ultrafine sulfoaluminate cement (CSA) clinker and quicklime were described in literature [23] as the same. Pure gypsum was purchased from Jiaozuo gypsum Co. Ltd. Anhydrite was prepared by heating the pure gypsum at 700°C for 4 h [24]. Figure 1 shows the X-ray diffraction pattern (XRD) of two kinds of calcium sulfate resources. e pure gypsum and anhydride were then ground until they passed a 25 μm sieve. e dissolution rate of two kinds of calcium sulfate resources in saturated lime water at room temperature was tested by an ion exchange neutralization method when the water-to-calcium sulfate ratio is 100 : 5, given in Table 1.

Mix Proportion of Experiment.
e better proportion of between ye'elimite, calcium sulfate, and quicklime can be obtained based on the previous research [23]. Depending on the calcium sulfate variety, the corresponding specimen will be named hereafter CSA-G and CSA-A. e water-to-cement ratio was 1 : 1. Besides, the content of gypsum in CSA-G was increased appropriately in order to maintain the consistency of calcium sulfate content in the two groups. e water consumption was also reduced accordingly in CSA-G correspondingly.

Compressive Strength and Dimensional Stability Tests.
20 mm × 20 mm × 20 mm and 40 mm × 40 mm × 160 mm paste samples were prepared according to the mix proportion as shown in Section 2.2. After forming, the samples were cured under standard conditions immediately. e 20 mm × 20 mm × 20 mm samples were used to measure the compressive strength. After curing for 2 h, the 40 mm × 40 mm × 160 mm samples were removed from the mold, and the initial length was measured and recorded. e expansion ratio was tested by using the method described in literature [23].

Hydration Heat Test.
A conduction calorimeter ( ermometric TAM Air) was used to determine the rate of hydration heat liberation during the first 24 h. e specimen was added to a glass container after weighing. en, the glass container that contained the specimen was placed into the calorimeter. e corresponding amount of water was added when the temperature of specimen was constant at 20°C. e mixing was done by a small stirrer for 1 min. e total heat of hydration after 24 h was determined by integration of the heat flow curve.

X-Ray Diffraction (XRD) and DTA-TG Analysis.
e hydrated specimen was stopped by immersing in alcohol for 24 h. e hydrated specimens were dried by using a vacuum drying box. e temperature and the vacuum degree were 35°C and 0.08 MPa, respectively. en, the hydrated specimen was ground by hand below 0.063 mm. An X-ray diffraction analyzer (Smart-Lab, Cu kα radiation) was used to determine the mineralogical composition of the hydrated specimen. e test range was 5-80°at a scanning rate of 10°min. e DTA-TG curves were tested by using a Hengjiu synchronous thermal instrument with 10°C/min heating rate. e air atmosphere was selected to test the thermal decomposition characteristics of each hydration product, which can guide the thermal stability of grouting materials in practical application environment.

Scanning Electron Microscopy (SEM) Test.
Unground specimen was examined by scanning electron microscopy (ZEISS EVO18) using a secondary electron image and energy dispersive X-ray (EDS) analysis.

Isothermal Calorimetry.
e hydration exothermic evolution is shown in Figure 2. It can be seen from Figure 2(a) that the hydration exothermic rate can increase rapidly with the increase in hydration time in two specimens and reaching the maximum hydration exothermic rate in a very short time. Subsequently, the hydration exothermic rate decreased rapidly. e maximum exothermic rate of CSA-G is significantly higher than that of CSA-A. e second exothermic peak in CSA-A can be detected after 1.58 h. Figure 2(b) shows the cumulative exothermic curve, and it can be found that the cumulative heat release in the early stage of CSA-G is significantly higher than that in CSA-A. e total heat of hydration is 425.5J/g for CSA-A and 425.1J/g for CSA-G after 24 h, which are very close to each other. Figure 3 shows the X-ray diffraction analysis of CSA-G under different hydration ages. It can be seen that strong diffraction peaks of ettringite can be detected only after hydration for 10 min. ereafter, the diffraction peaks of ettringite gradually increased with the increase in hydration time, but the growth is small. e diffraction peak intensity of ettringite does not increase significantly after 2 h. e diffraction peaks of the gypsum crystal can no longer be detected after 2 h, indicating that it has been depleted. Similarly, the diffraction peak of ye'elimite disappears after 8 h. Figure 4 shows the X-ray diffraction analysis of CSA-A under different hydration ages. e characteristic peaks of ettringite are obviously weak after 10 min. However, the diffraction peak intensity of ettringite obviously increased with the extension of age. e diffraction peak of ye'elimite gradually weakens with time until it disappeared after 1 day of age. e diffraction peak strength of anhydride also decreases with age, but still can be detected after 28 d. e diffraction peak of gypsum is not detected, indicating that anhydride did not change into gypsum in all ages.

DTA-TG.
e DTA-TG data for CSA-G are shown in Figure 5.
e main hydration products are composed of ettringite, monosulfate, and aluminum hydroxide. A strong endothermic peak of ettringite can be detected only at 10 minutes of hydration. e endothermic peak of ettringite does not change significantly after 2 h. e endothermic peak of gypsum is near 150°C, which cannot be detected after 2 h. Besides, a small amount of monosulfate was detected at all ages. On the whole, the results of thermal analysis and XRD show the same rule. Figure 6 shows the DTA-TG curves for CSA-G at different ages. It can be seen that the main hydration products of CSA-A also included ettringite, monosulfate, and aluminum hydroxide. Unlike CSA-G, only little ettringite is detected after 10 min in CSA-A. However, the amount of ettringite can increase rapidly with the prolongation of age. e gypsum cannot be detected throughout the age period. ese thermal analysis results of CSA-A are also consistent with the XRD results. Besides, the monosulfate disappears after 7 d, possibly all of them converted into ettringite.
According to the method described in literature [26], the content of ettringite in two groups of specimen is calculated, which is shown in Figure 7. For CSA-G, the production of ettringite is as high as 67 wt. % after 10 minutes of hydration. e production of ettringite is the largest 2 h later, which is about 80.7 wt. %. ereafter, the amount of ettringite hardly changed with age. For CSA-A, the amount of ettringite is only 11.4 wt. % after 10 min of hydration, which is significantly lower than CSA-G. However, the production of ettringite in CSA-A can increase by a large margin. e amount of ettringite in CSA-A reaches the maximum until 7 d. e great difference in the formation law of ettringite in two groups may be related to the dissolution rate of gypsum and anhydride. e dissolution rate of gypsum is faster than that of anhydride, which may be the reason for the formation of more ettringite in CSA-G in the early age.       e strength of sulfoaluminate cement mainly comes from the formation of ettringite. In general, the more the amount of ettringite, the higher the mechanical strength in the sulfoaluminate cement system. However, it can be found that the evolution of compressive strength values of CSA-G and CSA-A is never in complete agreement with the ettringite formation in the early hydration stage, as shown in Figures 7 and 9. Although a large amount of ettringite formation occurs in CSA-G at early hydration process, the early strength values of CSA-G are obviously lower than that of CSA-A. In fact, the distribution of ettringite also affects the strength of the hardened body. For CSA-G, despite a large number of ettringite precipitated in the early age, the fine acicular ettringite tends to be congregated around the clinker grain, which may effectively reduce the bridging effect between ettringite crystals, producing a negative effect on the compressive strength of the hardened body. For CSA-A, although only a small amount of ettringite is formed at the early hydration process, the precipitated ettringite can be uniformly distributed in CSA-A, showing complex overlapping between each other and ensuring good "skeleton effect" of ettringite, which may have a positive effect on the compressive strength of the hardened body.

Expansion Test.
e expansion ratio results are shown in Figure 10. For CSA-A, the hardened body shows significant expansion characteristic, and the expansion ratio is gradually increased with the extension of the hydration age. However, it is 1 d or so later that the expansion ratio of CSA-A is no longer significantly changed. e hardened body for the CSA-G specimen only shows very slight shrinkage characteristics during the early hydration stage. en, the shrinkage ratio decreases in the later hydration stage. e hardened body had nearly the same volume as in initial time after 28 d of hydration. e expansion ratio depends on the amount of ettringite. In the CSA-A specimen, the ettringite crystal always can be

Conclusions
In this study, the effect of calcium sulfate variety on the microstructure and properties of sulfoaluminate cementbased grouting material was investigated. e chief conclusions to be drawn from the present study are as follows.
Due to the relatively faster dissolution rate of gypsum, the formation process of ettringite can be completed during the first 2 h of hydration when gypsum is selected as the calcium sulfate source, causing nonexpansion characteristic of the hardened body in the later stage of hydration. In addition, ettringite tends to be precipitated near the clinker grain, which may have a negative influence on the "skeleton effect" of ettringite, bringing about adverse effects on the mechanical property of the hardened body.
Anhydrite has a relatively slower dissolution rate; thus, ettringite can be continually precipitated with the extension of the hydration age when anhydride is chosen, effectively resulting in a significant expansion. Besides, the precipitated ettringite can be uniformly distributed, showing complex interleaving, which can guarantee good "skeleton bridging effect" of ettringite and has a positive influence on the mechanical property.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest. Advances in Materials Science and Engineering 7