Effect of Calcium Chloride on Hydration Kinetics and Pore Structure of Hydrated Tricalcium Silicate

Chemical admixtures are frequently used to regulate the setting and strength development of concrete materials. In this study, tricalcium silicate (C3S) was used as a model of the cement system, and the influence of calcium chloride, an extremely useful accelerator, on C3S hydration and the pore structure of hardened C3S paste were investigated by the combination of the techniques of differential scanning calorimetry (DSC) and the N2 adsorption (BET). The results indicated that the addition of calcium chloride would significantly shorten the pre-induction and induction periods and enhance the specific surface area and porosity of hardened C3S paste. However, the presence of CaCl2 has little effect on the pores, with a width ranging from 2.5 nm to 5 nm. DSC technique has an advantage of measuring continuously the process of C3S hydration by changes of free water in hydrated C3S.


Introduction
Calcium chloride (CaCl2) is an extremely useful accelerator of tricalcium silicate (C3S) and cement hydration [1]. This accelerator is still frequently used in the unreinforced concrete. The higher the dosage of CaCl2, the more significant the accelerating effect. CaCl2 could promote the dissolution of cement, accelerate the formation of hydration products, and increase the heat flow rate of C3S hydration [2][3][4].
According to the heat flow curve, the hydration of C3S is generally divided into four stages:(1) pre-induction period; (2) induction period; (3) acceleration period; (4) deceleration period [5]. The hydration of C3S would release abundant heat, and produce amorphous calcium silicate hydrates and crystalline calcium hydroxide [6,7]. The evolution of hydration heat flow is frequently used to represent the hydration kinetics of C3S [8].
Calorimetry is a commonly used and effective method to measure the exothermic properties of cement hydration reaction, including heat of solution calorimetry, semi-adiabatic/adiabatic calorimetry, isothermal conduction calorimetry and differential scanning calorimetry (DSC) [9]. The approach for DSC is to measure the changes of free water in C3S and cement hydration, which can continuously measure the hydration regardless of the amount of heat released. In the study, with the same method, the influence of CaCl2 on C3S hydration was investigated.
With the progress of C3S hydration, the number of gel pores from C-S-H gels [1,10,11] would increase, whereas that of capillary pores would decrease. Since the capillary pores are the main channel of water transportation, it closely related to the deterioration of cement-based materials caused by ingress of foreign ions [12]. Therefore, the study of the pore structure of hydrated C3S is helpful to understand the C3S hydration products, microstructure and performance.
water with a liquid-to-solid ratio of 0.4 and 0.5, and CaCl2 was added with 1% wt of anhydrous C3S. Hydrated C3S pastes were placed into CryoTube vials (1.8mL) under N2-gas protection at room temperature and then were sealed in a vacuum bag. The samples of hydrated C3S without and with CaCl2 were respectively dried in a vacuum oven at 23℃ after 1, 3, 7 and 28-day hydration.
X-ray Diffraction. The X-ray diffraction tests of samples were conducted on a PNAalytical X'Pert Pro diffractometer equipped with a Co-target X-ray tube (λ=1.79 Å) and rapid X'celerator detector. The 2-theta ranges from 5° to 55°with a step of 0.02°. X-pert High Score Plus software was used to analyze the XRD patterns of the samples.
Differential Scanning calorimetry (DSC). Differential scanning calorimetry measurements were performed using a DSC Q2000 from TA Instruments. Around 50 mg of sample was mounted into an aluminum pan (ϕ5.4*2.0 mm), and then was sealed with the appropriate cover to avoid the leaking of water. Then, to the samples were equilibrated to -30℃ and kept at this temperature for 1 min, after which the samples were heated from -30℃ to -12℃ at 20℃/min and -12℃ to +25℃ at 4℃/min. The collected data were analyzed by a Q series software (version 5.5.22).
N2 adsorption (BET). The pore distribution of samples was measured by an accelerated surface area and porosimetry system (ASAP 2020) from Micromeritics (U.S.). The test temperature is around room temperature.

Results and Discussion
XRD Patterns. Hydration reaction of C3S hydration would form amorphous C-S-H and crystalline calcium hydroxide (Portlandite), which generally follows the following formula: (1) Allen et al. [13] determined x and y values in the above formula by using small-angle X-ray scattering and small-angle neutron scattering, and obtained the following equation: (2) It can be seen from Fig. 1 that CH and C-S-H (a hump at the d-spacing of 3.08 Å) are the only hydration products, and that CH was formed after 1-day hydration and the C3S were almost consumed after 28 days. This means that the addition of CaCl2 did not change the types of hydration products, and no new substances were produced.
Hydration kinetics of C3S. In this paper, the evolution of free water content (Free Water Index, FWI) was chosen to represent the hydration kinetics of C3S, by the technique of DSC which allows the quantitative analysis of free water content by integrating the melting peak of free water [9,14]. Fig. 2 presents the heat flow curves of hydrated C3S pates as a function of hydration time. With these heat flow curves, FWI can be calculated according to the following formula: FWI=∆H exp /ϕ w ΔH theor (3) where ∆H exp --the area of the water melting peak, J/g; ϕ w --the weight fraction of water in the paste; ΔH theor --the theoretical value of melting enthalpy of water in the C3S/water paste, 333.4J/g; The changes of free water in C3S pastes of the melting peak of C3S pastes with w/c of 0.4 and 0.5 over time is shown in Fig. 3. It can be seen that the addition of CaCl2 would shorten the pre-induction and induction period, which is consistent with the results in the literature. In the absence of CaCl2, where would be an induction period of around 3 hours, this period would almost disappear in the presence of CaCl2. For the system with a w/c of 0.4 as shown in Fig. 3b, similar results can be found that the addition of CaCl2 would accelerate the assumption of free water. Interestingly, it can be found that the FWI in the systems with a w/c of 0.4 are lower than those with a w/c of 0.5 regardless of the presence of CaCl2.
CaCl2 can significantly increase the nucleation rate of hydration products on the surface of C3S particles and has little influence on the growth rate [15]. The addition of would shorten the preinduction and the induction period as shown in Fig. 3. This means that CaCl2 would promote the dissolution of C3S particles and the release of Ca 2+ and SiO4 into the pore solution, and finally accelerate the saturation or supersaturation of hydration products. The critical effect of CaCl2 is to trigger the initiation of acceleration, to accelerate the hydration products of nucleation [16]. Analysis of Pore structures. N2 adsorption is a commonly used method to measure the specific surface area and pore size distribution of cement-based materials, especially for the determination of relatively small gel pores [17]. Hydrated C3S paste is porous, in which there are plenty of pores with the diameters ranging from nanometers to microns, such as gel pores and capillary pores. Fig. 4 presents the adsorption isotherm of C3S pastes in the absence and presence of CaCl2. From figure 4, it can be seen that the maximum adsorption of CaCl2-free samples reaches up to 58.8 cm3/g Seminar on Advances in Materials Science and Engineering STP at 28 days, while the 7-day maximum adsorption amount of samples is ranging from 20 cm3/g STP to 30 cm3/g STP. C-S-H gel has a high specific surface area, and these amorphous gels stack and is rolled to form gel pores. As the hydration time increases, more foil-like and layer-structure C-S-H gel are generated, increasing N2 adsorption. According to the International Union of Pure and Applied Chemistry (IUPAC) classification of N2 adsorption isotherms, the shapes of hysteresis loops of the samples with different hydration time are shown in Fig. 4. It can be seen that hysteresis loop type of the pores in the hydrated C3S sample transform from H4 type to H3 type with hydration. Hysteresis type H4 is associated with non-rigid aggregates of plate-like particles (e.g., calcium hydroxide), while hysteresis type H3 is associated with being filling with micropores, aggregated hydration product of C-S-H gels. The pores in C3S samples hydrated for 28 days show H3 hysteresis ring, and the adsorption amount was enormous when the relative pressure was high. The slit pores formed by the accumulation of flake particles are mostly laminated Ca(OH)2 crystals and C-S-H layered structure. Fig. 4 The adsorption isotherm of C3S in the absence (a) and presence (b) of CaCl2 The specific surface areas of Langmuir, BET, t-plot and BJH adsorption/desorption are shown in Table 1. The specific surface area of C3S pastes gradually increases with hydration time, which reflects the increase of C-S-H hydration products in the sample. Noticeably, the specific surface area of C3S pastes with CaCl2 is all larger than that without addition after the same hydration time. Also, Langmuir specific surface area is higher than the BET specific surface area. Theoretically, Langmuir adsorption is based on single-layer adsorption and focuses on measuring smaller pores, while BET specific surface area is based on multi-layer adsorption, which is suitable for testing larger pores. In Fig. 5, the pore diameter of the hydrated C3S is mainly distributed between 2.5 nm and 5nm at different hydration times. Furthermore, regardless of the presence of CaCl2 or not, the average pore diameter of the gel pores in the hydrated C3S samples are about 3.5nm. Although CaCl2 can accelerate the hydration reaction of C3S and produce more hydration products in the early stage, it has little effect on the size of the gel pores.

Summary
Differential scanning calorimetry offers the possibility of studying the effect of CaCl2 on the hydration kinetics of C3S hydration. Changes in free water in hydrated C3S pastes can easily be determined continuously by DSC.
The presence of CaCl2 does not change the types of hydration products, which is mainly C-S-H gel and calcium hydroxide. CaCl2, as an inorganic chloride salt, can shorten distinctly the pre-induction period and induction period during the hydration reaction.
As the progress of hydration, the specific surface area of hydrated C3S pastes gradually increase due to the formation of hydration products, such as C-S-H gel and calcium hydroxide, and the addition of CaCl2 would accelerate this increasing rate. Based on the shapes of H4 hysteresis loops of hydrated C3S pastes, the pores in hydrated C3S are of slit pores ranging from macropores to mesopores due to the generation of C-S-H gel with a layer structure and plate-like crystal calcium hydroxide. However, the presence of CaCl2 does not affect the pore width of mesopores of hydrated C3S, which is ranging from 2.5 nm to 5 nm.