Abstract
This work examines the influence of zeolite saturated with zinc (SZ) on the early hydration of Portland cement. The measurements and results indicate that the microcalorimetric method allows continuous observation and determination of SZ’s influence on kinetic processes in the early stages of hydration. According to the total heat released after 48 hours of hydration, higher SZ content results in lower heat values, while the maximum hydration occurs earlier. As SZ proportion increases, the rate of heat release and the degree of hydration decrease. Measured hydration degrees differ significantly from the calculated values when SZ is added to Portland cement, especially later in the hydration process.
1. Introduction
Cement production requires a large amount of electricity and is responsible for global CO2 emissions of 5–8%. CO2 is produced by the process of calcining limestone and burning fuel in the kilns [1]. One of the most effective methods of reducing greenhouse gas emissions and energy consumption in the cement industry is the partial replacement of cement with silicate and aluminosilicate materials [2]. As most of the supplementary cementitious materials (SCMs) are by-products of various industries, their use will not only reduce CO2 emissions and energy consumption but also prevent the disposal of these materials in the environment [3–5]. Furthermore, the use of SCMs has a positive effect on the properties of fresh and hardened concrete. Indeed, some SCMs improve the workability of fresh concrete, reduce leaching, and delay the setting time, which is particularly important in areas with high temperatures [6].
Zeolites are crystalline hydrated aluminosilicates of alkali and alkaline earth elements [7]. Recently, natural and modified zeolites have become very popular due to their ability to bind heavy metals and radioactive substances. The use of natural zeolites in the treatment of waste water contaminated with heavy metals is an essential aspect of environmental compatibility due to the economic efficiency and the excellent physical and chemical properties of zeolites [8]. Once zeolite is saturated with heavy metals, it becomes a harmful waste itself and should be treated in a way that does not endanger the environment [9]. A practical method of disposing saturated zeolites is to dispose them in a cement matrix [10–12]. However, the effect of zeolite saturated with heavy metals on the hydration process of Portland cement is still not clear.
The hydration of Portland cement is a series of chemical reactions between the clinker minerals, gypsum, and water, which occur simultaneously or at different rates. As the hydration process begins, the cement paste begins to thicken until it finally hardens. The hydration reactions continue until the reactants or the free space for the formation of hydration products disappear [13].
The progress of hydration is reflected in the development of the chemical, physical, mechanical, and electrical properties of the cement paste. Therefore, the progress of hydration can be monitored by measuring changes in chemical composition, enthalpy, ultrasonic velocity, volume changes, and changes in the strength of hardened cement paste [14]. A common method for monitoring the progress of the hydration reactions of Portland cement is to register the heat development curves [15, 16]. These curves do not provide a representative answer to mechanical questions but are useful to correlate experimental data with respect to an individual parameter, i.e., heat development. The hydration process is exothermic, which allows the monitoring of hydration reactions with calorimetric methods [17, 18]. Years of experience in the field of hydration process analysis of complex disperse systems offer the possibility to extend the application of calorimetric methods in the control of raw material quality and the control of properties of building composites [19, 20].
2. Development of a Mathematical Model for Kinetic Analysis of Portland Cement Hydration
According to the Krstulović–Dabić model, the hydration reaction of the cement composite can be divided into three basic processes, namely: nucleation and crystal growth (NG), interactions at phase boundaries (I), and diffusion (D) [21]. All three processes occur simultaneously, but the slowest determines the speed of the entire hydration process [22]. Therefore, it is necessary to determine the speeds of all three processes. Observations from previous studies suggest that nucleation and crystallization are the slowest processes at first, while interactions at phase boundaries and diffusion become more important over time [23–25].
In the nucleation and growth process, the basic Avrami–Erofeev equation describing hydration kinetics is often used for spherical particles and is represented by [26]where α is the reaction degree of reactants, KNR is the nucleation and growth constant, and t is time. The exponent n describes the geometry of crystal growth and usually falls in the range of 1 to 3.
During the second stage of hydration, the hydration reaction slows down, and the total rate is determined by the reaction interactions at the phase boundary, which can be shown bywhere KI is the phase boundary interaction constant.
In the third stage of hydration, when the process is further slowed down due to diffusion through the layer of products formed around the anhydride of the cement particle, the total rate is determined by the diffusion process described by the Jander relation:where KD is the phase boundary interaction constant.
α-t functions are considered the basis for the development of a mathematical model. Differential kinetic forms of the equations are obtained by deriving integral α-t functions, and the integral functions represent partial solutions. Each differential equation describes the change in hydration rate for each individual process defined in the time interval 0-∞. This paper focuses on the influence of zeolite saturated with zinc ions on the hydration kinetics of Portland cement composites and will contribute to the understanding of the hydration mechanism of cement with different additions of saturated zeolite.
3. Experimental
For all investigated samples, an ordinary Portland cement OPC type CEM I 42.5R was used. OPC was purchased from the cement plant CEMEX, Kaštel Sućurac, Croatia. Natural zeolite saturated with zinc ions from the Donje Jesenje site in Croatia was used as supplementary material to the cement. The saturation was performed with a solution containing 9.0 mmol dm-3 ZnSO4. Zeolite contains the main mineral clinoptilolite and impurities such as illite, feldspar, and quartz (Figure 1). The saturated zeolite (SZ) was dried to a constant weight at 60°C, ground, and sieved through a 90 μm sieve. The chemical compositions of OPC and natural zeolites are shown in Table 1. Natural zeolite (NZ) consists of 64.93 wt. % SiO2 and 13.66 wt.% Al2O3 and with such a composition can be classified as a potential pozzolanic material (the composition of silicon and aluminum should be over 70%).
Mix proportions of cement paste are shown in Table 2.
The mark C indicates OPC without addition, while C10SZ, C20SZ, and C30SZ are short for the cement samples containing 10, 20, and 30 wt.% saturated zeolite, respectively. The effect of SZ on hydration processes was determined by a differential microcalorimeter of the conduction-isoperibolic type at 20°C and water/binder ratio (W/B) 0.5. An ALMEMO 2390-8 data logger, Ahlborn, Germany, was used to trace the changes of voltage dU = f(t) due to changes of temperature in the samples during the first 48 hours of hydration. Computer programs calculate total heat released, heat release rate, and relative reaction degree from the measured thermovoltage values.
4. Results and Discussion
Figures 2 and 3 show the curves for hydration heat released rate and total heat of hydration of Portland cement samples with SZ.
Table 3 shows the end of the induction period and the duration of the acceleration period. Table 4 shows the maximum rate of hydration heat released and the total amount of heat released at different hydration times.
Based on Figure 2, all samples undergo similar hydration processes. Comparing the results of the influence of natural zeolite and zeolite saturated with zinc on hydration, both curves retain a similar shape, which indicates that neither natural zeolite nor saturated zeolite affects main hydration mechanisms. In terms of decreasing the amount of heat released and the rate of heat release, saturated zeolite has a greater impact [11, 28]. After mixing binder and water, a very sharp exothermic peak occurs due to the rapid dissolution of the clinker minerals and the rapid formation of ettringite and syngenite. The period after the initial period is the induction period, during which the hydration reaction begins to slow down. This period provides information on the workability time. The heat value drops sharply and remains constant for several hours (Figure 3).
Based on Table 3, sample C (pure Portland cement) requires an induction period of about 3.15 hours, whereas samples C10SZ, C20SZ, and C30SZ require about 2.26, 2.95, and 3.08 hours, respectively. In the hydration acceleration period, the first peak corresponds to the beginning of setting, and the second peak corresponds to the end of setting and the beginning of hardening of the cement paste. As can be seen in Table 3, adding SZ to the cement-water system does not affect the beginning of the acceleration period or the appearance of the main peak. It only affects the intensity of the main peak (by increasing the SZ, there is a decrease in the heat release rate), as can be seen in Figure 2. As a result of the SZ addition to the acceleration period, the Ca2+ ions are deposited on either the surface of the SZ or in the form of a C-S-H gel. The maximum value of the main peak decreases by 23%, from 9.10 J −1h−1 to 6.97 J g−1h−1 after incorporating 10 wt. % SZ. When the SZ content increases to 20 and 30 wt. %, the maximum value of the main peak is decreased by 6 and 13%, from 6.97 J g−1h−1 to 6.52 J g−1h−1 and from 6.52 J g−1h−1 to 5.67 J g−1h−1. The low hydration activity of SZ may be the main affecting factor.
Figure 3 shows that the total heat of hydration decreases with the increase of SZ content. The 12 hours’ heat released decreases by 24% from 74.50 J g−1 to 56.75 J g−1 after addition of 10 wt. % SZ. When the SZ content increases to 20 and 30 wt. %, the heat released decreases by 1% and 6% from 56.75 J g−1 to 56.03 J g−1 and from 56.03 J g−1 to 52.84 J g−1. The 48 hours’ heat released decreases by 14%, 8%, and 11% from 188.28 J g−1 (C) to 162.52 J g−1 (C10SZ), 148.93 J g−1 (C20SZ), and 132.04 J g−1 (C30SZ).
5. Hydration Kinetics Model of Portland Cement-Saturated Zeolite System
The mathematical model is based on the assumption that three processes occur simultaneously on the surface of a cement particle: nucleation and growth, phase boundary interaction, and diffusion. The slowest process controls the overall hydration rate. Using the assumed model, it is possible to estimate the values of the kinetic parameters of the hydration process in the cement-saturated zeolite system. The estimated values of the kinetic parameters are shown in Table 5. Theoretical and experimental hydration curves of the four samples are shown in Figures 4–7.
From Table 5, it can be seen that the value of the nucleation and growth constant (KNR) for sample C (Portland cement without addition) is 0.0480 h−1, and that for samples C10SZ, C20SZ, and C30SZ is 0.0390 h−1, 0.0370 h−1, and 0.0340 h−1. These results indicate that increasing the addition of saturated zeolite slows down the nucleation and growth process. The value of the exponent n for Portland cement without addition (C) is 2.20, while the samples containing saturated zeolite (C10SZ, C20SZ, and C30SZ) have a much lower value of this exponent. This indicates that the addition of saturated zeolite greatly affects the crystal growth geometry. The rate of the phase-boundary interaction process is described by the KI constant. The value of the KI constant for sample C is 0.0249 h−1. At lower zeolite addition (C10SZ), the KI constant assumes a lower value indicating a decrease in the rate of hydration processes when the interaction process at the phase boundary is a controlled process.
Increasing the content of saturated zeolite (C20SZ and C30SZ), the rate of hydration process increases. From the results of determining the transition time of tNR-I from one controlling process to another (NR⟶I), it is evident that in the system without the addition of saturated zeolite, the transition time is tNR-I = 9.47 h. In a system with the addition of 10 wt. % saturated zeolite, the duration of the NR process is shortened, while the effect of prolonging the duration of NR was observed in the system with increased addition. The process that follows the phase boundary interaction process is the diffusion process, which is described by the KD constant. To start the diffusion process, the phase boundary interaction process rate and the diffusion rate must be equal. The value of the KD constant for sample C is 0.0037 h−1, while the addition of saturated zeolite increases the KD value and it can be concluded that the addition of saturated zeolite increases the diffusion rate. The transition time tI-D for cement without admixture is 26.84 h, while for samples C10SZ and C20SZ, it is 28.42 h and 27.63 h. Sample C30SZ has a similar transition time tI-D as sample C.
The agreement of the experimental data α-t with the mathematical model is shown in Figures 8–11.
Considering the experimental data and the estimated kinetic model in Figure 8, the deviation of the model from the experimental data can be observed, especially in the area of the interaction between the phase boundaries. In Figures 9–11, this deviation is even greater, especially in the areas of interaction and diffusion. The proposed kinetic model can be applied to Portland cement without additives, while with the addition of saturated zeolite, some modifications of Jander’s model should be conducted to ensure the scientific and accurate characterization of the hydration of cement [29, 30].
6. Conclusions
Based on the measurements and the obtained results, it can be concluded that the microcalorimetric method allows continuous monitoring and determination of the influence of saturated zeolite on the kinetic processes of hydration in the period of early hydration. After 48 hours of hydration, the total heat released shows that higher saturated zeolite content leads to lower heat values, while the occurrence of maximum hydration time occurs earlier. The rate of heat release and hydration degree decrease as the proportion of saturated zeolite in the sample increases. Significant deviations are observed between the measured and calculated values of the hydration degree when saturated zeolite is added to Portland cement, especially in later hydration times.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.