Effects of Tuff Powder on the Hydration Properties of Cement-Based Materials under High Temperature

Abstract

In some instances, traditional mineral admixtures, such as fly ash and slag, have been insufficient, and tuff powder (TP) has been used as an alternate replacement. However, the mechanisms of the hydration of tuff powder have rarely been studied, which has restricted its application; therefore, this paper studied the hydration mechanisms of a cement–TP composite under different temperatures. In this study, the influence of TP on the hydration properties of cement-based materials under different curing temperatures was investigated by a compressive strength test, X-ray diffraction (XRD), and thermogravimetric and differential thermal analysis (TG–DTA). Our results showed that a high curing temperature effectively promoted the hydration of Portland cement and the pozzolanic reaction degree of TP and improved the mechanical and microstructural properties of cement-based materials. The high temperature was most conducive to the early development of strength. Additionally, different water-to-binder ratios showed different effects on the contribution coefficients of high curing temperatures. The effects of a high temperature on the pozzolanic reaction of TP may lead to greater Ca(OH)₂ consumption during hydration. This research provides a new way to improve the reaction activity of TP and lays a theoretical foundation for applying TP to precast concrete products, mass concrete, or concrete construction in hot seasons.

1. Introduction

In hydraulic concrete, high-quality mineral admixtures are commonly used and are referred to as the sixth component of concrete [1]. Adding mineral admixtures can reduce the dosage of cement and effectively improve the compactness of the concrete, reduce the hydration heat of the cementitious system, and improve the workability and durability of the concrete. In addition, the application of mineral admixtures has excellent environmental and economic benefits [2,3].

With the continuous expansion of infrastructure construction in China, the demand for mineral admixtures is increasing. However, in some underdeveloped areas, there is a lack of traditional mineral admixtures, such as fly ash and slag. Project construction costs in these areas have dramatically increased due to the long-distance transportation of traditional mineral admixtures. Thus, exploring new and local mineral admixtures to replace or partially replace traditional mineral admixtures is essential and can consequently ensure a project’s smooth development and effectively reduce construction costs [4,5,6].

Tuff powder (TP) is a widely distributed fine-grained volcanic clastic rock. Its main composition is volcanic ash, which has a loose and porous appearance. Its moderate hardness enables its mining and processing [7]. TP has been widely studied as an artificial aggregate [8,9,10]. Tuff is a stone formed from volcanic ash that contains pozzolanic material that reacts with calcium hydroxide (CH) to form calcium silicate hydrate (C-S-H) gel. The Al content in tuff is high, and can be dissolved and fixed in cement hydrates. Therefore, tuff powder can be used as an alternate mineral admixture of concrete [11]. To date, TP has been used as a mineral admixture in several hydraulic projects in China [12,13,14]. Studies have shown that the incorporation of pozzolanic materials (such as TP) into cementitious materials could improve the particle size distributions of cementitious materials. The activity and properties of TP can be improved by calcination at high temperatures and by grinding to a higher specific surface area (e.g., from 590 m²/kg to 805 m²/kg) [15]. In addition, with the increase in tuff powder content, the hydration rate decreases significantly with the heat release peak delayed, which is highly favorable for the temperature control of mass concrete [15].

Although research on the hydration mechanisms of cement and traditional mineral admixtures has become popular, studies on TP are still limited. Former investigations about TP were mainly focused on its pozzolanic properties and its influence on the workability and mechanical properties of cement-based materials [16,17,18,19]. For example, the effects of TP content, particle fineness, and water-to-binder ratio on the properties of cement-based materials have been widely studied [4,15]. Adding a proper amount of TP can also shorten the setting time, reduce the dry shrinkage rate, and improve durability [20,21]. However, compressive strength is known to decrease with the increase in TP content, which is explained by the reaction activity of TP being low compared with cement [14,19]. In a study by Peng [14], the 28d compressive strength was 54.2MPa, and reduced to 38.4MPa and 20.3MPa when the TP content increased from 0% to 20% and 50%. These studies mainly stay at the macro-strength performance level. On the micro-level, hydration mechanisms are rarely involved [5,15]. Furthermore, the effects of curing temperature on TP in cement-based materials remain to be studied. Curing temperature has an important influence on the hydration process of cement-based materials and may affect the hydration mechanisms of mineral admixtures. Therefore, understanding the role of TP in cement-based materials and its influencing factors is necessary and can be used to guide the application of TP in concrete.

In this study, the effects of high curing temperatures on the strength and microstructure of cement–TP composite cementitious materials were investigated. Considering the high temperatures of summer, 40 °C was chosen as the steam curing temperature, and a standard curing of 20 °C was used as a reference. The strength of the cement pastes with different TP contents at different curing temperatures was studied. Furthermore, the microstructure was studied using XRD and TG–DTA.

2. Experimental Procedure

Ordinary Portland cement PO 42.5 was used during the experiment, complying with the Chinese standard GB175-2007. The TP used in the experiment was obtained from tuff stone from Fujian Province, which was ground in a ball mill for 0.5 h after being washed, dried, and crushed. As shown in Figure 1, the microscopic morphology of the TP presents irregular polygonal debris-like particles [22], which is different from the spherical morphology of fly ash. The chemical compositions of cement and TP are shown in

Table 1. Chemical compositions of raw materials, mass/%.

Compositions	CaO	SiO2	Al2O3	Fe2O3	SO3	MgO	TiO2	K2O	Na2O	P2O5	MnO	Loss
Cement	57.95	25.67	4.89	3.29	3.29	0.15	0.35	0.76	0.28	0.06	0.13	2.15
Tuff powder	1.01	72.35	14.25	1.20	0.13	0.71	0.38	4.43	2.85	0.09	0.09	0.71

. It is noted that the chemical composition of TP is similar to that of pozzolanic materials. The main chemical components of TP are SiO₂, Al₂O₃, K₂O, and Na₂O. According to ASTM C618-02, one of the necessary conditions for pozzolanic materials is that the total content of SiO₂, Fe₂O₃, and Al₂O₃ should exceed 70%. Thus, the TP used in our study met the requirements for pozzolanic materials. In addition, TP contains many alkalies, i.e., Na₂O and K₂O, which may theoretically activate the hydration of silica.

The mixing ratio is shown in

Table 2. Mix proportions of the cement pastes/g.

Samples	Cement	TP	Water
H-0	800	0	400
H-15	680	120	400
H-30	560	240	400
H-45	440	360	400
L-0	800	0	240
L-15	680	120	240
L-30	560	240	240
L-45	440	360	240

. According to the literature reviewed and preliminary tests, too high a TP content will lead to a serious decline in compressive strength. Therefore, the TP content was controlled within 0–45%. Two water-to-binder ratios (W/B), 0.3 and 0.5, were selected. This was mainly because the preliminary experiment found that with a W/B higher than 0.5, cement paste is likely to produce bleeding, and with a W/B ratio lower than 0.3, the uniformity of cement pastes is poor. The paste specimens of 40 × 40 × 40 mm³ with W/B of 0.5 (H group) and 0.3 (L group) were prepared and cured for 24 h under >90% RH and 20 ± 1 °C before demolding. After that, the specimens were cured in different conditions: Half were cured at 20 °C, and the other half were cured at 40 °C in a steam curing box. All specimens were cured for 3, 7, 28, and 90 days. The compressive strengths of the pastes at different ages were tested by a WAY-2000 according to GB/T17671-1999. After the sample was cured to a predetermined age, the H group samples were broken with a small hammer, and the pieces were collected and put in a bottle filled with ethanol [23]. By replacing free water in the samples, ethanol can be used to stop hydration [24]. The samples were taken from the ethanol, rapidly ground into powder in an agate mortar, and were then dried in a vacuum desiccator to reduce the effect of carbonization. Samples were examined by X-ray diffraction (XRD) and thermogravimetric and differential thermal analysis (TG–DTA).

The morphology of the specimens was observed with an SEM (JSM-5610LV, Tokyo, Japan). XRD was performed by an X’Pert Pro (PANalytical, Almelo, The Netherlands) at a scanning speed of 10°/min. The TG–DTA was analyzed using a Diamond TG/DTA analyzer (Perkin Elmer Instruments Plant, Shanghai, China), and the temperature increased from 0 °C to 900 °C.

3. Results and Discussion

3.1. Compressive Strength

Table 3. Compressive strength of different samples/MPa.

Samples	20 °C				40 °C
	3 Days	7 Days	28 Days	90 Days	3 Days	7 Days	28 Days	90 Days
H-0	25.35	29.43	48.21	57.85	31.06	38.64	54.97	67.01
H-15	18.78	28.30	39.02	50.72	26.79	34.25	45.22	53.52
H-30	14.97	22.60	29.58	39.05	19.21	24.86	31.86	40.14
H-45	8.89	15.35	18.85	24.88	11.75	18.62	22.41	30.11
L-0	38.87	46.45	60.69	71.77	52.64	56.39	73.84	89.55
L-15	26.57	31.52	43.95	58.01	35.33	43.51	58.94	76.36
L-30	23.93	30.17	38.27	48.21	28.08	34.56	45.25	57.59
L-45	18.45	22.15	28.95	38.21	20.64	22.61	31.15	42.05

shows the compressive strength of the pastes. It is clear that the compressive strength of the samples increases with age. In addition, the compressive strength of the cement paste decreased with the increase in TP. This is consistent with a previous study by Li [19] in which the 90d compressive strength of the pure cement paste was 57.9 MPa but reduced to 39 Mpa with a TP content of 30% and 12 Mpa when the value increased to 45%. Overall, the compressive strength decreased with the increase in TP content, which may be because the reaction activity of TP is low compared with cement [14,19]. In addition, the compressive strength of each group after steam curing was higher than after the standard curing for the same period. This result is understandable because, according to the chemical equilibrium theory, the hydration reaction of cement and the pozzolanic reaction of TP would increase with the curing temperature.

In order to quantitatively analyze the effects of a high temperature on the strength of the pastes, a steaming index γ, which reflects the change rate of strength under steam curing, is introduced and calculated as shown in Equation (1). Here, R_s is the compressive strength of the samples under 20 °C (standard conditions) and R_h is that under high temperature steaming conditions. When γ > 100%, this indicates that the strength of the samples increased under steam curing. The results of the steaming index are shown in Figure 2.

γ = (R_s/R_h) × 100%

(1)

The compressive strengths of both groups after steam curing improved compared with the standard curing. However, the effects of different water-to-binder ratios (W/B) on the contribution coefficients of steaming were different. For the samples at water-to-binder ratios of 0.5 and 0.3, the steaming index of cement pastes reached the maximum value at 7 days and 3 days, respectively. As steam curing accelerates a cement specimen’s reaction rate, the water consumption speed in the cementitious system is also accelerated. With a lower water-to-binder ratio (0.3), it is more likely that the water is insufficient for cement hydration. Accordingly, the strength development of the samples was affected due to insufficient water. The water content in the paste with a water-to-binder ratio of 0.5 was higher, so the paste had a long strength development. At the later hydration stage, the steaming index of the pastes showed little change, which may be because the early hydration reaction of ordinary Portland cement under steaming is so intense that a large number of hydration products were produced in a short time. Accordingly, low-density hydration products with a looser microstructure were produced, adversely affecting strength development. The high steaming index in the early stage and the decline in the later stage are consistent with the conclusions of other researchers [14,25].

With the incorporation of TP, the steaming index decreased over time, and the tendency at early stages was similar for different water-to-binder ratios. Overall, the samples with a water-to-binder ratio of 0.5 obtained a higher steaming index. The higher the TP content in the pastes, the lower the steaming contributed to the strength. As the water absorption rate of TP is high, hydration reactions cannot be provided with sufficient water, which results in a low steaming index. However, the contribution of steaming at an early stage is prominent, even more so than it is for pure cement paste. This contributed to a high hydration rate with sufficient water at the early stage and to the high activity of TP under steam curing.

3.2. Hydration Products

3.2.1. X-ray Diffraction (XRD)

Figure 3 shows the hydration products of cement specimens under different curing conditions. The intensities of Ca(OH)₂ characteristic diffraction peaks in the pure cement paste with steam curing increased significantly, which indicates that steam curing effectively promotes cement hydration and accelerates the formation of hydration products.

Figure 4 shows the hydration products of the samples with TP under different curing conditions. Contrary to pure cement paste, the intensity of the Ca(OH)₂ characteristic diffraction peak is weakened after steam curing, and attenuation also increases with the increase in TP content, which may be related to the pozzolanic reaction of the TP. This is consistent with the observation by Liu [15] that with the increase in TP content, the peak intensity of CH decreases, indicating that the hydration of TP consumes CH and produces C-S-H.

High curing temperatures promoted the hydration of Portland cement in the system, with more Ca(OH)₂ generated. High temperatures also stimulate the pozzolanic reactions of TP and with Ca(OH)₂ [26]. When the consumption amount of Ca(OH)₂ is higher than the Ca(OH)₂ generated, the amount of Ca(OH)₂ decreases, and thus the intensity of the corresponding characteristic diffraction peak is weakened. XRD analysis may also indicate that high temperatures contribute to the improvement of the pozzolanic activity of TP.

3.2.2. Thermogravimetric and Differential Thermal Analysis (TG–DTA)

Generally, all samples showed similar TG–DTA curves (Figure 5), which indicates that TP has little effect on the type of hydration products of the cementitious materials [27]. In the references, the DTA peaks were located at around 120 °C, 155 °C, 450 °C, 520 °C, and 750 °C. The peaks at 120 °C and 155 °C were mainly due to the dehydration of C-S-H, AFt, and calcium aluminum oxide hydrate. The peaks at 450 °C and 520 °C were due to the decomposition of CH and calcium aluminum oxide hydrate. The peak at 750 °C was caused by the phase changing of cement [15]. The following conclusions can be reached by combining the TG–DTA images of this experiment: A pronounced endothermic peak was found between 400 °C and 500 °C in the DTA curve, which was caused by the decomposition of Ca(OH)₂. The corresponding TG curve observed a significant weight loss slope in this temperature region. Some samples showed weak endothermic peaks near 700 °C, which were caused by the decomposition of CaCO₃. In addition, some samples showed weak endothermic peaks with mass loss at around 110 °C, which may have contributed to water evaporation and the dehydration of ettringite. However, the peak shape characteristics were not prominent, indicating a low ettringite content.

The quantity of some hydration products can be calculated according to the TG–DTA curve. Based on this calculation, the effects of curing temperature on the hydration of cement-based materials can be indirectly analyzed [15,27].

(a)

There is a mass loss and an endothermic effect caused by the dehydration decomposition of Ca(OH)₂ at 400–500 °C, so Ca(OH)₂ content in the system can be quantitatively calculated. The reaction equation is as follows in Equation (4). Here, W_CH is the mass of Ca(OH)₂, and W_H is the mass of water.

Ca(OH)₂ → CaO + H₂O

(2)

The value corresponds to the mass of the water lost by Ca(OH)₂, which can be used to calculate the Ca(OH)₂ content based on Equation (2), that is, W_CH = 4.111W_H.

(b)

The CaCO₃ in the system is also converted from the carbonization of Ca(OH)₂ produced by hydration, so Ca(OH)₂ involved in the conversion reaction should also be calculated. The CaCO₃ decomposition near 700 °C is as follows in Equation (3). Here, W_C is the mass of CaCO₃, and W_CO2 is the mass of CO₂.

CaCO₃ → CaO + CO₂

(3)

The mass loss is the mass of CO₂, and the carbonization of Ca(OH)₂ is as in Equation (4).

Ca(OH)₂ + CO₂ → CaCO₃ + H₂O

(4)

The relationship between Ca(OH)₂ and CO₂ mass is based on Equation (4), that is, W_CH = 1.682 W_CO2.

(c)

The total Ca(OH)₂ content is the sum of the two parts, as shown in Figure 6.

The amount of Ca(OH)₂ increased over time due to cement hydration [15]. However, when the composite cementitious materials contained TP, the amount of Ca(OH)₂ decreased over curing time. This was because of the pozzolanic reaction of TP and its consumption of Ca(OH)₂. Therefore, for the samples containing TP, when the consumption amount of Ca(OH)₂ was higher than the newly generated Ca(OH)₂ amount, the total content of Ca(OH)₂ in the system decreased.

Similar to the effect of curing time, the amount of Ca(OH)₂ in pure cement cured at 40 °C was higher than that at 20 °C, which indicates that high temperatures promote the hydration of cement. For the samples containing TP, the amount of Ca(OH)₂ in samples cured at 40 °C was lower than that at 20 °C. These results also indicate that the pozzolanic reaction of TP that consumes Ca(OH)₂ can be promoted by high temperatures. This was because a relatively high temperature increased the early strength of the composite cementitious materials [23].

As the amount of TP increased, the cement content in the system decreased. Figure 6 shows that the amount of Ca(OH)₂ was significantly reduced because of the decrease in Ca(OH)₂ produced by cement hydration and the consumption of Ca(OH)₂ via pozzolanic reaction, which is consistent with XRD analysis results.

4. Conclusions

This paper aimed to determine the role of TP in cement-based materials and its influencing factors. The following conclusions were made:

(1)

With the increase in TP content, the compressive strength of the specimens gradually decreased, which was more obvious in the case of a high W/B ratio.

(2)

A high curing temperature was beneficial for developing the compressive strength of cement pastes incorporating TP because it promoted cement hydration and the pozzolanic reaction of TP simultaneously. High curing temperatures can partially compensate for the strength decline disadvantage caused by increased TP content. The high curing temperature was most conducive to the development of strength at the early stage; however, its impact was weak at a later stage.

(3)

XRD and TG–DTA tests showed that the high curing temperature promoted the hydration of cement in the system, with more Ca(OH)₂ being generated. At the same time, some Ca(OH)₂ was consumed during the pozzolanic reaction of the TP. When the pozzolanic reaction of the TP was stronger than the cement’s hydration, the total Ca(OH)₂ content decreased. The Ca(OH)₂ content was reduced with the increase in TP content and curing temperature for the pastes containing TP.

(4)

The activation of the pozzolanic activity of TP can be prompted by high curing temperatures, which provides a theoretical foundation for applying TP to concrete.