Effect of Sodium Stearate on the Microstructure and Hydration Process of Alkali-Activated Material

Abstract

In response to the high water sorptivity of alkali-activated slag–red mud–steel slag cementitious material (AASRSS), water and aggressive ions are easily transferred to the interior of cementitious materials, which poses a significant threat to their durability. In order to limit the transfer of moisture and aggressive ions to the interior of AASRSS, sodium stearate (NaSt) was utilized in this paper to reduce its water sorptivity. The aim of this study was to elucidate the role of NaSt in the AASRSS system. The changes in the compressive strength and water sorptivity of AASRSS mortar were tested by mixing different amounts of NaSt into the mortar. The changes in the hydration process and microstructure of AASRSS after mixing with NaSt were analyzed by using the test methods of the isothermal calorimetry, resistivity, SEM and MIP. The results showed that NaSt plays an important role in the pore structure characteristics of AASRSS and that the use of NaSt significantly reduces its water sorptivity. The improvement in the water sorptivity was caused by two main mechanisms, namely the optimization of the pore structure (reduction in defects of the microstructure and alteration of pore size distribution) and the introduction of a hydrophobic film on the pore surface. The addition of NaSt did not change the type of AASRSS hydration product, but it inhibited the hydration reaction, leading to a reduction in the hydration product. The combination of the increased porosity and reduced hydration products mainly accounts for the decrease in the compressive strength of AASRSS due to NaSt.

1. Introduction

Portland cement (PC) has become a widely used bonding material because of its excellent strength and durability, as well as its economic advantages. However, with the growing greenhouse effect, people are looking for alternative adhesive materials to PC due to the large amount of CO₂ emitted during the production of PC [1]. Alkali-activated material (AAM) is an adhesive that has been developed to replace PC, and the aim was to develop an environmentally friendly adhesive [2]. Some studies have shown that the use of AAM can reduce CO₂ emissions by 40–80% compared to PC [3].

AAM is manufactured by activating a solid silica–aluminate precursor material with an alkali activator [4]. Silica–aluminate-rich materials generally include metakaolin, silica fume, fly ash, steel slag, red mud and blast furnace slag [5,6]. Considering the current high emissions and low utilization of red mud and steel slag, their large stockpiles bring a series of environmental problems [7,8], and it is necessary to develop eco-friendly technologies regarding waste (e.g., steel slag and red mud) resources [9]. In this paper, we use blast furnace slag, red mud, steel slag and an alkali activator to produce a clinker-free binder called alkali-activated slag–red mud–steel slag cementitious material (AASRSS). It is a promising binder because the development of AASRSS can not only relieve the economic and environmental pressure caused by steel slag and red mud [10], but it can also replace part of the cement and reduce the production of cement, and, thus, reduce CO₂ emissions [11], which is in line with the national strategy of “sustainable development” in China.

As a potential substitute for cement, AASRSS has good mechanical properties and an excellent resistance to chemical attack [12]. Despite its favorable performance, the widely recognized applications of AASRSS in the industry have been hindered, partly due to concerns related to its long-term durability, such as extensive shrinkage and micro-cracking, rapid carbonation, and potential alkali–aggregate reactions. One of the most significant conundrums impeding the implementation of AASRSS is its higher water sorptivity, caused by the high content of alkali-containing salt in red mud and the poor volumetric stability [13]. This means that AASRSS is more susceptible to durability issues due to the fact that water transport in porous matrices is a source of multiple factors for material deterioration or loss of function [14,15]; for example, a high water sorptivity can lead to the easy penetration of water carrying corrosive ions into the interior of the material, which can destroy the microstructure of the material, leading to a reduction in the compressive strength of the material. Therefore, reducing water absorption is an important issue to be addressed before the promotion and application of AASRSS.

Hydrophobic treatment is one of the measures commonly employed for cementitious materials [16]. This faces a practical difficulty, however, in that some water repellents developed for OPC lose their functionality in alkali-activated systems, such as sodium methylsilicate, a common OPC water repellant; the highly alkaline environment of an alkali-activated system causes the water-repellent film formed by sodium methylsilicate to hydrolyze and lose its water repellency [17]. Recently, stearates have gained wide attention as waterproofing materials due to their low cost and effectiveness [18,19,20]. Some studies have shown that the incorporation of stearates can effectively reduce the water absorption rate of alkali slag cement stones [21]; however, there are very limited studies related to the application of stearates in alkali-excited multivariate composite systems. In addition, to the authors’ knowledge, a study on the application of sodium stearate (C₁₇H₃₅COONa, NaSt) to the AASRSS system has not been reported. Therefore, in order to elucidate the role of sodium stearate in the AASRSS system, in this paper, alkali-activated materials were prepared using blast furnace slag, steel slag and red mud. NaSt was used as a water-repellent reagent. The influence law of NaSt on the water sorptivity, mechanical properties, hydration process and microstructure of alkali-activated slag–red mud–steel slag materials was analyzed using the results of compressive strength, isothermal calorimetry, electrical resistivity, SEM and MIP.

2. Materials and Methods

2.1. Raw Materials

The ground-granulated blast furnace slag was from Guigang Iron and Steel Technology Group, Guiyang, China, and was ground to a specific surface area of 400 m²/kg with a density of 2.91 g/cm³ using a ball mill. The Bayer red mud was from Guangxi Baise Pingguo Aluminum Company, China, and was ground to a specific surface area of 670 m²/kg with a density of 3.11 g/cm³ using a ball mill. The converter steel slag, ground to 600 mesh, was from Guangxi Liuzhou Iron and Steel Company, China, with a specific surface area of 499 m²/kg and a density of 3.26 g/cm³. The main chemical elemental compositions of red mud, blast furnace slag and steel slag were determined by X-ray fluorescence spectroscopy (XRF), and the results are shown in

Table 1. Chemical compositions of slag, red mud and steel slag (wt.%).

Binder	SiO2	Al2O3	Fe2O3	MgO	CaO	Na2O	K2O	SO3	TiO2	MnO
Slag	30.79	13.56	0.96	8.18	41.59	0.40	0.46	2.12	0.94	\
Red Mud	13.49	20.39	30.58	0.32	15.82	9.15	0.19	0.66	7.80	\
Steel Slag	26.95	5.66	24.15	5.33	29.67	0.09	0.08	\	0.57	5.39

. The main phase compositions were determined by XRD, and the results are shown in Figure 1. In this paper, a mixture of sodium silicate and NaOH was chosen as the activator. The ratio of total SiO₂ to total Na₂O of the activator mixture was 1.5. The modified component was Aladdin’s USP-grade sodium stearate (density: 1.103 g/cm³).

2.2. Experimental Program

2.2.1. Specimen Preparation and Test Procedure

The alkali-activated slag–red mud–steel slag mortar specimens used a sodium silicate modulus of 1.5 and had a ratio for the mass of slag, red mud and steel slag of 4:4:2; the alkali equivalent (in Na₂O) was 5% of the mass of the cementitious material and the water–cement ratio (W/B) was 0.4. The NaSt dose was 0.5%, which was 1.5% of the mass of the cementitious material. When forming, the slag, red mud, steel slag and standard sand were first mixed in a mortar mixing pot at a slow speed for 2 min, and then the pre-mixed sodium silicate and water were poured into the pot and mixed at a slow speed for 60 s. The mixing was stopped for 30 s in the middle, the slurry at the edge of the mixing pot was scraped down, and finally, the mixture was loaded into a 160 mm × 160 mm × 40 mm mortar mold after mixing fast for 60 s. After being placed on the vibration table and vibrated 120 times, the surface was covered with a polyethylene film, the material was put into a standard maintenance room for 24 h, and then it was demolded. After demolding, it was maintained under the standard conditions of a temperature of 20 ± 2 °C and a humidity of >95% until the specified age. The mortar specimen compound is shown in

Table 2. AASRSS mortar mixture proportions (per group).

Group	NaOH/g	Sodium Silicate/g	Sand/g	Water/g	Slag/g	Red Mud/g	Steel Slag/g	NaSt/g
AASRSS/Control	15.55	118.25	1350	114.38	180	180	90	--
AASRSS/0.5%	15.55	118.25	1350	114.38	180	180	90	2.25
AASRSS/1.5%	15.55	118.25	1350	114.38	180	180	90	6.75

.

2.2.2. Test Procedures

(1)

Compressive strength: The test was performed with reference to the Chinese national standard GB/T17671-1999 [22].

(2)

The capillary water absorption coefficient test was carried out according to ASTM C1585 [23]. The specimens were first dried in an oven at 40 °C for 7 days, then placed in a drying dish at 20 °C for 48 h to achieve a uniform distribution of relative humidity inside the specimens and, finally, sealed around with aluminum foil tape (exposing the bottom surface). The 40 mm × 40 mm surface was immersed in water to a height of 3–5 mm and the mass of the specimen was weighed at 1, 3, 5, 8, 12, 16, 20, 25, 30, 35, 40, 45, 50, 55 and 60 min (after which, the mass of the specimen was basically stable). Then, the water absorption rate coefficient was calculated according to the following formula:

$$i=a+S{t}^{0.5}$$

where i is the volume of water absorbed per unit area (mm³/mm²); S is a material constant called the sorptivity (mm/min^0.5); a is a constant (mm); and t is the time elapsed (min).

(3)

Isothermal calorimetry: A hydration heat test was performed using an I-CAL4000/8000 isothermal calorimeter from Calmetrix, MA, USA. The test should be conducted in a 25 °C environment, and the instrument should be run for 24 h before the experiment to ensure the stability of the instrument. The water and alkali activator needed for the experiment also needed to be put into the apparatus in advance so that their temperature was kept at 25 °C. For the test, the total water–cement ratio was maintained at 0.4 and the effect of the incorporation of sodium stearate on the heat of the hydration of the alkali-activated slag–red mud–steel slag was determined. After weighing the cementitious material proportionally, the material was quickly mixed, homogenized and immediately placed in the test channel, and the test time was updated to start the test. The instrument recorded data every minute for 72 h.

(4)

Resistivity test: The CCR-3 induction resistivity of China Zhong heng Gang ke was used for the test. The mold was installed in advance, the stirred paste was poured into the mold, the paste in the mold was gently shaken up and down several times by hand on the standoffs and, finally, the mold was sealed well. The temperature was kept at around 20 °C during the test.

(5)

FTIR analyses: To determine the chemical groups in the reaction products, an FTIR analysis was performed using the KBr particle method. The 28-day paste samples were dried in an oven (temperature: 40 °C) for 3 days, and then the samples were crushed and ground into a powder. The tests were performed using a Thermo Nexus 470 Fourier-transform infrared spectrometer (FTIR) from Nicolet, USA. The tests were acquired using a resolution of 8 cm⁻¹ and a 64-scan program to obtain the IR spectral data in the wavelength range of 400 to 4000 cm⁻¹.

(6)

X-ray diffraction (XRD): Prior to testing, 28-day-old paste samples were dried in a drying oven at 40 °C and ground to a powder with a mortar that was able to pass through a 200-mesh square-hole sieve; then, the sample under the sieve was collected for X-ray diffraction testing and analysis. The crystalline phases in the samples were identified using a PANantical X’Per PRO diffractometer with a 40 kV voltage, a 40 mA current and a scanning speed of 2°/min. The scanning range was 5° to 80° (2θ).

(7)

MIP test: At 28 days, the paste samples were cut into 1 cm cubic specimens and then soaked in anhydrous ethanol for 24 h. Finally, the specimens were dried under a vacuum at 40 °C for 24 h. The pore structure of the samples was tested using a Mike AutoPoreV 9600 mercury piezometer (MIP) under the following conditions: pressure range: 0.2–60,000 psi and pore size range: 800 μm–3 nm. The minimum and maximum pressure applied corresponded to a cylindrical pore size of 800 μm and 3 nm, respectively. The test temperature was 19 °C; the adv. contact angle was 130.0°; and the rec. contact angle was 130.0°.

(8)

SEM test: The 28-day-old samples were taken out, hydration was terminated using anhydrous ethanol and the samples were dried under a vacuum at 40 °C to a constant weight. Then, each sample was glued to the test base with a conductive adhesive for surface gold spraying. Finally, the surface morphology of the samples was observed using a GeminiSEM 300 electron scanning electron microscope (SEM) from Zeiss, Germany, and the elements in the reaction products were determined using energy spectrometry (EDS). The SEM test conditions used a voltage of 5–15 kV.

3. Results and Discussion

3.1. Effect of NaSt on the Compressive Strength and Water Absorption Rate of AASRSS Mortars

3.1.1. Compressive Strength

Figure 2 shows the results of the compressive strength for each group of mortar specimens with different ages. As a whole, the compressive strength of all the groups of specimens increased gradually with age, and the compressive strength of the AASRSS control exceeded 60 MPa at the age of 28 days. After adding NaSt, the compressive strength of the AASRSS mortar specimens decreased slightly for a given age, and the more NaSt added, the greater the strength decreased. The addition of 0.5 wt.% and 1.5 wt.% NaSt resulted in approximately a 10% and 16% reduction in the 28-day compressive strength, respectively.

Jiang et al. [24] investigated the effect of NaSt on OPC. In their study, the addition of 0.5% NaSt (mass ratio) to OPC concrete with a W/B ratio of 0.5 resulted in a decrease in the compressive strength of about 50% at 28 days. This is related to the strong hydrophobic effect of NaSt, which inhibits the hydration of ordinary silicate cement [17,25]. According to the results obtained in this study, NaSt has a similar effect on the development of the compressive strength of AASRSS, but the effect on AASRSS is much lower than that on OPC. It seems that NaSt is more applicable to alkali-activated systems.

3.1.2. Water Sorptivity

Figure 3 shows the results of the water sorptivity for each AASRSS mortar group at 28 days. It can be seen that the AASRSS control group exhibited a value of water sorptivity as high as 6.09 × 10⁻¹ mm/min^0.5, which is about 10 times larger than that of an OPC sample with a similar W/B or the same compressive strength level. Encouragingly, the water sorptivity of AASRSS was reduced to 2.36 × 10⁻¹ mm/min^0.5 and 1.24 × 10⁻¹ mm/min^0.5, respectively, when 0.5 wt.% and 1.5 wt.% of NaSt was added. The water sorptivity of AASRSS was significantly reduced by up to approximately 80%, showing that NaSt does effectively reduce the water absorption rate of AASRSS.

To further explore the mechanisms behind these findings of compressive strength and water sorptivity, the reaction process, reaction products and pore structure were carefully examined, and the results are presented in the following sections.

3.2. The Influence of NaSt on the Hydration Process of AASRSS

3.2.1. Isothermal Calorimetry

The exothermic rate and cumulative heat release of AASRSS are shown in Figure 4. It was found that the development of the exothermic rate and the cumulative exothermic curve of AASRSS did not change significantly with the addition of NaSt, while the peak value of exothermic activity around 10 h decreased and the cumulative exothermic heat after 72 h decreased, indicating that NaSt inhibited the hydration reaction of AASRSS. The main reasons for the inhibition of AASRSS hydration by NaSt are as follows: (1) NaSt adsorbs on the surface of silica–aluminate precursor particles, which prevents the precursor and alkali activator from reacting in contact with each other [24]. (2) The wrapping effect of NaSt on the solution in the slurry reduces the liquid-phase distribution of the system to some extent, resulting in a reduction in the contact area between the precursor particles and the alkali activator [26].

3.2.2. Electrical Resistivity

It is well recognized that the electrical resistivity of AASRSS during the hydration process is related to the ion concentration in the solution and the pore structure of the hardened matrix. The lower the ionic concentration and the lower the pore transport in the matrix, the higher the resistivity of the AASRSS hydration system, and vice versa [27,28]. As can be seen from Figure 5, the resistivity profile of the AASRSS control group could be divided into four stages:

Stage 1 (0 to 2.5 h): the resistivity decreased rapidly because of the dissolution of precursor particles (slag, red mud and steel slag), resulting in a rapid increase in the concentration of Ca²⁺, SiO₄⁴⁻ and AlO₄⁴⁻ in the solution, and thus causing an improvement in the electrical conductivity of the alkali-activated system. Stage 2 (2.5~10 h): the resistivity remained stable and entered into an induction period with a relatively slow hydration rate as the hydration products adhered to the surface of the precursor particles to form an encapsulation layer, hindering further hydration. Stage 3 (10–20 h): the resistivity increased rapidly as the hydration products continued to increase, and the matrix became denser progressively; the reduction of porosity in the hydration system played a dominant role and the electrical conductivity decreased quickly. Stage 4 (20 to 72 h): the resistivity increased at a uniform rate and the hydration reaction entered a stage of steady development controlled by ion diffusion, as the layer of hydration products on the surface of the precursor particles thickened.

Interesting findings can be found in the development of electrical resistivity when NaSt was introduced into the AASRSS. In Stage 1, the resistivity of the AASRSS with NaSt became higher, indicating that the ion concentration at this stage decreased, which could be attributed to the hydration inhibition of precursor particles by NaSt, hindering the ionic dissociation. It should be noted that NaSt inhibits the hydration reaction of the precursors, which usually leads to a decrease in the resistivity due to the fact that the compactness of the microstructure is reduced. However, in Stages 3 and 4, the resistivity of the AASRSS increased significantly after NaSt was incorporated into it. An alternative explanation is that more air pores were formed in the specimens during the mixing and preparation process of the AASRSS due to the air-bubble-stabilizing function of stearate [29], and the connectivity between pores was blocked, making the migration of internal ions difficult and leading to a decrease in the electrical conductivity. More detailed information can be seen in the results of the hydration products and pore structure.

3.3. Effect of NaSt on the Hydration Products and Pore Structure of AASRSS

3.3.1. XRD

The XRD results of the 28-day AASRSS samples are shown in Figure 6. Calcite (CaCO₃, 2θ = 29.41° and 47.5°), alite (2θ = 22.24°, 27.55°, 29.36° and 32.21°) and hematite (Fe₂O₃, 2θ = 24.2°, 33.12°, 35.6° and 49.52°) could be detected. An important indication in Figure 6 is the peak in the X-ray diffraction pattern at 30° and 49.5°, which is attributed to C(N)-A-S-H. The difference in the intensity of this peak is somewhat indicative of the amount of C(N)-A-S-H gel produced; specifically, the intensity of C(N)-A-S-H slightly decreased with the addition of NaSt at around 30° and 49.5°. This indicates that the addition of NaSt slightly reduced the formation of C(N)-A-S-H in the AASRSS, which is consistent with the heat-of-hydration results.

3.3.2. FTIR

Due to the poor crystallinity of most of the products of the alkali slag–red mud–steel slag, which is difficult to detect using XRD, FTIR tests were carried out. By distinguishing between the typical wave numbers and the corresponding transmittance, the relative amounts of chemical reaction products can be estimated. The results of the AASRSS FTIR at 28 days are shown in Figure 7. The absorption peaks of the [O-H] stretching vibration of the water molecule are shown at 3460 cm⁻¹ and 1631 cm⁻¹ [30]. Three bands are present at 1471 cm⁻¹, 1419 cm⁻¹ and 876 cm⁻¹, corresponding to the antisymmetric stretching vibrations and out-of-plane bending vibrations of the CO₃²⁻ ion [30]. The peaks at about 2918 cm⁻¹ and 2850 cm⁻¹ correspond to the asymmetric stretching vibrations and symmetric stretching vibrations of the CH₂ group of NaSt [31]. The bending vibrational absorption peak and the stretching vibrational absorption peak of [Si-O] were found at 987 cm⁻¹ and 466 cm⁻¹, respectively [32]. Combined with the results of the XRD pattern, we could attribute these absorption bands to the C-A-S-H gel in the hydration products. In summary, the intensity of the [Si-O] absorption peaks became slightly lower with the addition of NaSt, meaning that the amount of C(N)-A-S-H in the AASRSS was reduced, expectedly. Moreover, the wave number of [Si-O] shifted from 987 cm⁻¹ to the slightly higher value of 994 cm⁻¹ with the addition of 5 wt.% NaSt, which means that the structure of C(N)-A-S-H was probably changed. To explain this, further research is needed.

3.3.3. MIP

The results of the MIP tests are shown in Figure 8. As shown in Figure 8a, the AASRSS with 0.5 wt.% NaSt added showed a comparable level of total cumulative pore volume to the control group, and the incorporation of 1.5 wt.% NaSt resulted in a larger porosity of the AASRSS. It is thus plausible that the combination of the increased porosity and reduced hydration products mainly accounted for the decrease in the compressive strength of the AASRSS due to the NaSt.

In addition, from Figure 8b, it can be seen that the threshold pore size of AASRSS with the incorporation of 1.5 wt.% NaSt moves towards a higher value, and the distribution of pore size in the range of the mesopore scale was changed. The details of the representative pore size distribution percentages are shown in

Table 3. Pore distribution of AASRSS pastes with/without NaSt at 28 days.

Sample	Volume Fraction (%)
	<9.1 nm	9.1~21.1 nm	21.1~95.3 nm	>95.3 nm
AASRSS/Control	88.7	5.6	1.1	4.6
AASRSS/0.5%	82.3	6.1	1	10.6
AASRSS/1.5%	75.9	13.8	2.3	8

, from which it is clear that the pore size of AASRSS was coarsened by NaSt in the mesopore region. This is beneficial to a reduction in capillary suction, which is one main driving force for water sorptivity according to the Young–Laplace equation.

3.3.4. SEM

To further clarify the effect of NaSt on the pore structure of AASRSS, the microstructure was observed using SEM, and the results are shown in Figure 9. It can be clearly observed that, on one hand, with the incorporation of NaSt, the microcracking of AASRSS was reduced significantly; on the other hand, the matrix became denser and a large amount of micro-closed pores formed inside the specimen (darker part). EDS was used to determine the elements on the surface of the AASRSS pore wall. Since light elements such as C and O are sometimes difficult to detect due to their low energy, using EDS for light elements increases the acquisition time and reduces the acceleration voltage. To avoid the interference of C elements in air, we minimized the contact between the sample and air during the sample preparation. The EDS results show that the AASRSS pore wall contains a large amount of C elements. The trace carbonization that occurs during the sample preparation process should not produce so many C elements. Thus, the EDS results indicate that the pore wall is covered with a wax-like layer of a NaSt compound, which has hydrophobic groups [33]. All these together result in a decrease in the water sorptivity of AASRSS with the incorporation of NaSt.

4. Conclusions

The following conclusions were drawn regarding the pattern and mechanism of the effect of NaSt on the microstructure and hydration of AASRSS gelling materials:

• The incorporation of NaSt effectively reduced the water absorption rate of AASRSS, with a dose of 1.5 wt.% reducing the water absorption rate of 28-day mortar specimens by about 80%.
• The incorporation of NaSt did not change the type of hydration products of AASRSS, and the trend in the exothermic rate and cumulative exothermic curve of the hydration exothermic process did not change either, but it had some adverse effects on the intensity of the exothermic activity peak and cumulative hydration exothermic heat.
• The addition of NaSt inhibited the hydration reaction of AASRSS, leading to an increase in porosity. The combination of an increased porosity and reduced hydration products mainly accounts for the decrease in the compressive strength of AASRSS due to NaSt.
• Based on the experimental results, the improvement mechanism of NaSt can be summarized through two main reasons. The first is the optimization of the pore structure of AASRSS, which involves a reduction in the defects of the microstructure and an alteration of the pore size distribution. On the other hand, when NaSt is added, a hydrophobic membrane is formed on the surface of the pore wall of the AASRSS, which is effective at preventing water from entering the AASRSS.