Mechanical Properties and Hydration Characteristics of Weathered Residual Soil of Granite-Based Geopolymer

Highlights

What are the main findings?

• Calcined WRSG can be activated by a sodium silicate activator. An optimum formula for activation consists of a sodium silicate modulus of 0.9, an alkali content of 14%, and a water-to-soil ratio of 0.5. The formula can achieve 28-day compressive strength of 53.5 MPa.
• WRSG geopolymer exhibits a characteristic of high early strength. Its compressive strengths are 25.6 and 38.1 MPa at one and three days respectively, which account for 48% and 71% of the value at 28 days.
• Chemically bound water increases continuously with curing age, which shows a positive trend with compressive strength.
• Mineral phases of geopolymer mainly consist of amorphous gels and primary minerals from WRSG.

What are the implications of the main findings?

• Compressive performance of WRSG geopolymer enables WRSG as an alternative precursor material to metakaolin.
• Rapid development in strength shortens curing time required for WRSG geopolymer coatings.
• Content of chemically bound water can be regarded as an index of hydration degree.
• Primary minerals from WRSG can hardly be activated by sodium silicate, and thus should be eliminated in preprocessing.

Abstract

Geopolymer coatings exhibit outstanding corrosion resistance, high-temperature performance and thermal insulation. This thus holds broad application prospects in anti-corrosion of metals, protection of building structures, and functional coatings. However, the large-scale application of geopolymers is constrained by the availability of precursor materials. In South China, construction waste soil is predominantly composed of weathered residual soil of granite (WRSG), which is rich in silicate and aluminosilicate minerals. This soil can serve as a precursor for geopolymer synthesis upon activation. In this study, geopolymers were prepared using activated WRSG as the precursor material. The mix proportion of the geopolymers was optimized through single-factor experiments. Additionally, the hydration process and products of the geopolymer were characterized. The experimental results show that both high alkali content and low water-to-soil ratio contribute to achieving high compressive strength. The geopolymer has early strength characteristics. Its one-day compressive strength can reach 48% of 28-day value. The hydration products of the geopolymer mainly consist of amorphous sodium–aluminum–silicate–hydrate gel and primary minerals such as quartz and albite. With the increasing age, the content of chemically combined water and gel clusters grows, which densifies the microstructure and elevates the degree of hydration reaction of geopolymers.

1. Introduction

Geopolymer coatings are a new type of inorganic coatings with geopolymers as the film-forming material. They are three-dimensional network polymer coatings formed by the geopolymerization of precursor materials under the action of alkaline activators [1]. Geopolymer coatings have early strength and rapid hardening [2,3], resistance to chemical medium erosion [4,5,6], and high-temperature resistance [7,8,9]. They have great application potential in the fields of anti-corrosion and fireproof coatings for steel structures and functional coatings. The author has developed a fireproof coating with geopolymer and aerogel as the main components. The coating with thicknesses of 20 and 25 mm can improve thermal insulating efficiency of steel plates by 84 and 108 min and enable steel beams to achieve the second and first levels of fire resistance, respectively [10]. The main precursors of geopolymers are metakaolin and fly ash [11]. The former is expensive and mainly used in the paper and ceramic industries [12]. The latter has been widely used as an admixture in cement concrete. It is also increasingly scarce due to the strict restrictions on high-carbon emission industries such as coal-fired power generation [13]. Traditional geopolymers are facing a shortage of raw materials. Finding a cheap silicate mineral raw material has become the key to developing low-cost geopolymers.

In recent years, the rapid urbanization process has generated a large amount of construction waste soil, but the utilization rate of the waste soil is very limited in China [14]. In most areas, the waste soil is still disposed of by transporting it away or piling it up in an open area, which occupies a large amount of land and even causes serious accidents [15,16]. The construction waste soil in South China is mainly composed of WRSG, with a thickness of 20–35 m in Guangdong and Fujian Provinces [17]. WRSG is characterized by large volume and wide distribution. It is rich in silicate minerals such as quartz and kaolinite [18]. After being calcined at high temperatures, the kaolinite changes from a crystalline state to amorphous metakaolinite, which has good cementitious properties [19]. Within the range of 541 to 850 °C, as the activation temperature increases, a large amount of hydroxyl groups in kaolinite are removed, the structure becomes disordered, and the activity gradually increases [20]. The high hydroxyl content of kaolinite makes its pozzolanic activity after calcination stronger than that of illite and montmorillonite, which gives it potential for engineering utilization [21]. Moya et al. [22] pointed out that kaolinite calcined between 600 and 800 °C is very suitable for use as an ecological pozzolanic additive for ordinary Portland cement and as a precursor for geopolymers. Zhou et al. [23] ground WRSG finely and calcined it at 700 °C for 2 h, using it as a sustainable supplementary cementitious material. Experimental results shows that the metakaolinite generated during the calcination process promotes the secondary hydration of cement. Using this material to replace 10%–30% of the cement does not reduce the compressive strength of cement mortar. Compared with the landfill scheme, the WRSG recycling method proposed in this study can save 1204.1 MJ/t of energy consumption and 326.1 kg of carbon dioxide equivalent. Ferone et al. [24] used sodium silicate to activate the calcined clay sludge and made geopolymers, but the 28-day compressive strength was only 12.17 MPa. The strength of the specimens presented in [23,24] is unsatisfactory. The reason is that the quartz and other minerals mixed in the clay are difficult to be activated by high temperature, which weakens the activity of the clay and the secondary hydration reaction or geopolymerization reaction, resulting in poor mechanical properties of the products. Yuan et al. [25] studied the effect of particle size on the activity of calcined engineering waste soil in alkaline solution and found that as the particle size decreases, the quartz content in the waste soil decreases and the clay mineral content increases. After calcination, the reaction degree of large-particle-size (1.18–2 mm) waste soil with alkaline solution is limited, while the small-particle-size (<0.075 mm) waste soil obtained a high amorphous phase content of up to 90.61%. Therefore, screening the engineering waste soil to remove the inert coarse particles is a necessary prerequisite for efficient activation of the waste soil. Wu et al. [26] screened the engineering waste soil to increase the proportion of kaolinite, and then thermally activated the screened fine particles to replace cement as a cementitious material, achieving good mechanical performance. Li et al. [27] pre-treated the engineering waste soil with water washing to increase the kaolinite content to 72.4% to 83.1%, and then calcined it at 800 °C for 2 h to prepare limestone calcined clay cement, whose compressive strength is comparable to that of ordinary Portland cement. Xiao [28] and Chen [29] respectively screened fine particles from the granite weathering residual soil for calcination activation and prepared geopolymers and engineering geopolymer composites. The former’s 28-day compressive strength exceeded 40 MPa, and its carbon emissions and preparation cost were reduced by 36% and 20% respectively compared with Portland cement. The latter’s 28-day tensile strength and tensile strain reached 4.17 MPa and 4.71% respectively, and its carbon emissions and material cost were reduced by 6% and 19% respectively compared with engineering cementitious composites.

It can be seen from the above literature that WRSG is rich in silicate minerals. After screening and calcination, it contains a large amount of amorphous metakaolinite and can replace metakaolin and fly ash to become the precursor material of geopolymer. However, the current related research mainly focuses on the mechanical performance, economic benefits and carbon emissions of WRSG geopolymer. There is little research on the hydration characteristics of activated WRSG in alkaline solution. This makes it difficult to explain the mechanism supporting the excellent engineering performance of WRSG geopolymer and is not conducive to the engineering promotion of this material. Therefore, this study explored the effects of sodium silicate modulus, alkali content, water-to-soil ratio and curing age on the compressive strength of WRSG geopolymer, and optimized the mix proportion with the highest compressive strength. The hydration process and hydration products of WRSG geopolymer were investigated by the method of chemically bound water, degree of reaction, thermal analysis, X-ray diffraction and scanning electron microscopy. The relationship between chemically bound water and compressive strength of the WRSG geopolymer was established. The results will expand the raw material range of the geopolymer, reduce the material cost of the geopolymer, and promote the resource utilization of WRSG.

2. Materials and Experimental Methods

In this study, compressive tests were first performed to investigate the effect of formulation composition on the compressive strength of WRSG geopolymer at different ages. The formulation yielding the highest compressive strength was subsequently optimized. A chemically combined water test and hydration reaction degree test were then employed to examine the hydration process of the optimized geopolymer. A thermal analysis test, X-ray diffraction test and scanning electron microscopy test were adopted to verify the hydration products and observe the microstructure of the geopolymer, respectively.

2.1. Materials

The WRSG used in this experiment was taken from the foundation pit of a construction site in Panyu District of Guangzhou. Its main minerals are quartz, feldspar and clay minerals (

Table 1. Mineral composition of WRSG (wt%).

Calcite	Quartz	Plagioclase	Albite	Orthoclase	Rutile	Illite	Chlorite	Glauconite	Muscovite	Clay Minerals
0.59	13.29	27.81	3.98	2.63	0.26	30.19	17.78	0.56	2.91	55.14

), and the chemical composition is mainly silicate and aluminate (

Table 2. Chemical composition of WRSG (wt%).

SiO2	Al2O3	Fe2O3	CaO	MgO	K2O	Na2O	SO3	TiO2	MnO2	Ignition Loss
53.44	27.10	3.59	0.08	0.17	5.70	0.18	0.04	0.34	0.09	9.24

). The particle size mainly ranges from 3 to 400 μm (Figure 1). The particles larger than 150 μm were sieved out by a vibrating screen, and the remaining fine particles were calcined at 800 °C for two hours and then cooled to room temperature to be used as the precursor of the WRSG geopolymer.

The WRSG geopolymer adopted sodium silicate water glass solution as the alkaline activator. The water glass solution had a modulus (molar ratio of SiO₂ to Na₂O) of 3.26 and a concentration of 34.5%. Sodium hydroxide and water were used to adjust the water glass solution to achieve required alkali content (mass ratio of Na₂O to WRSG) and modulus.

During the test, the water glass solution is poured into the activated WRSG, stirred slowly for 120 s first, then stirred quickly for 120 s. Subsequently, the well-mixed slurry is poured into the mold and vibrated to be compact, and then cured in an environment of 20 °C and 95% relative humidity until the required age.

2.2. Experimental Design

In this study, modulus, alkali content, and water-to-soil ratio (mass ratio of water to WRSG) were selected as the study variables (

Table 3. Experimental variables and consumption of hydrochloric acid in reaction degree test.

No.	Modulus	Alkali Content	Water-to-Soil Ratio	Consumption of Hydrochloric Acid (mL)
1	0.3	14%	0.5	48.03
2	0.6			48.05
3	0.9			48.13
4	1.2			48.07
5	0.9	8%	0.5	45.59
6		10%		46.31
7		12%		47.39
8		14%		48.13
9		16%		- 1
10	0.9	14%	0.45	- 2
11			0.50	48.13
12			0.55	47.63
13			0.60	47.57
14			0.65	- 3

^1,2,3 The formulas failed to prepare geopolymer specimens.

). The levels of the variables listed in Table 3 were decided based on preliminary experimental results. To examine the effect of a single variable on the compressive performance of WRSG geopolymers, the other variables were kept constant during each test. WRSG geopolymer paste specimens were prepared according to the mix proportions given in Table 3. The specimens were subjected to compressive tests at 3, 7 and 28 days. This single-factor experiment aims to examine the influence of the variables, which contributes to optimizing the geopolymer with the highest compressive strength.

2.3. Experimental Methods

2.3.1. Compressive Test

The dimensions of WRSG geopolymer specimens were 20 mm × 20 mm × 20 mm. After curing to the specified age, the specimens were taken out and wiped clean. The side formed during specimen preparation was designated as the bearing surface. The specimens were then centered on the lower platen of the testing machine. The upper platen of the testing machine was equipped with a spherical hinge. The compressive loading was applied at a stress rate of 1.5 MPa/s until the failure of specimens. Three replicates were tested for each formula. The mean compressive strength of the three specimens was adopted as the compressive strength value for the formula. If the difference between either the maximum or minimum value and the median value exceeded 15% of the median, both the maximum and minimum values would be excluded, and the median would be retained as the compressive strength value for the formula. If the differences between both the maximum and minimum values and the median exceeded 15% of the median, the test results of the formula would be deemed invalid.

2.3.2. Chemically Combined Water Test

Chemically bound water is one of the reaction products formed between WRSG and alkali activators. Its content increases with the accumulation of hydration products, which characterizes the hydration process of WRSG geopolymer [30].

In the experiment, an alumina crucible was first heated to 950 °C and held at this temperature for 30 min in a muffle furnace. The crucible was then cooled and weighed. This procedure was repeated until the mass difference between two consecutive instances of weighing did not exceed 0.05 g, ensuring the crucible reached a constant weight. Subsequently, the WRSG geopolymer paste was ground into fine powder, which was dried to a constant weight in a drying oven at 50 °C. Then, 1.00 g of the dried powder was accurately weighed and transferred into the pre-weighed crucible. Finally, the crucible containing the powder was placed back into the muffle furnace, heated to 950 °C, and held for 30 min. After cooling, the crucible with the residual sample was weighed again. The chemically bound water content of the sample was calculated using Equation (1):

$$w_{\mathrm{c}\mathrm{b}}={\displaystyle \frac{m_1\times \left(1-w_{\mathrm{L}\mathrm{O}\mathrm{I}}\right)-m_2}{m_1\times \left(1-w_{\mathrm{L}\mathrm{O}\mathrm{I}}\right)}}\times 100\%$$

(1)

w_cb represents the content of chemically combined water in the paste. m₁ is the mass of dried paste powder before calcination, which consists of unreacted solid precursor, unreacted solid activator, and reaction products. m₂ is the mass of paste powder after calcination. w_LOI is the ignition loss of calcined WRSG, which can be measured according to the burning difference method mentioned in GB/T 176-2017 [31].

2.3.3. Hydration Reaction Degree Test

Metakaolinite exhibits excellent stability in hydrochloric acid with a pH range of 0–2, whereas other minerals are susceptible to dissolution in hydrochloric acid. Thus, the residue remaining after the acid dissolution reaction is likely unhydrated WRSG. A higher residue content corresponds to a lower degree of geopolymerization, resulting in reduced formation of sodium–aluminum–silicate–hydrate gel. This, in turn, leads to increased porosity and decreased compressive strength [32]. This acid dissolution method is suitable for testing the hydration reaction degree of a geopolymer because it can separate unreacted kaolinite from the reaction product sodium–aluminum–silicate–hydrate gel. Calcite in the WRSG can also dissolve in hydrochloric acid. However, the content of calcite in calcined WRSG is lower than 0.1%. Its interference with the experimental results can be ignored.

After seven days of curing, the WRSG geopolymer paste was ground into fine powder, immersed in anhydrous ethanol for 24 h, and subsequently dried to a constant weight in an oven at 55 °C. Then, 10 g of dried geopolymer powder was added to 500 mL of hydrochloric acid solution with a pH value of 0 and stirred for one hour. The hydrochloric acid solution was prepared by diluting 83.3 mL of hydrochloric acid with a concentration of 37 wt% to a total volume of 1000 mL using distilled water. During the stirring, the pH of the solution was monitored using a pH meter. Hydrochloric acid solution of 37 wt% was added every 15 min to keep the pH value of the solution at 0. The total dosage of hydrochloric acid of 37 wt% used for each formula is given in Table 3. The temperature of the experimental environment was maintained at 20 °C. Finally, the solution was filtered to collect the insoluble residue. The residue was rinsed with distilled water and dried to a constant weight in an oven at 55 °C. The degree of reaction was the mass ratio of the difference between paste powder and dried residue to the paste powder.

2.3.4. Thermal Analysis Test

WRSG geopolymer pastes were prepared using the optimized mix proportion. After seven days of curing, the pastes were immersed in anhydrous ethanol to terminate hydration, followed by drying in a vacuum dryer and grinding into powder with particle size less than 0.075 mm. Then 0.3 g of geopolymer powder was heated at a rate of 10 °C/min to 900 °C in a nitrogen atmosphere using a simultaneous thermal analyzer (NETZSCH, Selb, Germany) (Figure 2), enabling the characterization of thermogravimetric and heat flow changes.

2.3.5. X-Ray Diffraction Test

Optimized WRSG geopolymer pastes were cured for 1, 3, 7 and 28 days, respectively. After curing, the pastes were immersed in anhydrous ethanol to terminate hydration, followed by drying in a vacuum dryer and grinding into powder with particle size less than 0.075 mm. The phase composition of the geopolymer powder was analyzed using a Rigaku SmartLab X-ray diffractometer (Rigaku, Tokyo, Japan) (Figure 3). The scanning range was from 5° to 80°. The scanning step was 0.01°, and the scanning rate was 10°/min.

2.3.6. Scanning Electron Microscopy Test

Optimized WRSG geopolymer pastes were cured for 1, 3, 7 and 28 days, respectively. After curing, the pastes were broken into fragments with fracture surface of 0.5 cm². The fragments were then immersed in anhydrous ethanol to terminate hydration and dried in a vacuum dryer for five hours. The microstructural morphology of the fracture surface was observed using a Zeiss Crossbeam-350 scanning electron microscope (Zeiss, Oberkochen, Germany) (Figure 4).

3. Results and Discussion

3.1. Effect of Moduli of Water Glass Solution

When the water-to-soil ratio was 0.5 and the alkali content was 14%, the compressive strength and reaction degree of WRSG geopolymers with various water glass moduli are given in Figure 5. The variable coefficients of the compressive strength ranged from 0.012 to 0.052. At a modulus of 0.3, the 7-day compressive strength of the geopolymer was only 1.8 MPa, with a reaction degree of 53.3%. As the modulus increased to 0.6, the compressive strength increased sharply. With a further increase in the modulus to 0.9, the 7-day compressive strength of the geopolymer reached a peak value of 42.7 MPa, and the reaction degree also peaked at 65.2%. When the modulus exceeded this value, the setting rate of the paste accelerated, and efflorescence appeared on the surface of the specimens. This phenomenon of efflorescence is also observed in the literature [29]. The geopolymer was prepared using calcined WRSG and water glass with alkali content of 15% and modulus of 1.2. White carbonates were found on the surface of specimens after curing.

High-modulus water glass solutions can provide sufficient SiO₃²⁻ for the geopolymerization reaction, while accelerating the reaction rate and shortening the setting time [33]. However, excess sodium silicate reacts with carbon dioxide in the air to form sodium carbonate on the specimen surface, which adversely affects the appearance. The 3-day and 28-day compressive strengths of WRSG geopolymers followed the same trend as the 7-day value. Therefore, the optimal water glass modulus is determined to be 0.9.

Under constant water-to-soil ratio of 0.5 and alkali content of 14%, the relationship between compressive strength (f_c) and water glass modulus (0.3 ≤ m_wg ≤ 1.2) was established by fitting experimental data as Equation (2).

$$f_{\mathrm{c}}=\left\{\begin{array}{c}-70.556{m_{\mathrm{w}\mathrm{g}}^2+141.83m}_{\mathrm{w}\mathrm{g}}-34.8 \left(\mathrm{a}\mathrm{t} 3 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=1\right) \\ -93.611{m_{\mathrm{w}\mathrm{g}}^2+181.32m}_{\mathrm{w}\mathrm{g}}-43.925 \left(\mathrm{a}\mathrm{t} 7 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=0.9988\right)\\ -94.722{m_{\mathrm{w}\mathrm{g}}^2+201.05m}_{\mathrm{w}\mathrm{g}}-50.625 \left(\mathrm{a}\mathrm{t} 28 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=0.9886\right)\end{array}\right.$$

(2)

3.2. Effect of Alkali Contents

When the water-to-soil ratio was 0.5 and the modulus was 0.9, the compressive strength and reaction degree of WRSG geopolymers with various alkali contents were as illustrated in Figure 6. The variable coefficients of the compressive strength ranged from 0.014 to 0.106. At an alkali content of 8%, the 7-day compressive strength of the geopolymer was merely 1.9 MPa. As the alkali content increased, both the compressive strength and reaction degree of the geopolymer increased monotonically, reaching their maximum values at an alkali content of 14%. When the alkali content was further elevated to 16%, the geopolymer paste underwent rapid setting, rendering it difficult to cast.

This phenomenon can be attributed to the fact that OH⁻ ions facilitate the depolymerization of Si-O and Al-O bonds in metakaolinite, thereby accelerating the geopolymerization reaction [34]. A higher alkali content corresponds to a faster reaction rate, which in turn enhances the strength of the geopolymer. However, excessively high alkali content may also induce flash setting of the geopolymer, which is detrimental to construction operations. Therefore, the optimal alkali content is determined to be 14%.

Under constant water-to-soil ratio of 0.5 and modulus of 0.9, the relationship between compressive strength (f_c) and alkali content (8% ≤ c_a ≤ 14%) was established by fitting experimental data as Equation (3).

$$f_{\mathrm{c}}=\left\{\begin{array}{c}-11125{c_{\mathrm{a}}^2+3020.5c}_{\mathrm{a}}-168.03 \left(\mathrm{a}\mathrm{t} 3 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=0.973\right) \\ -13313{c_{\mathrm{a}}^2+3618.3c}_{\mathrm{a}}-201.45 \left(\mathrm{a}\mathrm{t} 7 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=0.9845\right)\\ -14563{c_{\mathrm{a}}^2+4046.3c}_{\mathrm{a}}-226.88 \left(\mathrm{a}\mathrm{t} 28 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=0.984\right)\end{array}\right.$$

(3)

3.3. Effect of Water-to-Soil Ratios

When the modulus was 0.9 and the alkali content was 14%, the compressive strength and reaction degree of WRSG geopolymers with various water-to-soil ratios were as presented in Figure 7. The variable coefficients of the compressive strength ranged from 0.015 to 0.052. At a water-to-soil ratio of 0.45, the prepared paste was excessively dry and could not be cast for molding. As the water-to-soil ratio increased, the fluidity of the geopolymer improved. However, the excess water in the paste diluted the alkali activator, leading to a reduction in the reaction degree. During the subsequent hardening process, the evaporation of water left abundant pores in the geopolymer, which decreased its compressive strength. When the water-to-soil ratio reached 0.65, the geopolymer remained unhardened even after 24 h of curing, thus lacking practical application value. Therefore, the optimal water-to-soil ratio was determined to be 0.5. The optimum mix proportion of WRSG geopolymer was a water glass modulus of 0.9, a water-to-soil ratio of 0.5, and an alkali content of 14%.

Under constant alkali content of 14% and modulus of 0.9, the relationship between compressive strength (f_c) and water-to-soil ratio (0.5 ≤ w/s ≤ 0.6) was established by fitting experimental data as Equation (4).

$$f_{\mathrm{c}}=\left\{\begin{array}{c}-1240{\left({\displaystyle \frac{w}{s}}\right)}^2+1154\left({\displaystyle \frac{w}{s}}\right)-225.3 \left(\mathrm{a}\mathrm{t} 3 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=1\right) \\ -1000{\left({\displaystyle \frac{w}{s}}\right)}^2+1232\left({\displaystyle \frac{w}{s}}\right)-413.3 \left(\mathrm{a}\mathrm{t} 7 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=1\right)\\ -280{\left({\displaystyle \frac{w}{s}}\right)}^2+114\left({\displaystyle \frac{w}{s}}\right)-67.1 \left(\mathrm{a}\mathrm{t} 28 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s},R^2=1\right)\end{array}\right.$$

(4)

3.4. Reaction Process of WRSG Geopolymer

WRSG geopolymer specimens were prepared using the optimized mix proportion and cured for 1, 3, 7, 14 and 28 days. After curing, the specimens were subjected to a compressive test and chemically bound water test. Their chemically combined water and compressive strength at various ages are shown in Figure 8.

The compressive strength and chemically bound water content of optimized WRSG geopolymers increase continuously with curing age. At 1 day of curing, the chemically bound water content and compressive strength were 7.27% and 25.6 MPa, respectively. When the curing age increased to 7 days, the chemically bound water content and compressive strength increased by 14.9% and 80.5%, respectively. As the curing age further extended from 7 days to 28 days, the increments in chemically bound water content and compressive strength slowed to 13.5% and 15.6%, respectively. The decelerated growth rates of chemically bound water and compressive strength indicate that the geopolymerization reaction is approaching completion. The relationship between compressive strength (f_c) and chemically bound water (w_cb) was established by fitting experimental data as Equation (5) with correlation coefficients R² of 0.9974. The fitted result shows the compressive strength and chemically bound water content are positively related, yet different from the linear one of cement mortar proposed by Gu et al. [35]. This is because the rapid increase in the early strength of geopolymers is attributed to the formation of a three-dimensional Si-O-Al framework rather than the fixation of bound water. As hydration proceeds, water molecules enter the micropores of the sodium–aluminum–silicate–hydrate gel and become chemically bound water, making the microstructure more compact. However, the main structure of the geopolymer has been built, causing a slower increase in strength than that in the early stage.

$$f_{\mathrm{c}}=-65505w_{\mathrm{c}\mathrm{b}}^2+12223w_{\mathrm{c}\mathrm{b}}-516.59 \left(7.27\%\le w_{\mathrm{c}\mathrm{b}}\le 9.48\%\right)$$

(5)

3.5. Hydration Products of WRSG Geopolymer

The results of X-ray diffraction on optimized WRSG geopolymer at various ages are shown in Figure 9. The image exhibits prominent diffuse diffraction peaks in the 2θ range of 20–30°, which correspond to amorphous gel products generated by the hydration reaction of WRSG. This observation indicates that the silica–alumina phases in WRSG react with the alkali activator, forming a new amorphous polymeric structure [36]. A portion of amorphous gel transforms into zeolite phases. Furthermore, crystalline minerals including albite, quartz, and illite persist in the geopolymer across various curing ages, which indicates that the primary minerals in WRSG cannot react with alkaline solutions.

The results of thermal analysis on optimized WRSG geopolymer at 7 days are shown in Figure 10. The total weight loss of the specimen was 13.8%. A distinct endothermic peak was observed at approximately 116 °C, which was attributed to the extensive evaporation of free water in the geopolymer paste. By 200 °C, the weight loss rate of the specimen reached around 10%, indicating that the water in the alkali activator had not fully participated in the geopolymerization reaction. As the temperature increased from 200 to 600 °C, the specimen continued to lose weight but at a decelerated rate. This phenomenon was caused by the dehydroxylation of the geopolymer paste, leading to a reduction in structural water. When the temperature reached 700 °C, the weight loss process was essentially complete, and no new phases were detected.

3.6. Microstructural Morphology of WRSG Geopolymer

The microstructural morphology of the optimized WRSG geopolymer at various ages is shown in Figure 11. At one day, abundant granular gel substances were formed on the surface of WRSG, accompanied by unreacted flaky metakaolinite. After three days of curing, the loose gel particles adhering to the WRSG surface decreased in quantity, while continuous accumulation led to the formation of flocculent cementitious gel. By the 7th curing day, the unreacted metakaolinite lamellae in the specimens were significantly reduced, indicating a deepened reaction degree. The gel matrix became denser, and the pores and interfaces between the unreacted metakaolinite and the gel phase became more distinct. During the curing period from 7 to 28 days, the state and appearance of the specimens showed no significant changes and tended to be stable. The reaction rate of geopolymers in the early stage is primarily governed by the dissolution rate of the system. As the reaction proceeds, the yield of the gel phase increases continuously, and the framework structure of the system is gradually refined. At this stage, the microstructure of geopolymers can be observed to become increasingly dense under a scanning electron microscope.

Although Figure 9 shows zeolite existing in the geopolymer, hardly any zeolite was observed in Figure 11, especially at 7 and 28 days. This demonstrates that sodium–aluminum–silicate–hydrate gel is the main hydration product of WRSG geopolymer. The influence of zeolite and crystalline minerals in the long-term performance of WRSG geopolymer will be studied in the future.

4. Conclusions

This study investigates the effects of sodium silicate modulus, alkali content, and water-to-soil ratio on the mechanical properties and hydration characteristics of WRSG geopolymers. The optimal mix proportion of WRSG geopolymers was determined, and the geopolymerization mechanism of WRSG in alkaline solutions was elucidated. The key conclusions are as follows:

(1)

The compressive strength of the geopolymer increases and then decreases with increasing sodium silicate moduli, increases with increasing alkali equivalent, and decreases with increasing water-to-soil ratio. The optimal mix proportion is defined as a sodium silicate modulus of 0.9, an alkali content of 14%, and a water-to-soil ratio of 0.5.

(2)

The early compressive strength and reaction degree of the geopolymers increase rapidly. The 1-day compressive strength reaches 48% of that at 28 days.

(3)

The hydration products of the geopolymers are dominated by amorphous sodium aluminosilicate hydrate gels, interspersed with primary minerals such as quartz and albite.

(4)

With increasing curing age, the gel content on the WRSG surface increases continuously. The framework structure of the system is gradually refined, and the microstructure becomes denser.