Mechanical and Thermal Properties of Self-Curing Metakaolin-based Geopolymer Mortar in Hot-arid Climate

2.1. Dry Ingredients

This study utilized moderate-grade kaolin clay, sourced from Tabelbala in southern Algeria, as the primary aluminosilicate material for producing metakaolin-based geopolymer mortars (Fig. 1). The kaolin clay was processed by crushing it with a jaw crusher, followed by grinding it using a disk mill, and then sieving it to achieve a particle size of less than 20 µm. The resulting powder was calcined in a muffle furnace at 900 °C for durations of 1, 2, 3, and 4 hours, with a controlled heating rate of 10 °C per minute, to convert the kaolin into metakaolin (MK). The optimal thermal activation conditions for the kaolin were identified by evaluating the strength activity index (SAI %) [32].

The raw and calcined clay samples were analyzed by X-ray fluorescence, Fourier Transform Infrared (FTIR) spectra, and X-Ray Diffraction (XRD) analysis to determine the chemical and mineralogical compositions. Then, physical characteristics such as BET-specific surfaces were defined. We added Commercial Silica Fume (SF), obtained from GRANITEX (Algeria region) with a specific surface >15 m²/gr and a high percentage of amorphous silica (SiO₂ = 94.57%), to the geopolymer mixtures to enhance the geopolymerization process.

The reference sample mortars were prepared using Ordinary Portland Cement (OPC) type CEMII /A-p 42.5 N from STG SIDI MOUSSA Adrar. Mortar specimens were prepared using DIN-EN 196-1 standard sand (S) with a high percentage silica content of 98% as fine aggregate. Table 1 presents the chemical composition of the starting materials, metakaolin and silica fume, as determined by an X-Ray Fluorescence (XRF) spectrometer.

2.2. Alkaline Activators

The main ingredients in the alkaline activator solutions were sodium hydroxide (NaOH), potassium hydroxide (KOH), and liquid sodium silicate (Na₂SiO₃). We purchased all these ingredients from Sigma Aldrich. The NaOH powder has 99% purity with a 2.13 g/cm³ density, and the KOH pellets have 85% purity with a 2.04 g/cm³ density. At the same time, the sodium silicate solution contains 10.6% SiO₂ and 26.5% Na₂O with a density of 1.39 g/cm³ and a pH of 11.30. We prepared the NaOH and KOH solutions with 19 M and 13.389 M molarities and cooled them at room temperature (Fig. 2).

Four distinct activator solutions were prepared by combining NaOH and KOH solutions with liquid sodium silicate or silica fume. These solutions were left to rest at room temperature for 24 hours to allow the formation of soluble silicate species. The prepared activator solutions were mixed with metakaolin and standard sand and mechanically stirred until a homogeneous mixture was achieved. The resulting mixtures were poured into 40 × 40 × 160 mm moulds for compressive and flexural strength testing. The moulds containing the resulting mixtures were vibrated for 2 minutes before curing to minimize air entrapment.

2.3. Mix Proportions

Table 2 presents the proportions of each ingredient used in the formulation of 10 different geopolymer mix types, where the quantity of one ingredient was varied in each mix. Each mix was repeated thrice to ensure accuracy, resulting in 60 mixes. The samples were then subjected to two different curing methods: ambient and solar drying.

The calculations accounted for the total moles of silica provided by the metakaolin and silica fume. Similarly, we determined the water content by considering the contributions from all sources within the alkaline activators, including the hydroxide, potassium, and sodium silicate solutions, and added water. This experiment proposes replacing the sodium silicate solution with silica fume, using 49% silica fume by the total weight of the binder. All geopolymer mortars were prepared with a binder-to-sand ratio of 1:2.75. The liquid-to-binder ratio varied between 0.3 and 0.55. The SiO₂/M₂O ratio of 1 was selected to optimize gel formation while avoiding excessive alkali content that could negatively impact durability [33]. Varying the molar ratios SiO₂/Al2O (3.3–3.6), M₂O/SiO₂ (0.28–0.3) and H₂O/Na₂O (8.1–14.7) were investigated (Table 3).

The samples were prepared for comparison using a mixture of different materials, with the reference cement mortar being prepared according to the specifications outlined in EN 196-1 [34]. The reference cement mortar contains 1350 g of standardized sand, 450 g of cement, and 225 g of water. This mixture served as the baseline for comparison with other samples, which were prepared using different methods to test the properties of the mortar and the resulting reactions.

2.4. Curing Condition

The eco-friendly treatment of mortars was selected to minimize costs and offer an environmentally sustainable alternative. In this context, we opted to utilize the abundant solar energy in hot regions to achieve this objective. The samples were divided into two groups: One of the parts of samples was stored at the same ambient temperature (E), while the other was placed outside to achieve sun resistance conditions (S). Solar curing was conducted outdoors with temperatures ranging from 35°C to 45°C, as measured using a calibrated thermometer. Samples were exposed to direct sunlight daily for 28 days. The temperature and solar radiation recorded during the experimental period are presented in Fig. (3).

3.1. Compressive and Flexural Strengths

The mechanical property tests of specimens, cured at 3, 14, and 28 days, were conducted following the procedures outlined in EN 196-1 [29] for MK-based geopolymer mortar and OPC reference mortars. The flexural strength of the mortar (40 mm × 40 mm × 160 mm) prisms was measured using the three-point bending test method by standard testing protocols. In contrast, the compressive strength of the mortar was determined by subjecting the specimen to a uniaxial load in a compression testing machine, ensuring accurate measurements of load-bearing capacity. Each test was conducted at least three times. For each experimental condition, a total of 20 geopolymer mixes were prepared. These included variations in curing temperature (ambient and solar). Each condition was repeated in triplicate to ensure reproducibility and statistical validity.

3.2. Drying Shrinkage and Bulk Density Test Method

Thirty-two different geopolymer mortar mixtures (M1, M4, M7, and M9) were prepared to study drying shrinkage (8 mixes), thermal conductivity (8 mixes), bulk density, and porosity (8 mixes). The variation in curing conditions (ambient and solar) was ensured within the experimental design, with an appropriate number of samples allocated for each curing condition to guarantee the accuracy and reproducibility of the results.

The drying shrinkage test was conducted according to French standard NF P 15-433 [35] using a length comparator with a dial gauge accuracy of 0.0001 inch. Three mortar specimens (40 × 40 × 160 cm) were produced for each type, OPC and MK-based geopolymer, with a water–binder ratio of 0.5 Fig. (4).

After the test pieces were moulded, they were cured under standard conditions (20 ± 1°C and relative humidity, RH 90%) for 1 d, and then the mould was removed. Then, one part was conserved in the laboratory room under environmental temperature conditions, and the other was placed outside to achieve solar cure conditions.

The bulk density and porosity of the prism specimens were evaluated following ASTM C642-97 [36]. The specimens were heated at 105°C for 24 hours, cooled, and weighed (Wd). After soaking in water for 48 hours, the saturated surface dry weight (Wi) was measured. The specimens were then boiled for 5 hours and cooled, and the saturated surface dry weight (Wb) was recorded. The apparent weight in water (Ws) was obtained by suspending the specimen.

We measured thermal characterization tests on specimens of dimensions 100×100×50 mm³ in the ENPOran laboratory using the ISOMET 2104 equipment.

Additionally, residual fragments from the mortar compression tests were ground and sieved to a particle size < 45 μm. These samples were dried at 60°C for 24 hours before being subjected to X-ray diffraction analyses (XRD).

4.1. Metakaolin Activation

Thermal activation is widely recognized as a critical process for enhancing the reactivity of kaolin clay. The findings presented in Fig. (5) indicate that the optimal conditions for achieving the highest strength activity index occur at a calcination temperature of 900 °C for one hour.

During the calcination of raw clay, the hydroxyl groups (OH)₂ are removed as water (H₂O), transforming the crystalline structure into an amorphous material and enhancing its reactivity [37, 38]. During the calcination of raw clay, the hydroxyl groups (OH)₂ are removed as water (H₂O), transforming the crystalline structure into an amorphous material and enhancing its reactivity [37, 38]. Infrared spectroscopy (FTIR) confirmed the dehydroxylation of kaolin clay, indicated by the removal of the stretching (-oh) bands at 3617–3696 cm⁻¹, as shown in Fig. (6a). The chemical composition of the treated kaolin aligns with astm c-618 standards for natural pozzolanic materials, supporting its application in concrete [39].

The high-temperature calcination weakens kaolinite’s diffraction peaks and enhances quartz’s diffraction peaks (Fig. 6b). Based on the analysis of the crystallinity of the raw and calcined samples by Origin software, it was found that the crystallinity of kaolin clay before calcination is 56.38% and that after calcination for 1h is 21.77%. Due to the moderate purity of the Tabalbala deposit, the obtained metakaolin contained about 27.87% of impurities, mainly quartz (18.9%), and the remaining 72% comprised metakaolinite and metaillite as shown in Fig. (7).

Al₂Si₂O₅(OH)₄ (kaoliniteAl₂Si₂O₇ (Metakaolinite)+ 2H₂O(Water) [40]

KAl₃Si₃O₁₀(OH)₂ (Illite) KAl₃Si₃O₁₀ (Metaillite) + H₂O(Water) [41]

The above results allowed us to determine the amount of amorphous silica and aluminium in the calcined clay. The geopolymer formulation will include SiO₂=38% and Al₂O₃=32%.

4.2. Compression and Flexure Strengths

The compressive strength results for geopolymer and Portland cement mortars at 3, 14, and 28 days are shown in Fig. (7). Most geopolymer samples displayed significantly lower early compressive strength (3 days) than the control specimens (TE and TS). Example: M3E (12.8 MPa, ambient) and M3S (14.3 MPa, solar) are much weaker than TE and TS at 3 days. The early strength of geopolymer materials is indeed influenced by slower geopolymerization, which relies on the alkaline activation of aluminosilicate precursors. This process typically results in lower initial strength than the rapid hydration of Ordinary Portland Cement (OPC), which forms calcium silicate hydrate (C-S-H) and provides higher initial strength [42, 43].

In terms of strength at 14 and 28 days, solar curing significantly enhances geopolymer strength, often surpassing TS. Example: M6S achieves 35.6 MPa at 14 days and 37.4 MPa at 28 days, outperforming TS (30.7 MPa and 32.8 MPa, respectively). Ambient-cured geopolymers like M6E (34.3 MPa at 14 days, 36 MPa at 28 days) also approach or exceed TE’s compressive strength.

Ambient-cured geopolymers showed strength enhancement over time. A similar trend was observed in solar-cured samples, whose long-term compressive strength may equal or exceed that of ambient-cured OPC mortars. This strength gain is associated with the continuous geopolymerization process, which contributes to improvements over time [11, 44].

As shown in Fig. (7), the early strength (3 days) of TE and TS outperforms most geopolymer samples in early flexural strength. For example, M3E (4.1 MPa) and M3S (4.25 MPa) are among the few geopolymer samples to meet or exceed TE's flexural strength at 3 days. In terms of strength at 14 and 28 days, solar-cured geopolymers often surpass TS in flexural strength at later ages. For example, M6S reaches 8.6 MPa at 14 days and 10.8 MPa at 28 days, well above TS. Ambient-cured samples like M6E also exhibit superior flexural strength (7.4 MPa at 28 days) compared to TE (6.2 MPa).

4.3. Effect of using NaOH or KOH as an Alkaline Activator

Silica fume combined with KOH results in higher strength due to the larger ionic radius of K⁺ compared to Na⁺, forming a more stable aluminosilicate gel enriched with reactive silica [45]. As a result, network formation is enhanced, and the geopolymer matrix becomes denser, leading to improved compressive and flexural strength. While the sodium silicate solution already contains a high proportion of reactive silicates, the larger K⁺ ions can delay condensation and lower compressive strength during early curing stages due to reduced aluminosilicate dissolution. This results in a less compact geopolymer network. Na⁺ interacts more effectively with the sodium silicate, enhancing the dissolution and polymerization processes and leading to better strength development than KOH [46].

4.4. Effect of Curing Conditions on Compressive and Flexural Strengths

The effect of curing conditions on the compressive and flexural strengths of GP mortars is also evaluated in this study and is shown in Fig. (8a and 8b). The results demonstrated that solar curing accelerates geopolymerization reactions, leading to significantly higher compressive and flexural strength values at all ages (3, 14, and 28 days) than ambient curing.

Fig. (8a-d) demonstrates that solar curing enhances compressive strength by approximately 3–5% while increasing flexural strength by 10–25%, primarily due to the elevated temperatures provided by sunlight. These higher temperatures increase the solubility and reaction rates of silica and alumina in the source material, which are crucial for forming the geopolymer gel (e.g., sodium aluminosilicate hydrate (N-A-S-H), potassium aluminosilicate hydrate (K-A-S-H), or calcium aluminosilicate hydrate (C-A-S-H)). Furthermore, the elevated temperatures create conditions that facilitate faster condensation of the dissolved precursors, resulting in a denser and stronger geopolymer matrix [47].

4.5. Effect of H₂O/M₂O Ratios

Compressive strength often diminishes when H₂O/M₂O ratios increase, as elevated water content develops a porous structure during drying. Fig. (8a, b) 1 demonstrate that samples with lower ratios (e.g., 8.1 or 8.44) consistently achieve higher compressive strengths (e.g., up to 37.4 MPa under solar curing).

In contrast, elevated ratios (e.g., 13.5 or 14.4) achieve weaker geopolymers, evidenced by decreased strengths over the same curing periods. (Fig. 8c, d) demonstrate that reduced H₂O/M₂O ratios (e.g., 8.1) provide increased flexural strengths attributable to the denser and more cohesive gel network. Conversely, elevated ratios result in decreased flexural strength due to increased porosity. Compressive and flexural strengths often diminish when H₂O/M₂O ratios increase, as elevated water content develops a porous structure during drying. In contrast, elevated ratios achieve weaker geopolymers, evidenced by decreased strengths over the same curing periods [48].

4.6. Drying Shrinkage

The variation in mortar shrinkage over time for different activators (NaOH, KaOH, NaOH + Na₂SiO₃, and KaOH + Na₂SiO₃) is illustrated in Fig. (9). At a Water-to-Binder (W/B) ratio of 0.5, significant differences were observed after 3 days, with shrinkage values ranging from a minimum of 187.5 μm (M4E) to a maximum of 5237.5 μm (M9S) with a convergence of values between the other activators. We estimated the increase in shrinkage at 153% than the reference sample. The same trend was observed at 7 and 14 days. However, after 28 days, significant differences in shrinkage values were observed, with values ranging from 4331.25 μm to 4587.5 μm for M4S and M9E, and from 6918.7 μm to 8718.75 μm for M1E, M1S, M7E, and M9S. M7S recorded a maximum value of 10,875 μm, representing an increase of 224%.

The addition of silica fume to metakaolin-based geopolymer mortars reduced the drying shrinkage in MK geopolymer mortars by enhancing the matrix's density and reactivity. As a highly reactive material, silica fume contributes additional silicate content, promoting better geopolymerization and filling voids in the geopolymer matrix. Consequently, porosity decreases, and the capillary tension caused by water loss, both primary factors in drying shrinkage, is minimized [49].

Additionally, silica fume and solar curing may significantly enhance shrinkage resistance. The heat from the sun-curing process activates the pozzolanic effects of silica fume, improving bonding within the geopolymer network and reducing the risk of shrinkage-induced cracks [50].

4.7. Bulk Density, Porosity, and Thermal Conductivity

Fig. (10) illustrates the compressive strength, bulk density, porosity, and thermal conductivity of mortar specimens, including OPC and geopolymer mortars with various binders (e.g., silica fume, sodium silicate) and alkaline activators (NaOH, KOH), under ambient and solar curing conditions.

At a fixed w/b ratio, metakaolin geopolymer mortars demonstrate lower bulk density than OPC mortars, indicating increased porosity and enhanced thermal conductivity.

Silica fume mortars exhibit lower thermal conductivity (0.633–1.17 W/mK), primarily due to their higher porosity (20.65%–23.38%) and lower bulk density (1.92–1.97 g/cm³). Under solar curing, thermal conductivity increases slightly due to reduced porosity and enhanced microstructural densification. In contrast, sodium silicate mortars have a higher thermal conductivity (1.19–1.68 W/mK), attributed to their lower porosity (13.09%–15.21%) and higher bulk density (2.01–2.09 g/cm³). The incomplete geopolymerization in high Si/Al ratio samples (e.g., 3.6) results in less dense geopolymer matrices [51]. Additionally, the porous structure significantly hinders heat transfer within the geopolymer network, effectively reducing its thermal conductivity (k) [52].

4.8. XRD Analysis of Metakaolin Geopolymer Mortars

In the conditions provided by the experiment, the formation of an amorphous geopolymer gel was expected due to the high silicon-to-aluminium (Si/Al) ratios (3.6 and 3.3), which promote the development of the amorphous gel while suppressing the growth of crystalline zeolite phases [53, 54]. The addition of silica fume increased the availability of reactive silica, which may lead to the stabilization of the amorphous gel and prevent secondary crystallization, including zeolite formation, even under high alkalinity conditions (13.389M and 19M), consistent with previous studies [55].

Based on X-ray diffraction analysis Fig. (11), Leucite (PDF: 01-076-2298) was observed in most sun-cured samples, likely due to pore closure and leucite formation as the microstructure evolves under elevated temperatures [56]. Quartz (PDF: 01-070-3755) was the dominant mineral in the samples, showing strong peaks at angles (20.8° and 26.6° in 2θ), which overshadowed weaker signals associated with zeolite and the amorphous gel, consistent with the findings of Juengsuwattananon et al. [57].

FUTURE WORK

This study is a preliminary independent work, i.e., original experimental research on using a new binder based on local materials within a 3-phase timeframe: 3 days, 14 days, and 28 days, carried out in a region with an arid climate. It will be followed by another independent study focusing on the property of 'durability,' with the duration extended to 90, 180, and 365 days to assess the long-term durability of the mortar with this new binder.