1 Introduction

Cement is a major contributor to greenhouse gas pollution, climate change, and ozone layer depletion because of the sulfur oxide, nitrogen oxide, and carbon monoxide it emits. Due to the enormous production volume of conventional concrete, specifically ordinary Portland cement (OPC), the environment is also adversely affected. First and foremost, manufacturing cement involves mixing limestone, silicates, and other substances in a cement plant. One of the major causes of global warming is the cement industry's emissions of carbon dioxide (CO2) [1, 2]. Around 0.89 tons of CO2 are generated during the processing of 1 ton of OPC. From the cement industry, fossil fuels used in electricity generation produce a large proportion of CO2. There are approximately 4100 million tons of cement produced annually in the world; these factors make it difficult to meet the criteria for sustainable growth [1]. The search for alternative OPC materials in concrete has been carried out in several studies. This aim was to produce eco-friendly concrete by improving concrete properties, enhancing microstructure, and reducing CO2. Sustainable geopolymer concrete (GC) is produced using various aluminosilicates found in nature and industrial waste. As a result, the latter uses our resources sustainably, has an energy-efficient process, and is environmentally friendly. Geopolymers are produced through chemical reactions between aluminosilicates (metakaolin (MK), fly ash (FA), slag, volcanic ash, etc.) and alkaline solutions (sodium or potassium silicate). Binder is formed by polymerizing silica and alumina from the aluminosilicate in an alkaline-activated solution [3].

Heat resistance is critical property of construction materials. Traditional concrete properties deteriorate under high temperatures due to physical and chemical changes [4,5,6,7]. Water loss, strength loss, reduction of modulus of elasticity, crack formation, expansion, and spalling are irreversible changes [8]. The heat resistance of traditional concrete is significantly influenced by the type of aggregate and binder used in its mix design. This is because of the heat and duration of the fire, the extent of the structural component, and the moisture content of traditional concrete [4, 8,9,10]. Color changes, cracks, and spalling are the first signs of heat-damaged concrete [11]. Traditional concrete is less fire resistant when it has a small invalid proportion. When exposed to 800 °C, regular- and high-strength concrete loses 60% of its primary compressive strength [7, 12]. The modulus of elasticity is also affected by fire, just as it is with strength [13].

A geopolymer's inorganic framework makes it more thermally stable than traditional cement. Geopolymer concrete has the properties that make it suitable for high-temperature applications, such as wall panels, thermal insulation, and furnace linings. Thermal expansion, melting point, spalling, strength retention, and thermal conductivity are several macroscopic properties that affect an adhesive’s suitability for high-temperature applications. The thermodynamics, dehydration, morphology, and phase stability of geopolymers are all factors that determine their thermal stability. Geopolymer concrete has recently been studied for its thermal properties [7, 14].

Various studies have verified the performance of geopolymer concrete after heating up to 800 °C [15,16,17]. During a fire, the temperature is typically over 800 °C for 30 min, then slowly rises to approximately 1000–1100 °C within 2–2.5 h [18]. GC is not capable of performing at full mechanical strength over 600 °C [19, 20] because geopolymer pastes and aggregates differ microstructurally and are thermally incompatible [21].

Under ambient conditions, high-quality geopolymer concrete has similar mechanical properties to OPC. Fly-ash-based geopolymer concrete, cured at 60–80 °C temperature, exhibits high early mechanical strength and low dry shrinkage [22], excellent fire resistance [15, 23] and durability [24,25,26]. Zhang et al. [27] analyzed the influence of elevated temperatures on geopolymer concrete compressive and split tensile strengths. GC was produced by mixing MK and FA precursor mixtures with coarse and fine aggregates and adding an alkaline activator. Compressive and splitting tensile strengths of GC and OPC concretes are not significantly affected by exposure temperatures under 300 °C. At 300 °C, GC strength increases slightly [24, 28]. In the long term, at ambient temperatures and after exposure to high temperatures, GC exhibits similar compressive strength to OPC. However, it has more tensile strength than OPC.

Global energy consumption has increased dramatically over the past few decades. Most of the energy demand comes from heating and cooling buildings. Better construction techniques and improved materials technology can significantly reduce energy consumption for maintaining a comfortable indoor temperature. Thermal energy storage systems are one way to save and conserve energy [7, 29]. Jacob, Rhys, et al. used fly ash and black slag to store thermal energy at a high. When shaped and cured at low temperatures, geopolymer can create a variety of structural shapes. This reduces processing costs. The density and heat capacity of geopolymers were tested. Heat capacity increased from 1.38 kJ/kg at 100 °C to 1.64 kJ/kg at 450 °C, indicating excellent specific heat. The economic analysis suggests that using geopolymer fillers significantly reduces the cost of traditional 2-tank, sensible, and single PCM molten salt systems. Additionally, the geopolymer filler compares well with other potential fillers when it comes to thermal energy storage (TES) at high temperatures [30]. In this work, the material developed is meant for use in sensible thermal energy storage in solar energy power plants. This is because the heat in the storage system reaches around 500 °C. That is why the heat resistance of the materials prepared was studied in that temperature range [30]. There were two partial substitutes chosen (5 and 20 wt.%) for fly ash. Nine metakaolins were evaluated, each containing different levels of alumina. The purpose of this study is to investigate the effect of small and large substitution ratios based on some experiments. Moreover, MK is formed through the calcination process and therefore has a higher environmental impact. Because of this, only partial replacement was taken to keep the environmental impact low [31].

For sensible thermal energy storage applications when exposed at temperatures up to 500 °C, this paper investigates the influence of chemical composition and fineness of blends of FA-based geopolymers. Heat resistance behavior was correlated with fresh and hardened properties of MK-FA-based geopolymer pastes. An investigation was conducted on the mechanical, thermophysical, and microstructural properties of FA-MK geopolymer paste using slump tests, water absorption, porosity, X-ray diffraction (XRD), thermogravimetry/differential scanning calorimetry (TG/DSC), and scanning electron microscopy (SEM) at ages of 7 and 56 using elevated temperatures of 90, 300, and 500 °C.

2 Experimental procedure

2.1 Source materials

The FA used in this work was provided by Baumineral GmbH, Herten, Germany. Nine different types of MK used in this work were provided by Temcon Solutions GmbH, Alsfeld, Germany. MK has been classified based on aluminum oxide content due to its importance in heat resistance. The chemical compositions of FA and MK were determined by X-ray fluorescence (XRF), as listed in Table 1. Figure 1 illustrates the FA and MK particle size analysis performed by laser granulometry using Mastersizer 2000 from Malvern Instruments, UK. Table 2 shows the specific surface area and characteristic diameter of FA and MK determined according to the BET method and laser granulometry. FA and MK were activated with an alkaline activator made by adding technical-grade sodium silicate (Na2SiO3) and sodium hydroxide solution (NaOH). The chemical composition of Na2SiO3 was SiO2 = 30.2 wt. %, Na2O = 14.7 wt. %, H2O = 55.1 wt. % and the silica modulus (Ms = SiO2/Na2O mol/mol) equal to 2.12. The NaOH of 12 M was prepared by mixing sodium hydroxide pellets of 97–99% purity with distilled water. The alkaline activator was prepared by mixing Na2SiO3 and 12 M NaOH solution at a constant mass ratio of 5:2.

Table 1 Chemical composition in wt.% of FA and MK using XRF
Full size table
Fig. 1
figure 1

Particle size distribution of the FA and MK

Full size image
Table 2 Specific surface area according to BET and diameter characteristics of FA and MK
Full size table

2.2 Water demand

To investigate the flow and rheological properties of geopolymers, the water demand was first evaluated for all raw materials. The water demand was measured for FA and MK using the Puntke-method [32]. In this method, 100 g of binder was weighed into a beaker, then a spoon was added and the scale was adjusted. Water was then continuously added drop by drop and stirred and the beaker was tapped until additional water was absorbed by the binder. Finally, the beaker with the spoon, binder, and water was balanced. The water demand ratio was calculated using formula 1.

$$ n_{w} = \left( {Final \, reading/1.0} \right)/\left( {First \, note/\rho_{B} + \, Final \, reading/1.0} \right) $$
(1)

where:nw: water demand ratio, ρB: Binder density

2.3 Specimens preparation

The samples consisted of 95 wt.% FA + 5 wt.% MK and 80 wt.% FA + 20 wt.% MK, respectively. The sample name consists of the name of the MK used and its percentage by mass in the sample. MK1_5, for example, stands for a mixture of 95 wt.% FA and 5 wt.% MK1. The liquid activator to solid ratio was 0.5. The geopolymer paste specimens were prepared by mixing the FA and MK with the prepared alkaline activator liquid in a mixer for 2 min. After that, the mixture was poured into 20 × 20 × 20 mm3 cubic molds. The specimens were then vibrated on a vibrating table for 1 min to release any residual air bubbles. Subsequently, the molded specimens were wrapped using a thin plastic sheet to prevent water evaporation.

2.4 Slump flow

The slump flow of the fresh paste was measured before casting. In the slump flow test, workability was evaluated after mixing using a truncated cone (upper diameter = 70 mm; base diameter = 100 mm; height = 50 mm).

2.5 Curing regime

All wrapped molds were cured at room temperature 20 °C for periods of 7 and 56 days. The relative humidity was kept constant (approximately 70%).

2.6 Elevated temperatures exposure method

A part of the prepared geopolymer paste specimens was further exposed to elevated temperatures of 90, 300, and 500 °C for 7 and 56 days. They were placed in a furnace and heated at a fixed heating rate of 10 °C/min. The geopolymer specimens were kept at each elevated temperature for 2 h; afterward, the specimens were allowed to cool inside the furnace to room temperature. Meanwhile, the unexposed specimens were left in ambient conditions. The compressive strength test for the exposed geopolymers to elevated temperatures was performed one day after heating. Thus, all specimens were left to complete the aging period before testing their physical and mechanical properties. The compressive strength measurements were an average of three specimens. The mechanical properties for all specimens was tested using a compression testing machine (Toni Technik, Berlin, Ger many) according to EN 12,390–3.

2.7 Water absorption and apparent porosity

The water absorption and apparent porosity of the unheated and heated geopolymer paste specimens at 56 days were measured, and calculations were made using formulas 2 and 3.

$$ Water absorpation \left( {wt. \% } \right) = \frac{MSSD - MDry}{{MDry}} \times 100 $$
(2)
$$ Porosity \left ( {vol. \% } \right) = \frac{MSSD - MDry}{{MSSD - MW}} \times 100 $$
(3)

where:MSSD: mass of saturated surface dry sample in air. MDry: mass of dry sample in air. MW: mass of sample hanging on balance arm immersed in water.

2.8 Thermogravimetry/differential scanning calorimetry

The mass loss and heat change with rising temperature were measured for the unheated geopolymer paste specimens at 28 days by thermogravimetry/differential scanning calorimetry (TG/DSC). The measurement was conducted at a heating rate of 5 K/min with a gas flow (synthetic air) of 70 ml/min using Mettler Toledo TGA/DSC3 + . Fragments of the geopolymer paste specimens used for strength test analysis were powdered and analyzed in the TGA/DSC.

2.9 Phase composition investigation using X-ray diffraction analysis

X-ray diffraction (XRD) test was performed to investigate the mineralogical phase composition of both unheated and heated geopolymer pastes. The test was carried out on 56 days old powdered samples using Panalytical Empyrean Diffractometer (Malvern Panalytical Ltd, Malvern, UK) with Ni filter CuKα radiation (Cu Kα wavelength: 1.54060 Å), operating at 40 kV and 40 mA. In collecting data sets, the 2θ step size was 0.0131°, the counting time step was 0.1°/min, and the 2θ range was 5° to 65°. Rutile was used as an internal standard for Rietveld quantification.

2.10 Microstructural analysis of FA-MK geopolymer paste

The microstructure of the specimens was studied by Scanning Electron Microscope (SEM) analysis using ZEISS GeminiSEM500 NanoVP (Oberkochen, Germany), which operates in low-vacuum mode with 15 kV acceleration voltage. Fractured surfaces and fractured specimens were used in this test.

3 Results and discussion

3.1 Physical properties of FA-MK geopolymer paste

3.1.1 Water demand of raw materials and slump test of resulting geopolymer pastes

Figure 2 displays the results of water demand and the surface area on all the types of FA and MK powder investigated. In this figure, samples MK9 and MK8 have the highest water demand. That is likely due to their higher packing density. They also needed more water to fill all voids between particles in the packing due to their specific surface area. As the surface area increases, the water demand increases. This correlates with the evolution of the mean diameter of each raw material determined by particle size analysis (Table 2). The calculation of the mean diameter from Table 2 shows that MK8 and MK9 have the lowest values, 2.91 µm, and 3.30 µm, respectively. In contrast, the mean diameter of FA, which has the lowest water demand, is 38 µm.

Fig. 2
figure 2

Results of water demand and specific surface area for all the types of raw materials

Full size image

Figure 3 illustrates the results of slump flow tests of all the geopolymer pastes investigated. As compared with the results of the water demand and particle size distribution of raw material, one can see that geopolymer mixes with MK4 have the maximum flowability. Geopolymer obtained with metakaolin MK8 and MK9 has the lowest flowability. Also, the higher the percentage of MK in the paste, the lower the flowability. Due to their higher packing density and fineness, the mixes require a more liquid solution to become workable. As aluminum oxide percentage increases in MK, water demand increase.

Fig. 3
figure 3

Results of slump flow for all geopolymer paste investigated

Full size image

3.1.2 Water absorption and apparent porosity of geopolymer pastes

Figure 4 shows the water absorption and apparent porosity in all geopolymer pastes investigated (the mixtures MK1_5, MK2_5, MK3_5, MK1_20, MK2_20, and MK3_20 were not tested due to their poor mechanical and thermal properties). As the temperature increases, water absorption and porosity increase, and they stabilize at 300 °C. It is observed that water absorption depends not only on the type of MK but also on the replacement rate of metakaolin. For example, the water absorption of MK4_5 is higher than the water absorption of MK4_20 at different temperatures. The other MK used behaves similarly to MK4 in terms of water absorption at 5 wt. % is higher than at 20 wt. % in each case. Also, it is observed that apparent porosity does not only depend on the type of MK but also on the replacement rate of metakaolin. For example, the porosity of MK4_5 is higher than the porosity of MK4_20 at different temperatures. As with water absorption, the behavior is the same for the other MK used; the porosity at 5 wt.% of MK is higher than at 20 wt.% in each case. When the aluminum oxide content increases; the reaction increases, especially at early ages, and the matrix becomes more cohesive and denser, thus reducing both porosity and water absorption.

Fig. 4
figure 4

Results of a water absorption, b apparent porosity for all geopolymer pastes investigated

Full size image

3.2 Mechanical and thermo-physical properties of FA-MK geopolymer pastes

3.2.1 Role of MK on compressive strength evolution at room temperature

The results of the compressive strength evolution at 7 and 56 days of the unheated geopolymer with different types and replacement rates of MK are shown in Fig. 5. The mixtures MK1_5, MK2_5, MK3_5, MK1_20, MK2_20, and MK3_20 have poor mechanical properties at room temperature, as shown in Fig. 5 and Fig. 6 and were therefore not tested at elevated temperatures. The mixture that contains only fly ash (FA_100) does not give any strength after 7 days, due to the slow reaction of FA compared to MK at room temperature. The compressive strength increases significantly with the age of all samples. It ranges between 17 and 43 MPa at 7 days, and 29–81 MPa after 56 days. That is due to the geopolymerization process and the high volume of reaction products being formed over time. It is known that the more reactive phases in raw materials undergo more reactions with time. This leads to improved physical and mechanical properties [33, 34]. Therefore, the rate of strength improvement from 7 to 56 days can be correlated with the reactivity of the different metakaolins used. To highlight this, the rate of compressive strength improvement was calculated as the difference between the compressive strength at 56 and 7 days. As a result, the mixes obtained with MK9 developed a strength gain of up to 32 MPa from 7 to 56 days. In contrast, those with MK4 gave a strength gain of up to 26 MPa in the same time range. It is observed that the higher the percentage of aluminum oxide in MK, the higher the compressive strength at room temperature. It is also observed that such a strength improvement rate does not only depend on the type of MK but also on the replacement rate of MK. The MK4_5 gains only 11 MPa as compared to 26 MPa for the MK4_20. On the other hand, MK9_5 gains 29 MPa while MK9_20 gains 32 MPa. All this confirms that the higher the reactive phases in the metakaolin, the greater the improvement in compressive strength with age. Thus, MK9 is the most reactive of all. In addition,the surface area affects the rate of reaction, so it affects the early strength only, but its effect is slight on the final ages.

Fig. 5
figure 5

Compressive strength evolution at room temperature of geopolymer paste with various types and percentages of metakaolin

Full size image
Fig. 6
figure 6

Compressive strength evolution at different temperatures at a 7 days, b 56 days

Full size image

3.2.2 Role of MK on the thermal stability of FA-MK geopolymer paste

Figure 6 reports the compressive strength of the geopolymer paste after exposure to elevated temperatures. It is observed that independent of the age of the specimens, heating the geopolymers at 90 °C systematically improves their compressive strength. That increase is explained by the acceleration of reaction kinetics. This is because fly ash is less reactive than metakaolin and generally requires heat treatment to develop its physical properties [33, 34]. Beyond 90 °C, the compressive strength decreases because of the start of the partial destruction of the geopolymer network. Generally, at the same temperature and age, the more aluminum oxide content in MK, the higher the compressive strength.

To correlate the heat resistance with the alumina content and replacement rate of metakaolin, the rate of compressive strength changes of selected mixtures was calculated and presented in Fig. 7. The latter shows that the strength change varies considerably with the type, replacement rate, and alumina content of metakaolin. Moreover, the age of the samples also plays a significant role. At 7 days (Fig. 7a), the compressive strength is improved, independently of the heating temperature. However, the rate of that improvement decreases with the rising heating temperature. This means that the geopolymer network, which is responsible for strength development, is partially destroyed. At 56 days, different behavior is observed in strength change with rising temperature (Fig. 7b). Here it is characterized by strength gain and loss depending on the type of metakaolin, replacement rate, and the heating temperature. For example, heating at 500 °C systematically decreases the compressive strength of all MK at both replacement rates below the value of the unheated geopolymer at the same age. Thus, the highest negative values indicate the extent of deterioration of the geopolymer network. Additionally, some mixes exhibit strength loss at 300 °C, indicating that certain metakaolins do not contribute to the strength maintenance of fly ash-based geopolymer under heat. That is closely related to the reactivity of metakaolin during geopolymerization, as discussed above. The latter is also responsible for the different behavior of the strength change at an elevated temperature after 7 and 56 days of fly ash-metakaolin-based geopolymer. In this study, it was observed that the 56-day samples were most affected by heat, the reason being that after 7 days, geopolymerization is still in process and thus most sensitive to heat in order to accelerate the reaction kinetics. In contrast, a higher reaction turnover was already achieved after 56 days, resulting only in a deterioration of the geopolymer network.

Fig. 7
figure 7

Compressive strength evolution after exposure to elevated temperatures at a 7 days, b 56 days

Full size image

3.3 X-ray diffraction (XRD) analysis

Figure 8 shows crystalline the phases from XRD of selected geopolymer samples at room temperature and after heating at different ages. The geopolymerization process may result in the formation of novel crystalline phases and the disappearance of some minerals initially present in the aluminosilicates. It is noteworthy that during geopolymerization, the amorphous phase interacts preferentially with the crystal phase. The observed mineral changes depend on the type and composition of the alkali solution and the type and content of the calcium source in the mix design [35]. Table 3 reports the semi-quantitative phase analysis of the geopolymer before and after heat exposure using the Rietveld refinement method.

Fig. 8
figure 8

X-ray patterns of MK9_20 at different temperatures at ages of a 7 days and b 56 days

Full size image
Table 3 Semi-quantitative phase analysis for geopolymer pastes at different temperatures at 56 days
Full size table

The pastes exhibited similar crystalline phases as raw FA, including quartz, mullite, and the amorphous phase. Accordingly, the crystalline phase of the pastes is mainly governed by FA. At 500 °C, nepheline appears as a newly formed crystalline phase due to the partial crystallization of sodium aluminosilicate hydrate gel (N-A-S-H). Nepheline is one of the more porous crystalline phases that can adversely affect the compactness of the gel structure, resulting in a weaker gel structure [36, 37]. Calcium in the samples is responsible for the formation of nepheline crystal structure after exposure to high temperatures [38]. Since FA contains more calcium oxide than MK, as shown in Table 1, FA is responsible for the formation of nepheline in the samples. Therefore, when the percentage of MK increases as a substitute for FA, the formation of nepheline decreases, and the compressive strength increases.

It is observed that such a geopolymer matrix improvement rate depends not only on the type of MK but also on the replacement rate of metakaolin. For example at 500 °C, the MK4_5 contains 1.4 wt.% nepheline against 0.8 wt.% for MK4_20. On the other hand, the MK9_5 contains 0.8 wt.% while the MK9_20 contains 0.5 wt.%. All this confirms that increasing the proportion of MK tendentially reduces the proportion of nepheline and improves the geopolymer matrix with age. Thus, metakaolin MK9 is the most reactive among all. The patterns of MK9_20 before and after exposure to elevated temperatures of 90, 300, and 500 °C at ages of 7 and 56 days are shown in Fig. 8. Also, it is observed that by increasing the percentage of aluminum oxide in MK, the proportion of nepheline decreased. After alkali activation, the amorphous product forms a broad hump centered on a 2θ angle close to 28. As the temperature rises, the glassy phase dissolves more rapidly in fly ash, and amorphous reaction products are formed more rapidly. When the percentage of aluminum oxide increases and the percentage of silica oxide decreases in MK, this leads to the reduction of amorphous structures. Rutile was added as an inner standard for quantification, so it appeared in X-ray patterns.

3.4 Thermal analysis

Figure 9 shows the thermal stability of selected geopolymer mixtures in terms of mass change with the rising temperature at 28 days. There is a mass loss of 13–14 wt.% in the TG/DTG curves, which starts almost at room temperature and ends around 650 °C. That mass loss occurs in two stages. The first stage occurs when physically adsorbed water and/or loosely bound water evaporates from the geopolymer matrix below 100 – 150 °C [39]. The second stage is the dehydroxylation of the geopolymer network. As the aluminosilicate raw materials dissolve, they form aluminates and silicates that undergo condensation by forming chemically bonded water through hydroxyl groups (OH). The results of this stage provide useful information about the extent of geopolymerization [40, 41]. In the range of 650 °C to 1000 °C, mass does not change significantly. It is observed that mass loss depends on the type of MK. For example, the FA_100 loses only 11.6 wt.% against 14.1 wt.% for MK4_20. Also, the MK9_20 loses 12.31 wt.% while the MK9_20 loses 13.46 wt.%. Also, it is found that by increasing the percentage of aluminum oxide in MK, mass loss decreases. According to the DSC curves depicted in Fig. 10, there is a succession of three peaks with maxima at 90, 300, and 500–600 °C, respectively. The endothermic peak at around 90 °C is due to the evaporation of free water within the paste, whereas the exothermic peaks observed at around 300 °C can represent the partial crystallization of the amorphous N-A-S-H gel (hydration products). The last exothermic peak occurred at 500–600 °C and is due to the total crystallization of N-A-S-H gel to nepheline (NaAlSiO4) (as confirmed by XRD).

Fig. 9
figure 9

TG-DTG representing 2 stages of mass loss for geopolymer pastes investigated

Full size image
Fig. 10
figure 10

DSC curves for geopolymer pastes investigated

Full size image

3.5 Microstructural analysis

The photographic image of sample MK9_20 before and after exposure to elevated temperatures is shown in Fig. 11 (MK9_20 at 56 days). The change in color from greyish to light brown is observed for the samples after exposure to temperatures of 500 °C. That can be correlated with the changes in the different phases with rising temperatures as observed from XRD and DSC curves.

Fig. 11
figure 11

Images of MK9_20 pastes before and after exposure to elevated temperatures (90, 300 and 500 °C)

Full size image

Using SEM, cross-sectional microphotographs of the boundary specimens before and after exposure to elevated temperatures were obtained to provide additional insight into the structure of the FA-MK-based geopolymer pastes (Fig. 12). Table 4 shows the results of SEM analysis for MK9_20 at room temperature and after heating at 90and calculations were made using formulas , 300 and 500 °C at the ages of 7 and 56 days.

Fig. 12
figure 12

SEM micrographs of MK9_20 at ages of 7 and 56 days at temperatures a room temp., b 90 °C, c 300 °C and d 500 °C

Full size image
Table 4 The results of SEM analysis for MK9_20 at the different temperatures at ages of 7 and 56 days
Full size table

4 Conclusion

In this paper, FA-MK geopolymer paste is exposed to 90, 300, and 500 °C in order to determine its characteristics before and after exposure. Geopolymers exposed to elevated temperatures showed significantly different performance and properties. Because MK9 has a low packing density, pastes containing it have lower workability, which requires more water to fill all voids between particles. Increasing exposure temperature increases water absorption and porosity of geopolymer pastes. With increasing aluminum oxide percentages in MK, water absorption and apparent porosity decrease. Water absorption and porosity are reduced with a higher proportion of MK. The higher percentage of aluminum oxide in MK9 makes it the most effective in terms of water absorption and porosity. Due to the slow hydration process for unheated pastes, the early compressive strength of FA-MK pastes was low. But at later ages, they gave a higher strength. Gradual heating speeds up geopolymerization reaction rates at 90 °C, which results in maximum compressive strength for geopolymer pastes. MK with more aluminum oxide has a higher compressive strength at the same temperature and age. Compared to pastes containing 5 wt.% MK, pastes containing 20 wt.% MK showed higher compressive strength at different temperatures because MK makes the paste denser but less workable. When exposed to high temperatures, pastes containing MK9 showed high compressive strength. According to measurements of porosity and SEM micrographs, the geopolymer network suffers insignificant damage during heating because of the enormous amount of very small pores, which escape moisture during heating.

Geopolymers deteriorated in strength at 500 °C, and the deterioration increased with increasing temperature. The thermo-physical results of FA-MK geopolymer pastes revealed that both dehydration and dehydroxylation contributed to strength decline at 500 °C. At 500 °C, the compressive strength after 56 days is higher than the compressive strength after 7 days. The strength of the bond is enhanced by the ongoing geopolymerization of unreacted FA. Geopolymerizing the remaining FA at 500 °C increases bulk density by filling small pores and tiny cracks. This is in agreement with the results of SEM micrographs. Moreover, the amount of nepheline after 7 days is lower than after 56 days.

Compressive strength was increased by polymerizing unreacted FA at 90 °C. Despite no changes in crystalline phase, the compressive strength of N-A-S-H gel (hydration products) decreased gradually at 300 °C due to dehydration and recrystallization. During partial crystallization at 500 °C, nepheline is produced as a new crystalline phase. The large density reduction is due to nepheline being one of the more porous crystalline phases that adversely affect the compactness of the gel structure.

As a result of crystallization of iron oxide and partial crystallization of N-A-S-H gel, the samples’ colors change from greyish to light brown after exposure to temperatures of 500 °C.