One-part slag-mullite geopolymer: role of solid activators on connected porosity and mechanical properties Authors: Shahryar Kashi, Ebrahim Najafi Kani Published: 2026 ABSTRACT AbstractGeopolymer cements with reliable mechanical and durability properties could be a promising environmentally-friendly alternative to Portland cement. Connected porosity (CP) and compressive strength (CS) are critical parameters governing the durability of slag–mullite geopolymers. This study examines the relationship between connected porosity and mechanical performance in one-part slag–mullite geopolymer mortars activated by various solid alkaline salts, including sodium- and potassium-based activators. The results reveal a clear inverse relationship between connected porosity and compressive strength. Geopolymer mortars activated with sodium silicate and potassium silicate at silica moduli of 0.6 and 0.8 exhibited the highest compressive strength, along with the lowest connected porosity and water absorption. These improvements are attributed to the enhanced alkalinity and high ionic solubility of the silicate activators, which promote a denser geopolymeric network. In contrast, mortars activated with calcium hydroxide and sodium sulfate showed significantly higher connected porosity and lower strength, likely due to calcium sulfate precipitation and its detrimental effects on microstructural continuity. The incorporation of mullite further improved the material performance by promoting the formation of more densified aluminosilicate network and refining the planar silicate framework, resulting in reduced porosity. Overall, connected porosity is demonstrated to be a robust and quantifiable indicator of durability, reflecting the resistance of one-part slag–mullite geopolymers to fluid ingress and degradation under aggressive environmental conditions.Keywords: One-part geopolymer mortar, Mullite, Connected porosity, Compressive strength, solid activatorSubject terms: Engineering, Materials science Introduction Along with the development of human communities and population growth, the demand for Portland cement as a popular material for constructing roads and buildings has seen a rise, especially in recent years. This rising demand has led to concerns about environmental consequences, knowing that the production of Portland cement is responsible for 4–10% of total CO2 emissions globally1–4. Given these concerns, it is necessary to propose an alternative cementing material with appropriate properties to replace Portland cement in the mentioned application areas. Geopolymers refer to a class of mineral adhesives that are produced upon a reaction between aluminosilicate-containing substances and alkaline activators – a process called geopolymerization. In contrast to Portland cement, the production of geopolymers is less energy-intensive and leads to fewer CO2 emissions. Based on the aforementioned points, geopolymers can be considered a serious and advantageous properties for a promising alternative to Portland cement4–6. Geopolymers can be categorized into two classes based on the particular activation method used: one-part and two-part. To produce a one-part geopolymer, only water is introduced into a solid mixture of precursors containing aluminosilicate and a solid alkaline activator. For two-part geopolymer system, it is essential to first prepare an alkaline activator solution. This solution is then introduced into the solid mixture of aluminosilicate-containing precursors2,4,7,8. As a critical characteristic for evaluating the performance of a concrete or geopolymer mortar, compressive strength is affected by a multitude of factors, including porosity. In general, compressive strength is inversely related to porosity, so that a geopolymer mortar or concrete has its porosity dropped as its compressive strength increases. It is then expected that a buildup of porosity decreases the compressive strength. On the other hand, both compressive strength and porosity are known to affect geopolymer durability. This has provided the ground for many studies on the porosity and compressive strength concerning their effect on the durability performance of geopolymers. Previous studies have reported on the performance of metakaolin- and slag-based alkali-activated concrete with fine and coarse aggregates to investigate possible correlations between the pore structure and compressive strength in metakaolin-based porous geopolymers, study durability and strength characteristic compressive strength of geopolymer mortars containing waste concrete powder (connected porosity), and analyze the relationship among porosity, reactivity degree, and compressive strength in fly ash-based geopolymers, evaluate the effect of the curing temperature and alkaline activators on the compressive strength and porosity of geopolymer mortars based on fly ash and blast furnace slag and palm oil fuel ash (POFA) in presence of liquid activators9–13. The one-part geopolymers have also been the subject of various studies on their porosity and compressive strength and their relationship. Zhao et al. studied the characteristic compressive strength of on-part geopolymers with solid activators. Their findings indicated that increasing the sodium carbonate to solid sodium silicate ratio led to microcracks in particular zones within the geopolymer, thereby enhancing the overall porosity while reducing the compressive strength14. In another work, Wang et al. focused on the effects of different CaO/Na2SiO3 ratios on the mechanical and microstructural properties of geopolymers based on neutralized copper slag and blast furnace slag in the presence of solid activators. They figured out that an increase in the CaO/Na2SiO3 ratio to up to 3:1 resulted in significant C-A-S-H gelation in the geopolymer structure, thereby optimizing the structure and pore size while enhancing the compressive strength15. Other research works have reported on the performance of geopolymer mortars based on fly ash and blast furnace slag in the presence of solid activators, durability of fly ash-based one-part geopolymer mortar, development of high-strength one-part geopolymer mortar with solid sodium metasilicate, and increasing one-part geopolymer strength with the help of alkaline activator alongside bauxite waste and silica foam16–19. Previous research works have used typical porosity measurement methods like mercury intrusion porosimetry (MIP) or pore’s specific area assessment by Brunauer-Emmett-Teller (BET) analysis. Connected porosity is indeed a measure of connected pore space to the bulk volume of a porous medium, also known as effective porosity. Important novelty of this work is the use of non-conventional activators such as different Na- and K- based alkaline salts for producing one-part slag-mullite geopolymer and investigation of connected porosity which is among the most critical types of porosity and provides a measure of connected pore space. Despite the studies reported so far, none of them has investigated a wide range of solid activators and the performance of the resulting geopolymers. Nevertheless, authors of this work have not found references on the mutual relation of connected porosity and compressive strength of the designed one-part geopolymer mortar for predicting the durability of the final product. Another objective of the present work was to use mullite as an admixture in the geopolymer mortar mix design to check for its effect on the compressive strength, water absorption, and connected porosity. The reason behind our focus on the compressive strength, in this work, is its critical role in determining the durability and the fact that geopolymer samples of higher connected porosity are expected to exhibit lower compressive strength and poorer performance in harsh and corrosive environments. Accordingly, it can be stipulated that the assessment and prediction of geopolymer mortar durability in corrosive environments is a hot topic for a wide spectrum of research in the future. Materials and methods Materials In this study, the ground granulated blast-furnace slag (GGBFS) from Isfahan iron Company, with an average Blaine of 4400 cm2/g, was used as a precursor. Type-II Portland cement from Sabz-e-Var Cement Factory with Blaine above 2800 cm2/g, was also used to build control samples. To prepare mullite, the kaolin powder from Firuzkooh Company was heated in a furnace at 1000 °C for 6 h to form the mullite. It should be noted that the average Blaine of the mullite in the in this research was determined to be 5100 cm²/g. Table 1 reports compositions of the GGBFS, mullite as admixture, and Portland cement as control. Figure 1 shows the results of the FTIR analyses for the used mullite, and GGBFS. Table 1Chemical compositions of the materials used to prepare geopolymer mortar, including GGBFS, mullite, and Portland cement (wt %, XRF).SiO2Al2O3Fe2O3CaOMgOTiO2Na2OK2OSO3P2O5LOIGGBFS35.9611.460.9838.149.421.640.480.750.24-0.00Mullite53.5742.630.460.170.170.920.0570.69---Portlandcement21.994.314.1562.882.86-0.210.473.13-- Chemical compositions of the materials used to prepare geopolymer mortar, including GGBFS, mullite, and Portland cement (wt %, XRF). Portland cement FTIR spectra for the GBBFS, mullite. As seen in Fig. 1 The slag exhibits numerous peaks, while the mullite displays few characteristic peaks, which is represent of more oriented form resulting from the calcination process. In the FTIR spectra of the synthesized mullite, the characteristic peaks observed at 472 and 1097 cm⁻¹ are attributed to the Si–O and Si–O–Al vibrational modes, respectively. Furthermore, the absorption band at 814 cm⁻¹ corresponds to the Al–O bonds. Conversely, the slag exhibits a prominent peak at 973 cm⁻¹, which signifies the presence of T–O–Si (where T = Si or Al) stretching vibrations. This specific band is of paramount importance in the FTIR characterization of geopolymer systems. Figure 2 shows that the mullite, upon heating, exhibits characteristic few peaks in its XRD pattern. The XRD patterns also demonstrate that the crystalline intensity diminished following the thermal treatment (calcination) of the kaolin precursor, suggesting a transition to an mullite phase.The XRD pattern confirms the formation of the mullite phase. The primary reflection at 2ө =26.2, indicates the well-developed crystallization of the aluminosilicate framework. Furthermore, the presence of the characteristic peak at 33.2 along with higher-angle reflections at 40.8, 60.7, and 67.1, suggests a highly ordered lattice and a substantial degree of crystallinity within the sample. X-ray diffraction patterns Mullite. Since this research focuses on producing an unconventional cement mortar (geopolymer), a specific grade of silica sand was utilized in this study. Specifically, silica sand supplied by the Firuzkooh Company was employed, featuring two distinct particle size distributions. geopolymer mortar must exhibit adequate flowability, workability and easily placed into the mould. Therefore, determining the appropriate sand-to-binder ratio is of critical importance in the production of geopolymer mortar. Findings from previous studies indicate that a sand-to-binder ratio of 2 can product a geopolymer mortar possessing the aforementioned properties; therefore, a ratio of 2:1 was selected in the present research20–22. The sand mixture used in the geopolymer mortar was made from fine particles of 0.1–0.5 mm in diameter (50%) alongside coarse particles of 0.5–2.5 mm in diameter (50%).The two size ranges were selected for the aggregates in order to ensure equal participation of each in the production of the geopolymer mortar23,24. A variety of different solid alkaline salts containing Na, K, and Ca ions were used as activators for preparing the geopolymer mortar samples, as detailed in Table 2. Table 2Types and characteristic of the used solid alkaline salts.No.Solid alkaline materialPurity (wt%)1Anhydrous sodium carbonate99.5 industrial grade2Solid sodium silicate pentahydrate99 industrial grade3Anhydrous potassium carbonate99.5 industrial grade4Potassium silicate pentahydrate98 industrial grade5Calcium hydroxide98 industrial grade6Anhydrous sodium sulfate99 industrial grade Types and characteristic of the used solid alkaline salts. Noteworthily, the silica modulus for the sodium silicate pentahydrate (SiO2 = 32.6, Na2O = 33.0, H2O = 34.4) and potassium silicate pentahydrate (SiO2 = 24.1, K2O = 37.8, H2O = 36.1) is 0.98 and 0.63 respectively. four solid activators were used to prepare geopolymer mortar samples as detailed in Table 3, where they are organized into dual groups: Table 3The binary combinations of alkaline salts utilized in this study.Geopolymer mortarmix nameAlkali saltsA1sodium carbonate and sodiumsilicate pentahydrateA2potassium carbonate and potassiumsilicate pentahydrateA3anhydrous sodium carbonateand calcium hydroxideA4anhydrous sodium sulfate andcalcium hydroxide The binary combinations of alkaline salts utilized in this study. sodium carbonate and sodium silicate pentahydrate potassium carbonate and potassium silicate pentahydrate anhydrous sodium carbonate and calcium hydroxide anhydrous sodium sulfate and calcium hydroxide As detailed in Table 3, given the profound impact of solid activators on geopolymer mortars, the nomenclature of the mixtures in the present research is established based on the binary activator categories (A1 to A4). Consequently, the mixtures are designated as Mix A1 through Mix A4. Experimental procedure Preparing the samples A variety of different solid activators were used in this study, and this necessitates various mixture designs. For the silicate- and carbonate-based activators (mixes A1 and A2), it was necessary to define ranges for the silica modulus (Si2O/M2O, M = K or Na) and Na2O or K2O content (wt% by weight of slag). Given that mortars of desired properties (e.g., high strength) require significant amounts of silica, the lowest Si2O/Na2O and Si2O/K2O ratios for the geopolymer mortars containing mixes A1 and A2 were set to 0.4 and 0.2, respectively. The highest Si2O/Na2O ratio for preparing the geopolymer mortars was set to 0.8; this was based on the silica modulus of the sodium silicate pentahydrate activator (Ms = 0.98) as the main supply of silica. When it came to the geopolymer mortars (mix A2), considering the silica modulus of the potassium silicate pentahydrate activator (Ms = 0.63), the highest Si2O/K2O ratio was set to 0.6. Regarding the Na2O content, we know that an increase in Na2O content from either of the two supplies (sodium silicate pentahydrate and sodium carbonate) increases the compressive strength of the geopolymer mortar samples, imposing no limit on the set range of this species. Therefore, the range was set to 6, 8, and 10 wt% by weight of slag. For the activators containing potassium silicate pentahydrate and potassium carbonate, the K2O content was limited to 6, 8, and 10 wt% by weight of slag following the same line of reasoning as that for mix A1. In addition to what was mentioned, fundamental and applied studies have been conducted on the effect of silica modulus, alkali concentration, and slag on alkali-activated materials and geopolymers, which has significantly helped in determining a suitable range for Na2O% and silica modulus25,47. Used as an admixture, the mullite content was set to 5 wt% by weight of slag. This particular dosage was determined to avoid the disadvantages of lower and higher dosages on the produced geopolymer mortar, given the explanations introduced in our previous work26. Table 4 presents the considered mixture designs for preparing the geopolymer mortar samples based on the mixes A1 and A2 (solid activators carbonate and silicate) at different levels of silica modulus and Na2O and K2O contents. Additionally, the workability parameters and the respective w/b ratios for all mix designs have been detailed in Table 4. Before preparing the geopolymer mortar samples, the flow table test (ASTM C1437) for measuring spread diameter was conducted to determine the required amount of water for different geopolymer mortar samples27. The spread diameter for all samples was maintained constant at 100 ± 5 mm by adjusting the w/b ratio. For the production of all geopolymer mortar samples, the dry mixing method, which is the basis for one- part geopolymers, was use28. In all cases, dry materials (ground granulated blast-furnace slag (GGBFS), solid activators, and mullite (at 5 wt%) were ground in a homogenizer for 30 s to achieve thorough mixing. Next, the coarse - and fine-grain sands were blended with the dry mixture, and water was introduced at different ratios before molding the geopolymer mortars into special molds of 2 × 2 × 2 cm and 5 × 5 × 5 cm in dimensions in stages. To optimize the use of raw materials for geopolymer production, 2 cm3 molds (sets of 4) were used to measure compressive strength, and a 5 cm3 mold was used for coring and porosity measurement. The average values of 4 compressive strengths were taken as the final compressive strength value, and the average values of 3 connected porosities were taken as the final connected porosity value. Alongside all geopolymer mortar samples, specimens of Portland type II cement were further prepared as a control. The reason for using Type II Portland cement as a control sample is its properties such as resistance to sulfate environments, compressive strength, and high durability, which make it a suitable selection for comparison with geopolymer mortar samples.The molded samples were maintained in wet conditions (relative humidity of 98% at 25 °C) for 24 h before removing the molds and exposing the samples to hydrothermal conditions (80 °C) for 24 h to complete the curing stage. Different groups of geopolymers with different solid activators were prepared in paste mode (With out sand) and subjected to the same curing steps before they were powdered and got ready for the FTIR analysis. Table 4General mixture designs for geopolymer mortar samples based on mix A1 and mix A2 activators with different solid activators.Sample noMixnameWt.% ofNa2O or K2O%Ms = SiO2/M2OM = Na or KTotal solid content (100 g)w/b ratioMullite (%)Slag (%)1SS1A160.4-1000.3552SS2A180.4-1000.3493SS3A1100.4-1000.3514SS4A160.6-1000.3505SS5A180.6-1000.3466SS6A1100.6-1000.3437SS7A160.8-1000.3428SS8A180.8-1000.3459SS9A1100.8-1000.35210SSM1A1 (mullite)60.45950.38111SSM2A1 (mullite)80.45950.37712SSM3A1 (mullite)100.45950.37313SSM4A1 (mullite)60.65950.36614SSM5A1 (mullite)80.65950.36915SSM6A1 (mullite)100.65950.36216SSM7A1 (mullite)60.85950.37417SSM8A1 (mullite)80.85950.36518SSM9A1 (mullite)100.85950.36319PP1A260.2-1000.37520PP2A280.2-1000.38621PP3A2100.2-1000.37022PP4A260.4-1000.37323PP5A280.4-1000.37824PP6A2100.4-1000.37125PP7A260.6-1000.37026PP8A280.6-1000.37527PP9A2100.6-1000.36328PPM1A2 (mullite)60.25950.37229PPM2A2 (mullite)80.25950.39830PPM3A2 (mullite)100.25950.39531PPM4A2 (mullite)60.45950.39732PPM5A2 (mullite)80.45950.39533PPM6A2 (mullite)100.45950.39234PPM7A2 (mullite)60.65950.38035PPM8A2 (mullite)80.65950.38936PPM9A2 (mullite)100.65950.391SS: Sodium silicate pentahydrate and Sodium carbonate (related to mix A1). PP: Potassium silicate pentahydrate and Potassium carbonate (related to mix A2). SSM: Sodium silicate pentahydrate and Sodium carbonate with mullite (related to mix A1 mullite). PPM: Potassium silicate pentahydrate potassium carbonate with mullite (related to mix A2 mullite). General mixture designs for geopolymer mortar samples based on mix A1 and mix A2 activators with different solid activators. SS: Sodium silicate pentahydrate and Sodium carbonate (related to mix A1). PP: Potassium silicate pentahydrate and Potassium carbonate (related to mix A2). SSM: Sodium silicate pentahydrate and Sodium carbonate with mullite (related to mix A1 mullite). PPM: Potassium silicate pentahydrate potassium carbonate with mullite (related to mix A2 mullite). The calculations for preparing the geopolymer mortars with the mix A3 and mix A4 were based on the produced percent weight of NaOH upon compositing the two activators in the presence of the slag as precursor. On the other hand, production of a geopolymer mortar with desired mechanical properties requires significant amounts of NaOH, making us set the applicable range of this material to 4 ,5, and 6 wt%. The mentioned range for the NaOH weight% has been determined based on previous studies conducted in this regard, as well as the mix designs used in them29,30. In cases where mullite was used as an additive at 5 wt%, and slag was introduced into the calculations at 95 wt%. Table 5 lists the considered mixture designs for preparing the geopolymer mortar samples based on the mix A3 and mix A4 (solid activators carbonate and sulfate) at different levels of NaOH content along with the workability values and the respective w/b ratios for each mix design. For A3 mix and A4 mix, the additive concentrations were normalized and expressed as NaOH equivalent (wt%) to provide a standardized alkalinity index for all mixtures, ensuring a consistent comparative analysis of the activation performance. Table 5General mixture designs for geopolymer mortar samples based on mix A3 and mix A4 activators with different solid activators.Sample noMixnameNaOH (wt%)Total solid content (100 g)w/b ratioMullite (wt%)Slag and Activator (wt%)1SC1A34-1000.5312SC2A35-1000.5493SC3A36-1000.5474SCM1A3 (mullite)45950.5735SCM2A3 (mullite)55950.5696SCM3A3 (mullite)65950.5627SFC1A44-1000.4978SFC2A45-1000.4929SFC3A46-1000.48910SFCM1A4(mullite)45950.51011SFCM2A4(mullite)55950.50612SFCM3A4(mullite)65950.501SC: Sodium carbonate and Calcium hydroxide (related to mix A3). SCM: Sodium carbonate and Calcium hydroxide with mullite (related to mix A3 mullite). SFC: Sodium sulphate and Calcium hydroxide (related to mix A4). SFCM: Sodium sulphate and Calcium hydroxide with mullite (related to mix A4 mullite). General mixture designs for geopolymer mortar samples based on mix A3 and mix A4 activators with different solid activators. SC: Sodium carbonate and Calcium hydroxide (related to mix A3). SCM: Sodium carbonate and Calcium hydroxide with mullite (related to mix A3 mullite). SFC: Sodium sulphate and Calcium hydroxide (related to mix A4). SFCM: Sodium sulphate and Calcium hydroxide with mullite (related to mix A4 mullite). Compressive strength, connected porosity and water absorption One of the most important tests for evaluating the performance and characteristic compressive strength of a geopolymer mortar is the compressive strength test. In this research, compressive strength of the geopolymer mortar was evaluated using a 50 KN uniaxial hydraulic machine. While 20 mm cubic specimens were employed specifically to optimize material consumption and improve dimensional precision the experimental protocol remained consistent with the general requirements of ASTM C109, utilizing a constant loading rate of 0.5 MPa/s (1 mm/min). For each mix design, the reported values represent the mean of four independent replicates to ensure the reliability of the dataset.Connected porosity of the geopolymer mortars was measured by the gas porosimetry apparatus developed by Fetco (model R20), where nitrogen gas was devised and the Boil’s law at constant temperature was applied as follows Eq. (1): where: V1 = Expansion cell volume in cm3. V2 = (Vcell + Vexpansion cell – VG). Vcell = Sample cell volume, 110 cm3. Vexpansion cell = 41.30 cm3. P1 = Expansion cell pressure in psi. P2 = Matrix cell pressure in psi. The P1 and P2 pressures are read from the test rig in the course of the experiment. Solid volume of the sample, VG, is determined based on the known constant volumes Vcell and Vexpansion cell, which are obtained from the calculations related to the porosimetry apparatus. Finally, connected porosity was calculated through the Eq. (2): The water absorption measurement process started by dehydrating the cured samples in an oven. The following procedure shall be repeated until the mass stabilizes, as per the ASTM C642 standard code31: Two successive mass measurements are Compared. If the difference exceeds 0.5% of the lesser value, the samples are returned to the oven (40 °C) for another 24-hour drying period. This process continues until the difference between any two successive measurements is less than 0.5% of the lowest recorded value. Samples were then submerged in water for 48 h, as per the ASTM C642 standard code: After final drying, cooling, and mass determination, the samples are immersed in water at room temperature. Immersion of samples is carried out for 48 h, continuing until two successive surface-dried mass measurements, taken at 24 h intervals, show an increase of less than 0.5% of the larger value. After reaching this criterion, the sample is removed, dried with a towel to eliminate surface moisture, and its mass is recorded. Finally, the water absorption percentage is calculated using Eq)3(: where: Wd = Dehydrated sample weight before submergence in g. Ww = Sample weight after 48 h of submergence in g. WA = Water absorption in percent (%). Figure 3 demonstrates the experimental procedure and related analyses for determining optimal geopolymer mortar in the form of a flowchart. General workflow of the experiments in this research. Results and discussion Compressive strength and connected porosity Compressive strength and porosity are among the most critical characteristic compressive strength for measuring the overall performance of a mortar or concrete. In the present study, both properties were measured and analyzed. Figure 4 exhibits the results of connected porosity and compressive strength measurements (28-day curing age) on the prepared samples. Plots of compressive strength and connected porosity for the mortar samples based on different mixes: (a) A1 mix, (b) mix with mullite, (c) A2 mix, and (d) A2 mix with mullite. As seen in Fig. 4a, b, c and d a generally inverse relationship can be observed between the connected porosity and compressive strength, meaning that connected porosity increases with decreasing compressive strength, and vice versa. The progress of the geopolymerization reaction and formation of consolidated and stable structure results in escalated compressive strength and, hence, decreased connected porosity. As seen in Fig. 4a, b, c and d it is evident that the highest compressive strength and the lowest connected porosity are associated with the samples SS7, SSM4, PP9, and PPM8, showing compressive strength values of 57.00, 58.12, 58.55, and 59.74 MPa and connected porosity values of 6.68%, 6.32%, 6.30%, and 6.18%, respectively. For both groups of mixes A1 and A2, the presence of mullite was seen to decrease the connected porosity while increasing the compressive strength relatively. This can be attributed to the presence of active alumina and its impact on the resulting gel from the geopolymerization process, not to mention the effect of the increase in the formation of the planar structure made from silica and oxygen in the geopolymer matrix27,32. The ranges of compressive strength and connected porosity were seen to be wider for the mix A2 (see Fig. 4c and d) compared to the mix A1 (see Fig. 4a and b). This distinction could be due to the difference in the type of alkaline ion between these two groups of activators. Indeed, despite its higher chemical activity compared to Na+ ion, the K+ ion is less soluble due to its larger ionic size, while Na+ is more effective in dissolving and hence the geopolymerization process thanks to its smaller size. This is why the samples containing Na+ could achieve narrower ranges but higher levels of compressive strength and lower levels of connected porosity compared to those containing activators with K+. On the other hand, with increasing the silica modulus to 0.8 and 0.6 in the mix A1 and A2, respectively, the K+ concentration is high enough to establish, thanks to its higher chemical activity compared to Na+, the maximum compressive strength coupled with minimum connected porosity both with and without mullite (samples PP9 and PPM8)33–36. As seen in Fig. 4, among the samples with Ms = 0.8 and Na2O = 6%, the sample SS7 exhibited the highest compressive strength and connected porosity. The presence of mullite in the abovementioned samples made the sample SSM4 (Ms = 0.6 and Na2O = 6%) show the highest compressive strength and connected porosity compared to the corresponding sample without mullite. This could be attributed to the presence of reactive species, such as alumina, the mullite could establish a sample with better characteristic compressive strength. Higher levels of Ms = 0.8 and Na2O = 10% in sample SS9, however, led to no significant increase in connected porosity or decrease in compressive strength. This indicates that increasing the dosage of sodium silicate as an activator can improve the geopolymer mortar properties up to a certain level, beyond which it can adversely affect the geopolymerization process due to the adhesive and viscous nature of the silica37. In the samples with K-containing (mix A2), no significant trend was seen in the compressive strength and connected porosity for the silica modulus up to 0.4. However, with the increase in the modulus to 0.6 at Na2O, an abrupt rise of compressive strength was evident in sample PP9. This sharp rise can be attributed to high levels of silicate alongside the highly active K+ ions in the potassium silicate activators. On the other hand, the presence of mullite in this group of activators set the scene for achieving the maximum compressive strength and connected porosity at lower levels of the potassium silicate (Ms= 0.6 and Na2O = 8%). Given the explanations presented so far, it can be concluded that, in the samples containing mixes A1 and A2, compared to the sodium carbonate and potassium carbonate, the sodium silicate and potassium silicate (thanks to their nature and silica modulus), impose greater effects on the geopolymer mortar properties. Figure 5 depicts connected porosity and compressive strength variations for the geopolymer samples prepared with the mixes A3 and A4. Plots of connected porosity and compressive strength for mortars based on (a) mix A3, (b) mix A3 with mullite, (c) mix A4, and (d) mix A4 with mullite. Referring to the samples containing mix A3, Fig. 5a and b show different trends of connected porosity and compressive strength. Figure 5a indicates an initially decreasing and then increasing trend for connected porosity, but Fig. 6b shows a generally decreasing trend for connected porosity and a continuously increasing trend for compressive strength upon introducing the mullite. The highest compressive strength (31.97 and 33.60 MPa) and the lowest connected porosity (21.21% and 20.64%) were seen for the samples SC2 and SCM3, respectively. Water absorption for the mortar samples prepared based on (a) mix A1, (b) mix A1 with mullite, (c) mix A2, and (d) mix A2 with mullite. Figure 5c and d show connected porosity and compressive strength variations for the samples prepared with mix A4, highlighting the decreasing trend of connected porosity and increasing trend of compressive strength for the samples with and without mullite. These figures suggest that the highest compressive strength (23.2 MPa and 24.5 MPa) and the lowest connected porosity (24.59% and 23.65%) were exhibited by samples SFC3 and SFCM3, respectively. In general, the reason behind the fact that the highest connected porosity and lowest compressive strength were seen for the samples with mix A4 was the presence of CaSO4 in the geopolymer structure, which is known to cause efflorescence. Upon the reaction between the mix A4 in the presence of water, the calcium sulfate is decomposed into calcium and sulfate ions. These ions then migrate to the sample surface and form a solid white residue after dehydration38. Positive effects of mullite on the connected porosity and compressive strength of the resulting geopolymer mortar are evident. This can be explained by the presence of silica and active alumina in the mullite and increased development of planar silicates in the geopolymer matrix. This effect could be better characterized and verified by means of FTIR analysis, as presented in the following sections. Note that the connected porosity and compressive strength for the Portland cement, which was herein used as control, were measured at 14.92% and 34.76 MPa, respectively. The geopolymer mortars based on the mixes A1 and A2 outperformed those based on the mixes A3 and A4, with all geopolymer mortars showing significantly better performance than the Portland cement. The superiority of the mixes A1 and A2 could be attributed to the presence of silicate compounds in these activators, which resulted in the formation of a lower connected porosity coupled with a higher compressive strength. Thanks to their lower connected porosity and higher compressive strength, the samples based on the mixes A1 and A2 are expected to show better tolerance to corrosive and harsh environments, as compared to other samples studied in this work. Proving this expected performance, however, needs further investigation in future studies. Water absorption Following connected porosity and compressive strength, water absorption is the third important parameter for evaluating the performance of a geopolymer. Figure 6 depicts water absorption for the samples prepared with the mixes A1 and A2 with mullite. As seen in Fig. 6, samples SS7, SSM4, PP9, and PPM8 showed the lowest water absorptions for the samples prepared based on mixes A1 and A2, measured at 1.55%, 1.37%, 1.23%, and 1.13%, respectively. Figure 7 reports water absorption for the samples prepared with the mixes A3 and A4 with mullite. Water absorption for the mortar samples prepared based on (a) mix A3, (b) mix A3 with mullite, (c) mix A4, and (d) mix A4 with mullite. As seen in Fig. 7, the samples prepared based on the mixes A3 and A4 exhibit high levels of water absorption. Accordingly, samples SC2, SCM3, SFC3, and SFCM3 showed water absorption levels of 7.53%, 7.18%, 8.64%, and 8.32%, respectively. These results are in agreement with those reported in previous sections on the connected porosity and compressive strength of the samples, and prove the very high performance of the samples in terms of water absorption; i.e., the samples with the lowest connected porosity and highest compressive strength exhibit the lowest water absorption. The water absorption for the Portland cement control mix was recorded at 8.5%. In comparison to this baseline, mixes A1 and A2 exhibited significantly lower water absorption, indicating superior performance. Conversely, a relative decline in performance was observed for mixes A3 and A4; specifically, A4 surpassed the control specimen’s water absorption level. These findings underscore the pivotal role of the alkali activator type in dictating water absorption properties, an important parameter for geopolymer mortar mixes. Although the geopolymer mortar produced in the present study is of non-conventional type, the reported water absorption percentages show some consistency with studies on the effect of absorption on concrete properties39,40. FTIR analysis FTIR spectroscopy was performed for a detailed study of molecular structure and chemical bonds forming the geopolymer matrix. For each group of activators, the samples with the highest compressive strength and lowest connected porosity and water absorption were subjected to the FTIR analysis. Figure 8 shows the results of the FTIR analysis for four samples of geopolymer matrix prepared based on mixes A1 and A2. FTIR spectra for the geopolymer paste samples SS7, SSM4, PP9, and PPM8. Given the spectra shown in Fig. 8, the peaks appearing in the range of 875–881 cm− 1 are associated with the Al-O bond in the AlO4. The presence of mullite in samples SSM4 and PPM8 shifts the peaks toward 877 cm− 1 while reducing their intensity. This suggests that the mullite can optimize the orientation and activity of the reactants in the reaction medium, resulting in a more stable and stronger yet less porous geopolymer of lower water absorption. The presence of mullite has further moved the characteristic compressive strength peak of the geopolymers (within 1000 cm− 1) toward lower wavenumbers, as seen for the samples PPM8 and SSM4 (at 977 and 975 cm− 1, respectively). This has favored the formation of silicate structure along with aluminosilicate network in the primary gel of the geopolymer structure, thereby making the final sample more compact, less porous, and stronger than the mullite-free samples thanks to the planar silicate structure41–49. To sum up, all of these findings confirm the positive effect of mullite on the structure and, hence, connected porosity, compressive strength, and water absorption of the geopolymer, as explained in previous sections. Figure 9 demonstrates the FTIR spectra for the geopolymer paste samples prepared based on mixes A3 and A4. It highlights the presence of calcium sulfate (gypsum), which has led to the efflorescence and the resulting presence of S–O bonds in the samples containing mix A4, as characterized by the strong peaks at 1122–1124 cm− 1 for the samples SFC3 and SFCM3. A relatively strong peak is also seen in the range of 617–623 cm− 1, which can be attributed to the tensile vibrations of the SO42− group due to the presence of sodium sulfate, as activator, or calcium sulfate in samples SFC3 and SFCM3. FTIR spectra for the geopolymer paste samples SC2, SCM3, SFC3, and SFCM3. As explained in a previous section, the presence of calcium sulfate in the samples prepared with mix A4 contributes to reducing compressive strength while increasing connected porosity and water absorption compared to other samples based on different groups of activators. The presence of mullite in sample SFCM3 partially limited the efflorescence and, hence, calcium sulfate formation, as indicated by reduced intensity of the characteristic peaks for this species, although even this reduced intensity is enough to mask the characteristic peak for mullite in the spectrum for SFCM3. For the samples prepared based on mix A3, the presence of mullite resulted in a shift of the characteristic peaks for alumina and Al–O bond from 875 cm− 1 in SC2 to 877 cm− 1 in SCM3, not to mention the increase in the peak intensity. Therefore, the increase in compressive strength coupled with the decrease in connected porosity and water absorption can be explained by the further tendency toward forming silicate network with its planar structure and its positive effect on the mullite-modified geopolymer composition15,50. Relationship between compressive strength and connected porosity In order to investigate the relationship between compressive strength and connected porosity in more detail, one needs to plot these two variables by taking them as an independent–dependent pair of factors for different values of Ms. Table 6 presents the all plots along with their corresponding fitted trend line and regression coefficient for compressive strength and connected porosity at different levels of Ms and Na2O or K2O content (6, 8, and 10 wt%) and different mix groups. Table 6Regression coefficient and trend line equation for compressive strength and connected porosity.No.Ms = SiO2/ M2OM = Na or KNa2O or K2O (wt%)Mortar samples R 2 Trend line1.0.46SS10.999y = -4.2454 x + 85.0720.48SS20.410SS32.0.66SS40.9963y = -3.3639 x + 78.9040.68SS50.610SS63.0.86SS70.9993y = -4.5297 x + 87.3380.88SS80.810SS94.0.46SSM10.9998y = -4.9907 x + 91.200.48SSM20.410SSM35.0.66SSM40.9909y = -4.2383 x + 84.930.68SSM50.610SSM66.0.86SSM70.9976y = -4.2124 x + 84.900.88SSM80.810SSM97.0.26PP10.9974y = -4.6101 + 107.580.28PP20.210PP38.0.46PP40.9977y = -6.2258 x + 130.380.48PP50.410PP69.0.66PP70.9481y = -2.1199 x + 72.590.68PP80.610PP910.0.26PPM10.9573y = -2.4196 x + 76.6040.28PPM20.210PPM311.0.46PPM40.9753y = -1.886 x + 71.8940.48PPM50.410PPM612.0.66PPM70.9981y = -2.0923 x + 72.580.68PPM80.610PPM9 Regression coefficient and trend line equation for compressive strength and connected porosity. As seen in Table 6, all trend lines showed negative slopes, proving the inverse relationship between compressive strength and connected porosity. For the samples prepared based on mix A1 (sodium carbonate and sodium silicate pentahydrate with and without mullite), the regression coefficient (R2) was as high as 0.99, indicating the strong association of connected porosity with compressive strength. This very high correlation for the geopolymer mortar samples designated as mix A1 can be attributed to the good performance of the sodium ions in the mix A1 when it comes to dissolution and formation of the geopolymer matrix. On the other hand, in the samples prepared based on mix A2 (potassium carbonate and potassium silicate pentahydrate), the R2 value was below 0.99 for particular levels of Ms. This can be a result of differences in the behavior of the potassium ion in the activator at different levels of Ms, not to mention its lower solubility than the sodium ion and differences in the formation of the geopolymer matrix. Differences in the slope of the trend line for mixes A1 and A2 with mullite samples can be linked to the differences in the potassium ion behavior, which had led to the formation of samples with different connected porosity and compressive strength values. Table 7 reports the regression coefficient and trend line equation for the samples prepared based on mixes A3 and A4. Table 7Regression coefficients and trend line equations for the relationship between compressive strength and connected porosity at different wt% of NaOH.No.wt% of NaOHMortar sample R 2 Trend line14SC10.9944y = -1.7026 x + 68.13725SC236SC344SCM10.9828y = -2.213 x + 79.06555SCM266SCM374SFC10.9949y = -1.1684 x + 51.89485SFC296SFC3104SFCM10.9861y = -1.2327 x + 54.629115SFCM2126SFCM3 Regression coefficients and trend line equations for the relationship between compressive strength and connected porosity at different wt% of NaOH. Given Table 7, it is evident that the negative slope further confirms the inverse nature of the relationship between compressive strength and connected porosity. Furthermore, in all cases, the R2 was found to be close to 0.99, which indicates a strong relationship between compressive strength and connected porosity, similar to the case with mixes A1 and A2. For both mixes A3 and A4, the presence of mullite increased the slope of the trend line over that of the cases without mullite, because the mullite-containing samples exhibited lower connected porosity and higher compressive strength levels, with the slope of the plot being affected by this concept. Conclusion The following conclusions were drawn from this study: Due to the presence of silicate species in the Na2SiO3·5H2O and K2SiO3·5H2O related samples, the lowest connected porosity and highest compressive strength have been exhibited (mixes A1 and A2). Indeed, the silicate species and the Si–O bonds established compact and stable geopolymer structures. In contrast, the mix prepared based on the A3 and, especially, mix A4 showed the poorest compressive strength coupled with the highest connected porosity, as compared to other samples as well as Portland cement as control. Results of FTIR analysis on the geopolymer matrix confirmed the presence of gypsum (CaSO4) in the composition of the samples with the mix A4 (Ca(OH)2 and Na2SO4), with the gypsum being a known cause of efflorescence and, hence, increased connected porosity and reduced compressive strength of the samples. Due to obtained results, connected porosity could be mentioned as one of the important properties in evaluating the performance of geopolymer mortar in determination of durability. Connected porosity provides a measure of the ability of a porous media to transmit or store fluids within its structure. This concept plays a key role in the prediction of durability and strength of a geopolymer matrix upon exposure to corrosive and harsh solutions. That is, the higher the connected porosity in a geopolymer mortar, the more it is susceptible to degradation by the permeation of aggressively corrosive environment. Due to the presence of silicate species in the Na2SiO3·5H2O and K2SiO3·5H2O related samples, the lowest connected porosity and highest compressive strength have been exhibited (mixes A1 and A2). Indeed, the silicate species and the Si–O bonds established compact and stable geopolymer structures. In contrast, the mix prepared based on the A3 and, especially, mix A4 showed the poorest compressive strength coupled with the highest connected porosity, as compared to other samples as well as Portland cement as control. Results of FTIR analysis on the geopolymer matrix confirmed the presence of gypsum (CaSO4) in the composition of the samples with the mix A4 (Ca(OH)2 and Na2SO4), with the gypsum being a known cause of efflorescence and, hence, increased connected porosity and reduced compressive strength of the samples. Due to obtained results, connected porosity could be mentioned as one of the important properties in evaluating the performance of geopolymer mortar in determination of durability. Connected porosity provides a measure of the ability of a porous media to transmit or store fluids within its structure. This concept plays a key role in the prediction of durability and strength of a geopolymer matrix upon exposure to corrosive and harsh solutions. That is, the higher the connected porosity in a geopolymer mortar, the more it is susceptible to degradation by the permeation of aggressively corrosive environment. For all of the studied mortar samples, measured water absorption was less than connected porosity, which could be linked to the nature of the test fluid (water + gaseous nitrogen). In fact, due to the high diffusivity of the gas compared to water and the pressurized condition of the samples in the porosimetry test, one would expect a higher connected porosity than the water absorption. Results showed that the samples exhibiting the lowest connected porosity and highest compressive strength enjoyed a more compact structure and, hence, showed lower water absorption. Overall, the highest and lowest water absorption levels were obtained for the samples SFCM3 and PPM8, respectively. On the other hand, the samples prepared based on the mix A4 exhibited higher water absorption than the Portland cement because of the presence of CaSO4 (gypsum) and the resulting efflorescence, as confirmed by the outputs of the FTIR analysis. For all of the studied mortar samples, measured water absorption was less than connected porosity, which could be linked to the nature of the test fluid (water + gaseous nitrogen). In fact, due to the high diffusivity of the gas compared to water and the pressurized condition of the samples in the porosimetry test, one would expect a higher connected porosity than the water absorption. Results showed that the samples exhibiting the lowest connected porosity and highest compressive strength enjoyed a more compact structure and, hence, showed lower water absorption. Overall, the highest and lowest water absorption levels were obtained for the samples SFCM3 and PPM8, respectively. On the other hand, the samples prepared based on the mix A4 exhibited higher water absorption than the Portland cement because of the presence of CaSO4 (gypsum) and the resulting efflorescence, as confirmed by the outputs of the FTIR analysis. Across the investigated activator groups, the addition of mullite predominantly contributed to reduced connected porosity and enhanced compressive strength. However, the extent of these improvements was highly sensitive to the activator type, and in certain mix designs, localized deviations such as increased porosity were observed due to workability constraints and matrix-filler interaction dynamics. Where the reinforcing effect was dominant, it could be attributed to the presence of reactive silica and alumina within the mullite. The analytical results suggest that the release of alumina and the formation of robust Al–O bonds during geopolymerization promote a denser geopolymer structure. This process further encourages the formation of more densified aluminosilicae network with planar silicates within the matrix, thereby enhancing mechanical performance while reducing connected porosity and water absorption. Across the investigated activator groups, the addition of mullite predominantly contributed to reduced connected porosity and enhanced compressive strength. However, the extent of these improvements was highly sensitive to the activator type, and in certain mix designs, localized deviations such as increased porosity were observed due to workability constraints and matrix-filler interaction dynamics. Where the reinforcing effect was dominant, it could be attributed to the presence of reactive silica and alumina within the mullite. The analytical results suggest that the release of alumina and the formation of robust Al–O bonds during geopolymerization promote a denser geopolymer structure. This process further encourages the formation of more densified aluminosilicae network with planar silicates within the matrix, thereby enhancing mechanical performance while reducing connected porosity and water absorption. Considering the trend line equations for the studied samples, an inverse relationship was identified between connected porosity and compressive strength. On the other hand, the obtained values of the regression coefficient (R2) differed for different types of activators. That is, the samples prepared based on mix A2 exhibited a lower R2 than those based on mix A1. Differences in properties between the potassium and sodium ions of different activators can explain this. Due to its lower solubility (because of its larger particle size) and higher chemical activity, the potassium ion generates different levels of connected porosity and compressive strength over wider ranges. In contrast, the smaller yet more effective and soluble sodium ion produces better correlated values of connected porosity and compressive strength over narrower ranges. These differences can explain the lower value of R2 for the activators containing potassium ions rather than sodium ions. When it came to the mixes A3 and A4, the high correlation of compressive strength and connected porosity factors results in R2 values as high as 0.99. Considering the trend line equations for the studied samples, an inverse relationship was identified between connected porosity and compressive strength. On the other hand, the obtained values of the regression coefficient (R2) differed for different types of activators. That is, the samples prepared based on mix A2 exhibited a lower R2 than those based on mix A1. Differences in properties between the potassium and sodium ions of different activators can explain this. Due to its lower solubility (because of its larger particle size) and higher chemical activity, the potassium ion generates different levels of connected porosity and compressive strength over wider ranges. In contrast, the smaller yet more effective and soluble sodium ion produces better correlated values of connected porosity and compressive strength over narrower ranges. These differences can explain the lower value of R2 for the activators containing potassium ions rather than sodium ions. When it came to the mixes A3 and A4, the high correlation of compressive strength and connected porosity factors results in R2 values as high as 0.99. Author contributions Shahryar Kashi and Ebrahim Najafi Kani have contribution in writing and editing. Shahryar Kashi has contributed in investigations, formal analysis and visualization. Ebrahim Najafi Kani has contributed as supervision, methodology, visualization, and project administration. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data availability All data generated or analyzed during this study are included in this published article. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement All data generated or analyzed during this study are included in this published article.