Introduction

Concerns about the climate crisis affecting the environment have been the focus of global environmental politics. The key to mitigating worldwide climate-altering factors, such as CO₂ emissions into the atmosphere, lies in the concerted efforts to promote new ‘environmentally friendly’ technologies and facilitate their industrial and commercial deployment. The Portland cement industry is known for the substantial greenhouse gas production from the thermo-chemical clinkerization process of the raw materials approaching to 1300 ºC, which corresponds to approximately 5-10٪ of the global anthropogenic CO₂ emissions (Gao et al., 2021; Habert; Lacaillerie; Roussel, 2011; Van Deventer et al., 2010). Hence, the partial replacement of Portland cement by supplementary cementitious materials into concrete production, or total transition to alternative binder technologies with lower CO₂ emissions is crucial to improve environmental preservation.

With this purpose, metakaolin can be used as pozzolanic material in Portland cement concrete, offering several advantages, including reduction of cement consumption, lowered hydration heat, microstructure refinement, and enhanced mechanical strength (Menéndez et al., 2020). The material is derived through the calcination process of kaolinite clay minerals within a temperature range spanning from 550 °C to 950 °C temperature range – lower than that typically employed in clinker production – ensuring the amorphousness characteristics within the atomic structure (Kong; Sanjayan; Sagoe-Crentsil, 2007; Wan et al., 2017). The metakaolin’s abundant in silica and aluminum chemical composition, along with its specific production conditions, render it a valuable supplementary cementitious material for use in concrete production (Wei et al., 2019). From the concrete durability perspective, it is widely acknowledged that the application of metakaolin is subject to extensive research aimed at evaluating its potential for mitigating degradation mechanisms, such as alkali-aggregate reactions (Wei et al., 2019; Zhou et al., 2015).

The interaction between non-crystalline silica-based phases present in mineral aggregates and the alkalis within the binder (Na⁺, K⁺, and Ca₂⁺) results in a hygroscopic gel in concrete pore solution (Lindgård et al., 2012; Menéndez et al., 2020). This phenomenon initiates with the dissolution of amorphous silica, due the attack of hydroxyls (OH^-) on the siloxane bridges between silicon and oxygen atoms (Si-O-Si), thereby generating the silanol group (Si-O-H) (Kim; Olek, 2014b). The subsequent reaction releases siliceous ions (HSiO₃^2- and H₂SiO₃^-) due to the hydroxyl attack on the newly formed silanol group. These siliceous anions are then ionically linked to the alkali cations present in the microstructure, forming the gel that expands due to the humidity of the pore-solution (Guo et al., 2019).

Given the low crystallinity of silica-based pozzolanic additions, these materials undergo a reaction with Portlandite (Ca(OH)₂) present in the concrete pore solution. This reaction yields calcium silicate hydrate (C-S-H), thereby enhancing mechanical strength and matrix densification (Menéndez et al., 2021; Sun et al., 2020). It contributes to the mitigation of alkali-aggregate reactions by reducing water permeability in concrete, decreasing total alkalis through cement reduction and reducing alkali consumption via pozzolanic effects (Souza; Dal Molin, 2005). However, metakaolin is not solely comprised of amorphous silica. Its composition also exhibits significant quantities of aluminum, accounting for up to 40% of its mass composition (Kong; Sanjayan; Sagoe-Crentsil, 2007; Kumar; Revathi, 2016; Zhang; MacKenzie; Bronwn, 2009). These high amounts of aluminum contribute to the formation of additional products within the microstructure of Portland cement concrete, including calcium aluminosilicate hydrate (C-A-S-H) (Wei et al., 2019).

Certain reaction byproducts, referred to as geopolymers, may also arise within hydrated Portland cement matrices influenced by alkali-aggregate reactions. Geopolymeric materials represent a category of alkali-activated matrices with diminished calcium content, achieved through the alkaline activation of amorphous aluminosilicates, with metakaolin being a principal raw material in their production (Castillo et al., 2022; Davidovits, 1991; Deventer; Provis; Duxson, 2012; Khale; Chaudhary, 2007; Li; Sun; Li, 2010; Singh; Middendorf, 2020). Activator solutions necessitate high alkalinity and are typically derived from alkaline hydroxides and silicates. Sodium and potassium are the alkalis commonly employed in their production (Arnoult et al., 2019; Gharzouni et al., 2015). They ensure the metakaolin geopolymerization reaction, which unfolds in three sequential stages: dissolution of the precursor material, gel formation and nucleation, and then condensation of the gel. This process yields a cementitious composite characterized by a three-dimensional polymeric atomic structure (Król et al., 2019; Xu; Deventer, 2003).

The dissolution mechanism observed in both geopolymer and alkali-aggregate reaction (AAR) reactive aggregates is analogous. In alkaline solutions, metakaolin particles dissolve as a result of the medium’s pH, liberating monomeric units of silica (Si(OH)₄) and alumina (Al(OH)₄^-) in tetrahedral coordination (Singh et al., 2015; Singh; Middendorf, 2020). The T-OH (where T = Si or Al) ramifications undergo nearly instantaneous reaction with each other in a condensation process, liberating water molecules and giving rise to a three-dimensional polymeric structure (Gharzouni et al., 2015; Zhang; MacKenzie; Brown, 2009). Because of the valency of tetrahedral aluminum molecules, the alkalis present in the activating solution become integrated into the structure through ionic bonding (Duxson et al., 2006; Zhang et al., 2017).

After consolidation, the geopolymeric product has properties similar to hydrated Portland cement, being a promising option to take its place in the civil construction industry (Duxson et al., 2007; Khale; Chaudhary, 2007; Sarka; Dana; Das, 2015). It is also known that geopolymers with low calcium content are not prone to alkali-aggregate reactions, despite possessing alkalinity levels comparable to those found in cementitious matrices (Cyr; Pouhet, 2015; Fernández-Jiménez; Palomo, 2009; Pouhet; Cyr, 2014).

The absence of calcium alkalis (Ca₂⁺) in the chemical composition and pore solution of geopolymers may account for the lack of development of alkali-aggregate reactions (AAR). Numerous studies support the notion that calcium ions play a crucial role in the formation of the hygroscopic gel associated with AAR (Glasser; Kataoka, 1982; Guo; Dai; Si, 2019; Kim; Olek; Jeong, 2015; Leemann et al., 2011). Nevertheless, the dissolution of the amorphous mineralogical phases in aggregates persists in the alkaline medium, with their dissolution attributed not to calcium (Ca₂⁺), but rather to hydroxyl (OH^-), sodium (Na⁺), and potassium (K⁺) ions (Guo et al., 2019; Strack et al., 2020; Visser, 2018).

The dissolution of reactive aggregates releases silica-based phases, similar to the metakaolin process, suggesting that this silica can be assimilated into the geopolymer matrix (Singh et al., 2015; Singh; Middendorf, 2020). Similarly, substituting Portland cement with metakaolin in mortars and concretes may lead to the formation of geopolymeric products within the cementitious matrix due to hydroxyls and monovalent alkalis present in the pore solution (Wei et al., 2019).

Hence, the primary focus of this article was to examine the utilization of metakaolin for the production of geopolymeric and Portland cement mortars, particularly when exposed to alkaline solutions in the presence of aggregates susceptible to alkali-aggregate reaction (AAR). Given the resemblance between the alkaline dissolution processes of metakaolin and reactive aggregates, the study aimed to analyze the physical and mechanical properties of the resultant products to elucidate the underlying phenomena. Specifically, the investigation sought to ascertain whether metakaolin, when employed as a partial replacement for Portland cement, instigates the geopolymerization reaction in mortars. Furthermore, the study assessed the alterations in the microstructure of the matrices and evaluated the contribution of mineral phases susceptible to AAR to the consolidation of the geopolymeric product.

Experimental details

Materials characterization

Metakaolin geopolymeric mortars and Portland cement mortars with metakaolin addition were produced. Two crushing aggregates were used: one was susceptible to alkali dissolution into the pore-solution of the mortar, and the other one was non-reactive. The rheological properties of the fresh-state mortar were studied by the flow table tests to determine consistency (ABNT, 2016). These samples were also characterized by axial compressive strength, water absorption by immersion, and linear dimensional variation to the alkali-aggregate reaction (ASTM, 2021; ABNT, 2005, 2019). Microstructural characterization was also performed to understand the effects of metakaolin and deleterious aggregates on the cementitious matrices.

Figure 1, Table 1, and Figure 2 present the physical, chemical and mineralogical properties of the aggregates, respectively (ABNT, 2000, 2003, 2021b, 2021a, 2022). The X-ray fluorescence test was performed on SHIMADZU EDX 700 equipment. X-ray diffractogram tests were obtained from pressed powder tablets in a PANalytical EMPYREAN diffractometer. The measurement was performed between 5º and 75º in angular scanning 2θ, with a 0.02º 2θ step and 1s per step. A copper anode tube, with 40 kV/30 mA and a divergent cleft of 1º was used. The mineral chemical phases were identified by comparison with the ICDD (International Centre for Diffraction Data) standards and the intensity of the characteristic plagioclase peak for aggregate A (maximum at 1346 counts) and quartz for aggregate B (maximum at 1546 counts) where normalized to improve their representation.

The source rock of aggregate A (Figure 3) is a syenogranite with a massive structure and medium equigranular phaneritic texture (crystals between 1mm and 5mm). Pervasive hydrothermal alteration was also observed, which can generate secondary minerals, such as clay minerals and carbonates. While the source rock of aggregate B is an epidote-muscovite-plagioclase-alkali feldspar-biotite-quartz schist with a foliated structure and very fine lepidoblastic texture (crystals smaller than 1mm). The biotite and muscovite crystals are preferentially oriented, developing a spaced foliation/schistosity, and the feldspar crystals are incipiently clayified and sericitized, giving the crystals a cloudy appearance. These aggregates have larger quantities of powder fractions than quartz sands due to the crushing processes the rocks were submitted. Both aggregates are mainly rich in silicon and aluminum. This is coherent with their mineralogy, which presents quartz and feldspars, besides phyllosilicates present in aggregate B. All these mineralogical phases may be susceptible to the alkali-aggregate type reaction, where the main factor for its alkali-dissolution is their respective amorphicity (Leemann; Holzer, 2005).

It was used a CPII-F 32 Portland cement which is close to CEM II/A-L 32.5 N and CEM II/B-L 32.5 N types of cement of European standards according to BS EN 197-1 (ABNT, 2018; BSI, 2011). As an addition/precursor, a highly reactive metakaolin was used (ABNT, 2010, 2014). The X-ray fluorescence chemical characterization of the cementitious materials is shown in Table 2. Figure 4 demonstrates the XRD plot of the used metakaolin, where besides quartz (maximum at 150 counts) and kaolinite crystal phases an amorphous phase characterized by a bandwidth between 20º – 30º was also identified.

Mix proportion

The Si/Al/Na molar ratio of the geopolymer pastes was calculated from metakaolin’s chemical composition data (Table 3). Sodium hydroxide solutions were produced at 8 mol/l and 12 mol/l concentrations. Together with metakaolin and aggregate fractions, it resulted in geopolymer mortars. The metakaolin to aggregate ratio was 1:2.25, the same as that present at C1260 – Standard Test Method for Potential Alkali Reactivity of Aggregates (Mortar-Bar Test) (ASTM, 2021).

The mix proportion 1:2.25:0.47 (binder: aggregate: water) was used to produce the Portland cement mortars. To produce these samples, part of the cement binder was replaced by metakaolin (10% and 20%). The fresh state of the mortars was then evaluated through its consistency index (ABNT, 2016).

After the first 24 hours of mixing, the samples were kept in 1 mol/l sodium hydroxide solution at 80 ºC for 28 days. This curing time allowed the AAR catalysis and the verification of the effects of metakaolin in Portland cement mortars. The geopolymeric mortars were also kept in the same NaOH solutions to verify their susceptibility to the alkali-aggregate reaction. The amount of each material used in the mortar mixture is shown in Table 4 and 5.

Products characterization

Compressive strength, water absorption, and linear dimensional variation tests were performed. Microscopy images were obtained using a FEG-SEM, Tescan, Mira 3 microscope. The SEM also had SE/BSE detectors and an Oxford X-Maxn 50 (EDS) analytical X-ray microprobe. For this test, the specimens were reduced and superficially metalized with gold, and their fracture surface was used for analysis.

Results and discussion

The rheological characteristics of the mortars were examined by assessing the consistency index of the mixtures in their fresh state (Figure 5). Aggregate B demonstrated a greater fluidity in the mixtures compared to aggregate A, attributed to the higher water absorption levels of the latter (Table 1). Furthermore, it was observed that higher proportions of metakaolin addition in the Portland cement/metakaolin mortars resulted in reduced workable consistency. This phenomenon can primarily be attributed to the fineness and shape of the metakaolin particles (Vance et al., 2013).

The diameter of metakaolin particles is usually smaller than that of Portland cement particles due to normative requirements (ABNT, 2010), leading to a higher specific surface area for metakaolin addition compared to the binder (Bumanis et al., 2020). The higher water adsorption of metakaolin particles compared to Portland cement particles, stemming from this variance in particle size distribution, contributes to a less fluid consistency in the mixed binder mortars (Bai; Gailius, 2009; Li; Fan, 2022; Vance et al., 2013).

The geopolymer mortars were formulated with molar ratios conducive to the formation of crystalline products (Table 3), which facilitates microstructural characterization, albeit resulting in lower mechanical strength compared to Portland cement mortars (Sarkar; Dana; Das, 2015; Zhang et al., 2017). Consequently, the liquid-to-solid ratio of these mixtures was augmented until their consistency resembled that of Portland cement mortars (Table 4). However, this adjustment procedure led to further reductions in mechanical strength.

It is further noted that geopolymeric mortars produced with 12 mol/L solutions exhibited a marginally more fluid consistency in comparison to those prepared with 8 mol/L solutions. This phenomenon can be attributed to the heightened alkalinity of the activating solution, thereby diminishing the viscosity of the gel (Abdelrahman; Garg, 2022; Hou; Li; Lu, 2019). The reduction in the Al/Na molar ratio is consequently accountable for altering the rheological properties of geopolymer mortars, as illustrated in Figure 4. Although the variance in consistency may not be notably significant, the distinct molarity ratios engender changes in properties, matrix composition, and microstructure, as elaborated below.

For the assessment of the alkali-aggregate reaction, the linear dimensional variations of prismatic bars were analyzed. Notably, in samples featuring aggregate A, dimensional variations exceeding 0.1% were not observed in either geopolymer mortars or Portland cement/metakaolin mortars (Figure 6a). This finding indicates that aggregate A is inert, aligning with the specifications outlined in C1260 (ASTM, 2021). Nonetheless, it remained evident that the REF mortars exhibited the most pronounced expansions among the various samples, whereas the substitution of Portland cement with metakaolin resulted in expansions smaller than 0.05%. Geopolymeric mortars, on the other hand, displayed the lowest linear expansions, failing to exceed 0.02% of the total expansions.

Similar trends were observed in mortars containing aggregate B, with the exception that the REF samples exhibited expansions exceeding 0.2% - indicative of deleterious aggregate as per C1260 (ASTM, 2021). However, gradual reductions in these expansions were observed with increasing binder replacement by metakaolin addition, as depicted in Figure 6b. The incorporation of metakaolin in these mortars resulted in maximum expansion values of 0.1% for samples with 20% addition replacement, directly influencing the formation of alkali-aggregate reaction products. Moreover, the mixtures containing geopolymers exhibited the lowest linear expansions. These findings suggest that the dissolution of amorphous mineral phases into their microstructure does not contribute to degradation in alkali-activated matrices.

The decrease in linear expansions observed in Portland cement/metakaolin mortars is in line with findings reported in existing literature (Menéndez et al., 2021; Wei et al., 2019). Due to its pozzolanic potential, metakaolin particles engage in a reaction with portlandite present in the pore solution. This interaction yields calcium silicate hydrate, a phenomenon that mitigates the alkali-aggregate reaction in Portland cement concretes (Sun et al., 2020). Nevertheless, the mechanisms underlying the pozzolanic effect do not account for the aluminum content in metakaolin and the contributions of aluminate products within the hydrated matrix of Portland cement. Ramlochan, Thomas, and Gruber (Ramlochan; Thomas; Gruber, 2000) demonstrated that the hydrated product obtained through the incorporation of metakaolin in Portland cement composites can integrate sodium or potassium alkalis from the pore solution. Wei et al. (2019) similarly affirmed that metakaolin, when employed as a partial substitute for Portland cement in concretes, undergoes alkaline activation upon curing in sodium hydroxide solutions. These statements corroborate the fact that metakaolin is not only capable of producing calcium silicate hydrate but also geopolymers within the concrete microstructure under alkaline curing conditions.

Following alkaline activation, the aluminum content within metakaolin undergoes a change in its coordination number from 5 or 6 to 4. Result to this transformation, the monomeric Al unit linked to four OH^- ions become negatively charged (Al(OH)₄^-), which interacts by ionic binding with monovalent alkalis present in the solution (Duxson et al., 2006; Zhang et al., 2017). If such a phenomenon were to take place in the pore solution of hydrated Portland cement, alkali-activated metakaolin would probably generate geopolymers, thereby reducing the overall concentration of sodium alkalis in the medium. Since geopolymers are not susceptible to alkali-aggregate reaction (Cyr; Pouhet, 2015; Fernández-Jiménez; Palomo, 2009; Pouhet; Cyr, 2014), a mitigating mechanism against the degradation of composite mortars would ensue, akin to the pozzolanic effect.

The aggregates utilized also could serve as sources of aluminosilicates, and if they contain reactive phases, they can indeed be assimilated into geopolymer matrices (Drolet; Duchesne; Fournier, 2017; Guo et al., 2019; Khale; Chaudhary, 2007; Kim; Olek, 2014a; Król et al., 2019; Singh et al., 2015; Strack et al., 2020; Zhang; MacKenzie; Brown, 2009). This phenomenon arises from their dissolution mechanism, as the aggregates involved in alkali-aggregate reactions (AAR) exhibit similarities to metakaolin in geopolymers. In both dissolution processes, solubilized silica and aluminum are released. This hypothesis gains further support when considering the influence of aggregates on the mechanical strength of the mortar, where deleterious aggregate B contributes to enhanced geopolymer strengths. Data on compressive strength and water absorption through immersion are depicted in Figure 7.

It is underscored that the Si/Al molar ratio approaches 1, facilitating the formation of crystalline products without prioritizing high mechanical strength. Moreover, the geopolymeric mixtures fabricated with an 8 mol/L solution of sodium hydroxide exhibit an Al/Na molar ratio of 1.62, whereas mixtures utilizing a 12 mol/L solution possess an Al/Na ratio of 1.08 (as detailed in Table 3). The limited presence of sodium alkalis in the 8 mol/L samples impacts alkaline activation, thereby contributing to diminished mechanical strength due to the incomplete satisfaction of chemical bonds in tetrahedral aluminum after the coordination change during the geopolymerization process (Hou; Li; Lu, 2019).

The REF samples demonstrate that aggregate A yields higher compressive strength and water absorption compared to mortars containing aggregate B. Upon substituting Portland cement with metakaolin in mortars, samples incorporating aggregate B start exhibiting higher compressive strength alongside reduced water absorption. This observation suggests that reactive aggregates influence the products obtained through metakaolin utilization in these mixtures, thereby altering their properties.

Regarding geopolymeric mortars, their compressive strengths are indeed lower, and water absorptions are higher compared to those of Portland cement/metakaolin mortars, a phenomenon justified by the molar ratio adopted for their production (Table 3). Nevertheless, it is still notable that samples containing aggregate B exhibit higher mechanical strength than those containing aggregate A. This behavior mirrors that observed in Portland cement/metakaolin mortars, where the addition of metakaolin enhanced mechanical strength in specimens with aggregate B.

The obtained results can suggest the potential for solubilized aggregates to have condensed into the metakaolin geopolymers structure. It is recognized that geopolymers typically demonstrate higher mechanical strength as the Si/Al ratio approaches approximately 2 (Sarkar; Dana; Das, 2015; Zhang et al., 2017). It is also acknowledged that certain mineralogical phases based on silicon/aluminum can undergo polycondensation reactions, resulting in the formation of geopolymer chains (Arnoult et al., 2019; Wan; Rao; Song, 2017; Xu; Deventer, 2003). Given that aggregates A and B primarily consist of silica (Table 2), their dissolution in the alkaline medium would contribute to an elevation in the Si/Al ratio within the geopolymer. This, in turn, could lead to improvements in mechanical strength in the resultant product.

To scrutinize the microstructural morphology of the products acquired within the mortar’s microstructure, scanning electron microscopy (SEM) assays were conducted, and the resulting images are depicted in Figure 8 through Figure 13. This technique proved instrumental in delineating the influence of materials’ reactions in the matrices, thereby facilitating the characterization of the formed products.

The Portland cement mortars employed as reference mixtures (Figure 8) exhibit uniform pores in samples containing aggregate A, and they do not display any discernible AAR gel. Conversely, in specimens containing aggregate B, the characteristic gel associated with alkali-aggregate reaction was identified (Kim; Olek; Jeong, 2015; Leemann; Holzer, 2005). This outcome was anticipated based on the linear expansions observed in the mortar bars over 28 days in the sodium hydroxide solution (Figure 6a and 6b).

Figure 9 depicts the microstructure of Portland cement mortars with 10% metakaolin. Both samples differ from their respective reference samples, with the products found in the pores attributed to alkaline activation products. Specifically, in Figure 9a, materials exhibiting a slight degree of crystallinity are observed, and categorized as chabazite-type zeolites. These materials are obtained from the activation of alkaline aluminosilicates (Hasegawa et al., 2010; Li et al., 2011). The products observed within the pores in Figure 9b are predominantly amorphous, bearing a resemblance to non-crystalline geopolymeric phases (Król et al., 2019; Sarkar; Dana; Das, 2015).

In Portland cement mortars containing 20% metakaolin, crystalline products were observed in their microstructure. Specifically, for samples featuring aggregate A (Figure 9a), semicrystalline phases resembling those identified in Figure 10a were observed, alongside characteristic chabazite crystals (Du et al., 2017; Vallace et al., 2021). Similarly, the crystals observed in the microstructure of samples containing aggregate B (as illustrated in Figure 10b) bear resemblance to some geopolymers documented by Sanchindapong et al. (Sanchindapong et al., 2020). As previously noted, these compounds belong to the zeolitic material group, which can be obtained through the alkaline activation processes of aluminosilicates. The presence of zeolitic crystals in the microstructure of these samples suggests that metakaolin can generate geopolymers within this medium, thereby filling the hydrated Portland cement matrix.

Continuing to Figure 11, it showcases the geopolymeric mortars prepared using 8 mol/L sodium hydroxide solutions, wherein the geopolymerization reactions facilitated the formation of crystalline zeolites within the composite matrix. Moreover, these products were observed to nucleate on the surface of the aggregates in both samples. Kai and Dai (Kai; Dai, 2021) proposed that the interface between the aggregate and geopolymer is partially comprised of aluminum condensation in the Si-OH bonds of the crystalline silica aggregate, leading to water release into the medium. This phenomenon mirrors the solidification process of geopolymers, wherein monomeric units of silica and aluminum condensate to form the polymeric structure (Khale; Chaudhary, 2007; Singh et al., 2015; Singh; Middendorf, 2020; Wan et al., 2017). This suggests that the mineralogical phases present in the aggregates could serve as precursors of geopolymers, with their chemical composition and crystallinity being the primary factors defining their efficiency in utilization (Arnoult et al., 2019; Xu; Deventer, 2003).

Despite the challenges associated with obtaining energy dispersive X-ray (EDX) analyses in the samples due to their surface sinuosity and numerous pores, the utilization of this technique for geopolymer mortars with aggregate B and 8 mol/L solution proved successful. This analysis yielded approximate chemical composition data of the crystalline formations observed on the surface of the aggregates, as depicted in Figure 11b. The spectrum presented in Figure 12 indicates the presence of potassium within the zeolite crystal structure, which contrasts with the type of alkaline activator used for geopolymer production (Table 3). In this mortar, the sole source of potassium is the aggregate itself (Table 2), signifying that the mineral phases within the aggregate make it available.

For geopolymeric mortars prepared with 12 mol/L sodium hydroxide solutions, the elevated concentrations of alkalis and hydroxyls in the mixture led to the formation of non-crystalline products, contrasting with those observed in Figure 11 (Zhang et al., 2012). These compounds have been identified as sodalite, an amorphous to semi-crystalline zeolite precursor product commonly observed in geopolymers (Sarkar; Dana; Das, 2015; Zhang; MacKenzie; Brown, 2009). Additionally, upon examining the matrix density depicted in Figure 13, it is possible to visualize that mortars containing aggregate B exhibit shorter distances between geopolymer nucleation centers. This observation suggests that the presence of alkali-aggregate reaction (AAR) reactive mineral phases in these aggregates influences the microstructural products obtained in geopolymeric mortars.

From the conducted procedures, it became evident that both deleterious aggregates from alkali-aggregate reactions and metakaolin exert significant effects on the hydrated Portland cement and geopolymer matrices. The reactive mineral phases present in the aggregates promote the formation of expandable products within the microstructure of Portland cement mortars and concretes, consequently adversely impacting the durability of these structures.

However, the utilization of metakaolin can mitigate the effects of alkali-aggregate reactions in these mixtures, effectively controlling expansion and increasing mechanical strength due to the products obtained within the hydrated cement matrix. It is noteworthy to highlight that the deleterious mineral phases present in the aggregates also influence geopolymeric products. Interestingly, their presence contributes to the enhancement of mechanical strength in mortars containing reactive aggregates. This suggests that the mineral phases susceptible to alkaline activation for geopolymers can be related to those responsible for the development of alkali-aggregate reactions in Portland cement concrete.

Conclusions

Based on the conducted procedures and analysis of the results obtained from the evaluation of Portland cement and geopolymeric mortars with metakaolin under the influence of the alkali-aggregate reaction, several conclusions can be drawn:

1. the substitution of metakaolin in Portland cement mortars led to a reduction in linear expansion caused by alkali-aggregate reaction (AAR) in prismatic bars cured in sodium hydroxide alkaline solution, where the expansions observed in the bars were below the normative levels outlined in C1260 (ASTM, 2021);

2. metakaolin-based geopolymeric mortars exhibit no linear expansions in prismatic bars cured in an alkaline solution. This observation suggests that the dissolution of reactive aggregates in the microstructure of geopolymeric mortars does not generate expansive product;

3. the presence of deleterious phases within aggregates subjected to AAR, in conjunction with metakaolin, results in mechanical strength gains in both Portland cement and geopolymeric mortars;

4. within the microstructure of Portland cement mortars, when cured in sodium hydroxide solutions, metakaolin appears to undergoes a reaction which generate products that resembles geopolymers capable of filling the matrix. This hypothesis is substantiated by the characterized morphology of the products observed within the microstructure of the samples;

5. the presence of deleterious mineral phases within aggregates disrupts the morphology of products resulting from the alkaline activation of metakaolin and may integrate into the hardened geopolymer matrix;

6. the molarity of the geopolymer’s activating solutions also plays a significant role in determining the morphology and mechanical properties of the resulting geopolymers, where higher alkalis concentration demonstrated greater effectiveness in producing denser matrices with higher mechanical strength; and

7. the incorporation of metakaolin in Portland cement mortars has been generally effective in mitigating alkali-aggregate reactions (AAR) within the microstructure. The observed reaction products indicate that metakaolin substitutions lead to the formation of geopolymer-like compounds within the cement matrix. Concurrently, geopolymeric mortars appear to utilize AAR-susceptible aggregates, enhancing mechanical strength and resulting in a denser microstructure. This phenomenon presents a potential avenue for future research, with the objective of identifying similar geopolymer precursors and advancing this technology as a cost-effective alternative.

Acknowledgments

The authors would like to thank UEPG (State University of Ponta Grossa), C-LABMU (Complex of Multiuser Laboratories) and CNPQ (Brazilian National Council for Scientific and Technological Development) for financial support and infrastructure.

References

Edited by

Publication Dates

History