Synergistic Effects and Mechanisms of Plant Ash and Activator on Geopolymer Gel Formation, Hydration Evolution and Mechanical Properties

TL;DR

This study explores how plant ash and activator affect the setting time, hydration, and strength of eco-friendly cement materials.

Contribution

The study reveals the synergistic effects of plant ash and activator on geopolymer gel formation and mechanical properties.

Findings

Plant ash delays setting time and reduces early reaction speed while slightly lowering flowability.

APAG with 20% plant ash and 4% activator achieved 57.8 MPa compressive strength after 28 days.

Compressive strength correlates strongly with chemically bound water content and reaction degree.

Abstract

Against the backdrop of promoting green buildings and a circular economy, the development of efficient, sustainable, and low-carbon cementitious materials is of great significance for reducing resource consumption and carbon emissions. In this study, plant ash (PA) was used as a partial cement replacement, and a series of alkali-activated composite cementitious materials (APAG) were prepared by regulating the dosages of PA and alkali activator (AA). The evolution of their workability, hydration behavior, and mechanical properties was systematically investigated. The results show that the incorporation of PA effectively delayed the setting process of the system; compared with P0, the initial and final setting times of P20 increased by approximately 302% and 100%, respectively, thereby mitigating the excessively rapid early-age reaction of the alkali-activated system while causing only a slight reduction in flowability. In contrast, the addition of AA shortened the setting time of APAG and led to a gradual decrease in fluidity. When the PA dosage was 20% and the AA dosage was 4%, APAG achieved a 28 d compressive strength of 57.8 MPa while maintaining good workability. Further analysis revealed a strong linear correlation between compressive strength and chemically bound water content under different PA and AA dosages, indicating that the reaction degree is a key factor governing macroscopic mechanical performance. Microstructural characterization confirmed that the incorporation of PA and AA significantly altered the reaction pathways and the morphology of hydration products, providing a reasonable microstructural explanation for the evolution of macroscopic properties. These findings provide valuable insights into the high-value utilization of biomass waste and the broader application of green cementitious materials.

Figures

Full text

1. Introduction

Alkali-activated materials (AAMs) are cementitious materials synthesized from aluminosilicate precursors under alkaline conditions, in which Si-O-Al tetrahedra serve as the fundamental structural units [1]. Compared with conventional Portland cement, AAMs do not require high-temperature calcination during production [2], thereby offering significant advantages in reducing carbon emissions and energy consumption [3]. In addition, AAMs exhibit rapid early strength development and favorable durability performance [4,5,6]. However, conventional AAM precursors primarily rely on industrial by-products such as fly ash and blast furnace slag. With ongoing adjustments in the energy structure, the supply of these materials is becoming increasingly limited and their costs continue to rise [7]. Therefore, it is imperative to develop sustainable alternative precursors with abundant availability.

In recent years, the rapid development of the biomass energy industry has generated large quantities of biomass ash, and the treatment and utilization of these residues have become a significant challenge [8,9,10]. PA is mainly derived from solid residues produced during biomass power generation processes [11,12,13,14]. Previous studies have shown that PA is rich in CaO, SiO_2_, and Al_2_O_3_, indicating a certain potential cementitious activity [15,16], and thus it has been regarded as a promising supplementary cementitious material. However, when PA is used alone as a supplementary cementitious material in ordinary Portland cement systems, it often exhibits insufficient reactivity and a relatively high water demand [17,18].

The existing studies have indicated that incorporating PA into alkali-activated systems is considered an effective approach to activating its latent reactivity [19,20]. Hu Feng et al. [14] reported that the pozzolanic activity of PA is significantly enhanced under alkaline conditions, leading to a marked increase in its reaction degree. Abdulkareem et al. [21] further demonstrated that the Ca component in biomass ash can participate in the formation of gel products such as C-S-H and C-A-S-H under alkali-activated conditions, thereby promoting microstructural densification and improving the mechanical strength of the material. However, most existing studies have focused on the effects of either PA or AA on the performance of APAG from the perspective of single variables or limited performance indicators [22]. Stephen Adeyemi Alabi et al. [23] investigated the effect of partially replacing ordinary Portland cement with palm oil fuel ash (POFA) at replacement levels of 0%, 10%, and 15% on the compressive strength of concrete. The results showed that, within the investigated dosage range, the compressive strength of concrete increased with an increasing POFA content. Similarly, Ali, Sabir et al. [24] examined the performance of concrete in which cement was partially replaced with sugarcane bagasse ash at proportions of 0%, 5%, 10%, and 15%. The compressive strength, strength activity index, and water absorption were evaluated. The findings indicated that all mixtures containing sugarcane bagasse ash exhibited improved compressive strength, with the strength activity index reaching a peak value of 115.67% at a 10% replacement level. Under the combined action of PA and AA, the synergistic effects of different dosage combinations on workability, hydration kinetics, and macroscopic mechanical properties, as well as the evolution trends of these performance parameters with varying dosages, remain insufficiently understood and require further systematic investigation.

Based on the above considerations, this study employs alkali activation technology by introducing different dosage gradients of PA and AA to systematically investigate their effects on the workability, hydration reaction behavior, and macroscopic mechanical properties of APAG. Based on the experimental results, the variation trends and correlations among the relevant performance parameters are analyzed. In addition, X-ray diffraction (XRD) and scanning electron microscopy (SEM) are employed to elucidate the effects of varying PA and AA dosages on the characteristics of reaction products and the microstructural evolution of the APAG system. This study aims to provide experimental evidence and mechanistic insights for the rational utilization of PA in alkali-activated cementitious materials.

2. Results and Discussion

2.1. Workability of APAG

2.1.1. Effect of PA Dosage on Flowability

Figure 1 presents the flowability of APAG under different PA dosages. As observed from the figure, the flowability shows a mild linear downward trend with an increasing PA dosage.

Compared with the reference group, the flowability of the low-dosage groups (P10 and P15) decreases by 3.2% and 5.2%, respectively; for the relatively higher-dosage groups (P20 and P25), the decreases are 7.8% and 11.0%, respectively. This reduction can be attributed to the intrinsic properties of PA: its specific surface area (593.03 m^2^/kg) is larger than that of cement (348.70 m^2^/kg). As a result, PA has higher water absorption and is more susceptible to agglomeration than cement, which reduces the free water content in the system and thus impairs the flowability [25].

Notably, even for the highest-dosage group (P25), the flowability remains at 137 mm; this value fully meets the typical workability threshold in conventional construction. Moreover, the flowability decreases gradually with an increasing PA dosage, with a maximum reduction of less than 4% between any two adjacent mix groups, indicating no abrupt decline. This indicates that the PA dosage exerts a negligible impact on the flowability of APAG.

2.1.2. Effect of PA Dosage on Setting Time

Figure 2 shows the setting times of APAG with different PA dosages. As the PA dosage increased from 0% to 25%, the initial setting time of APAG markedly extended from 58 min to 348 min, while the final setting time increased from 183 min to 481 min, representing increases of 500% and 163%, respectively. This trend indicates that the incorporation of PA produced a pronounced retarding effect [26], and this effect intensified nonlinearly with the increasing dosage, becoming particularly sharp when the dosage exceeded 20%. This phenomenon is primarily attributed to the high water absorption capacity of PA [27] and the potential presence of a small amount of incompletely combusted organic matter. Consequently, a portion of the water involved in the hydration reaction is absorbed by PA, and some hydration products may be decomposed by the organic matter, leading to a progressively greater extension of the setting time as the PA dosage increases. Overmann, S. et al. [28] found that organic residues and unburned carbon in biomass ash increase the water demand and delay the formation rate of hydration products such as C-S-H, which manifests as a retardation of setting. Consequently, the extension of the setting time becomes increasingly pronounced with an increasing PA dosage.

It is worth noting that the interval between the initial and final setting first increased and then decreased with a higher PA dosage, rather than continuously widening as the overall setting time lengthened. The increase in the final setting time was greater than that in the initial setting time, indicating that the system requires a longer duration after the initial setting to complete its structural formation. When the PA dosage reached 25%, the setting interval slightly decreased. This may be due to the introduction of a large number of fine particles at high PA dosages, which act as micro-aggregates, filling voids and accelerating local densification. These results suggest that while PA incorporation extends the workable time, its dosage must be controlled to avoid excessively prolonged setting, which could adversely affect early strength development.

2.1.3. Effect of AA Dosage on Flowability

Figure 3 illustrates the effect of the AA dosage on the flowability of the APAG. As shown in the figure, the flowability decreases with an increasing AA dosage. At AA dosages of 2% and 4%, only a slight reduction in flowability is observed, with decreases of 5.8% and 9.7%, respectively. However, when the AA dosage reaches 6% and 8%, the flowability of the APAG declines sharply, with reductions of 15.3% and 23.4% compared to the previous group, respectively. The minimum flowability recorded is only 85 mm.

The possible reason for this phenomenon is that, with the addition of AA, the alkalinity of the APAG system increases linearly, thereby accelerating the depolymerization–repolymerization reactions of the silico-aluminate networks in PA and cement [29] and leading to the formation of more cementitious products. Tekle, B.H. reported that changes in plastic viscosity are closely related to the quantity and structure of cementitious products within the system. Interactions between solid particles are strengthened, the internal friction resistance increases, and the plastic viscosity rises significantly, which macroscopically manifests as a reduction in flowability [30].

It is noteworthy, however, that within the lower dosage range of 2% to 4%, the flowability remains above 131 mm, indicating that the system still maintains good workability within this interval.

2.1.4. Effect of AA Dosage on Setting Time

Figure 4 shows the effect of different AA dosages on the setting time of APAG. As can be seen, both the initial and final setting times decrease with an increasing AA dosage. It is observed that at an AA dosage of 8%, the initial setting time is less than 1 h, and the final setting time is less than 1.5 h. This is because AA neutralizes some of the organic acidic substances present in PA, accelerating the reaction process. At the same time, it provides a strongly alkaline environment that promotes the dissolution of silico-aluminous components in PA and their subsequent polymerization [31]. The synergistic effect of these two actions facilitates the rapid formation of a gel network, thereby shortening both the initial and final setting times.

An AA dosage of 4% appears to represent a transition point, at which the rate of decrease in setting time changes from sharp to more gradual. This trend is consistent with the influence of AA dosage on the flowability of APAG, suggesting that, at this dosage level, a more balanced interaction between setting kinetics and reaction progression is achieved.

As shown in the figure, the setting interval continues to decrease with the increasing AA dosage. This demonstrates that the addition of AA can substantially shorten the overall setting process, making the hydration reaction more concentrated and efficient.

2.2. Hydration Process of APAG

2.2.1. Effect of PA Dosage on Chemically Bound Water Content

Figure 5 presents the chemically bound water contents of APAG with different PA replacement ratios at curing ages of 3, 7, and 28 days. As shown in the figure, the chemically bound water content exhibits a non-monotonic variation with an increasing PA content, characterized by an initial decrease, followed by an increase, and then a subsequent decrease. Compared with the plain cement system (P0), the P10 mixture with a PA replacement ratio of 10% shows a significant reduction in the chemically bound water content at 28 days, with a decrease of 21.71%. This phenomenon can be attributed to the fact that, at low PA replacement levels, cement hydration remains the dominant controlling process. Partial replacement of cement by a small amount of PA reduces the effective concentration of cement particles, thereby retarding the hydration reaction and resulting in a lower amount of chemically bound water, which is consistent with the findings reported by Ukrainczyk, N [32].

For the P20 mixture, the chemically bound water content reaches its maximum value. Under this condition, the OH^−^ ions provided by AA effectively activate the silico-aluminate phases in PA, promoting their dissolution and releasing large amounts of silicate and aluminate monomers. On the one hand, these monomers react with Ca^2+^ released from cement hydration to form C-A-S-H gel; on the other hand, within a moderate alkalinity range, part of the dissolved species undergoes direct polycondensation to form N-A-S-H gel. This observation is consistent with the XRD results discussed in Section 2.3.1. However, when the PA replacement ratio increases to 25% (P25), the chemically bound water content decreases again. This reduction can be attributed to the synergistic antagonistic effects of the physical and chemical factors. Physically, the high specific surface area and strong water absorption capacity of PA competitively consume the free water required for reactions, thereby reducing the ion mobility [33]. Chemically, excessive PA may introduce trace organic components that adsorb onto particle surfaces, hindering the continuous dissolution of silico-aluminate phases and the polycondensation of gel networks, ultimately limiting the overall reaction degree [34]. ANOVA and Tukey HSD analyses indicated that the PA dosage exerted a highly significant effect on the chemically bound water content at all tested curing ages (3 d, 7 d, and 28 d, p < 0.001). For the 28 d curing period, an F-value of 130.189 with p < 0.01 was obtained, demonstrating an extremely significant influence. This conclusion was further validated by passing the Brown–Forsythe robust test (p < 0.01), confirming its high statistical reliability.

In addition, the chemically bound water content of all mixtures increases with curing age, indicating that although the incorporation of PA alters the reaction extent, it does not change the fundamental age-dependent development trend of the system.

2.2.2. Effect of PA Dosage on Leachate pH

Figure 6 shows the leachate pH of APAG specimens at curing ages of 3, 7, and 28 days with different PA dosages. As can be seen from the figure, except for the reference group (P0), the leachate pH shows an overall tendency to initially decrease, then increase, and finally decrease again with increasing PA content. This trend is generally consistent with that observed for the chemically bound water content in Figure 6, suggesting a potential correlation between the evolution of hydration products and the leaching behavior of alkaline species under different PA dosages.

Specifically, compared to P0, the leachate pH of P10 shows a minor reduction at 28 days (approximately 1.69%), which is within the experimental variability indicated by the error bars. As the PA dosage further increases, the leachate pH gradually rises and reaches its maximum at P20, with a 28 d pH value of 12.61. This behavior may be attributed to the dissolution of the alkali metal ions contained in PA, which are readily released into the aqueous phase during the extraction process, leading to an increase in the measured pH. However, when the PA dosage exceeds 20%, a decreasing trend in the leachate pH is observed. One possible explanation is that organic constituents present in PA may undergo secondary reactions with hydration products such as Ca(OH)2, forming soluble salts and thereby reducing the concentration of free alkaline species in the leachate [35]. In addition, excessive PA may promote secondary hydration reactions, resulting in the formation of C-S-H gel with a relatively low Ca/Si ratio, which has been reported to exhibit a stronger alkali-binding capacity than the C-S-H formed during ordinary cement hydration [36]. The combined effects of these processes may lead to a reduction in the measured leachate pH.

From the perspective of the curing age, the leachate pH of all mixtures shows an overall decreasing trend with an increasing age. Moreover, the decrease from 7 to 28 days is more pronounced than that from 3 to 7 days. This phenomenon can be explained by the progressive incorporation of alkaline species into hydration and alkali-activation products as curing proceeds. At early ages, a portion of alkali metal ions from PA can readily dissolve during the extraction process, resulting in relatively minor changes in the leachate pH. At later ages, the increased presence of gel phases with a higher alkali binding capacity leads to a more substantial reduction in the amount of leachable alkaline species, which is reflected by the greater decrease in leachate pH.

ANOVA and Tukey HSD analyses demonstrated that the PA dosage exerted a highly significant influence on the leachate pH at all tested curing ages (3 d, 7 d, and 28 d, p < 0. 001). This effect became even more pronounced at later stages, with F-values of 1363.7 and 1243.2 for 7 d and 28 d, respectively. The reliability of this conclusion was further corroborated by the Brown–Forsythe robust test (all p < 0. 001). Therefore, the incorporation of PA significantly alters the alkalinity of the reaction system, and its impact is both persistent and potent.

2.2.3. Effect of AA Dosage on Chemically Bound Water Content

Figure 7 shows the chemically bound water content of APAG at curing ages of 3, 7, and 28 days with different AA dosages. As the AA dosage increases, the chemically bound water content first rises and then declines, reaching its peak at an AA dosage of 4%. Compared to the reference group and the group with the lowest content, the growth rates of chemically bound water at 28 days are 30% and 24.66%, respectively. This indicates that the AA dosage has an influence on the chemically bound water content.

This behavior is closely related to the alkalinity-controlled dissolution and polycondensation processes of aluminosilicate phases in PA. PA primarily contains amorphous silico-aluminate components, which can be effectively dissolved under an appropriate alkaline environment. At moderate AA dosages, the concentration of OH-ions is sufficient to promote the dissolution of Si-O-Si and Si-O-Al bonds, releasing reactive silicate and aluminate species. These species subsequently participate in poly condensation reactions to form three-dimensional gel networks. Meanwhile, Ca^2+^ released from cement hydration reacts with the dissolved aluminosilicate species to form C-A-S-H gel, while part of the aluminosilicate species undergoes direct polycondensation to generate N-A-S-H gel. The synergistic coexistence of these gel phases enhances the overall degree of reaction, leading to an increase in chemically bound water. Therefore, an AA dosage of 4% represents an optimal balance between aluminosilicate dissolution and gel evolution.

When the dosage of the alkaline activator (AA) exceeds 4%, the chemically bound water content begins to decrease. Excessively high alkalinity accelerates early-stage reactions, leading to the rapid precipitation of hydration and alkali-activated products on particle surfaces. These products form diffusion barriers that hinder the continued dissolution of unreacted silico-aluminate and cementitious phases, thereby restricting further polycondensation reactions and gel development [37]. As a result, the chemically bound water content is ultimately reduced.

From the perspective of curing age, the increase from 3 to 7 days is far greater than that from 7 to 28 days. This is because the alkali-activated reaction primarily occurs in the early stages. The strength development in the later period is largely attributed to the refinement of the microstructure, such as reduced porosity, rather than a continued substantial increase in the degree of chemical reaction. ANOVA and Tukey HSD analyses revealed that the effect of AA dosage on the chemically bound water reached a highly significant level at all tested curing ages (3 d, 7 d, and 28 d, p < 0.001). For the 28 d curing period, the F-value was 354.39 with p < 0.01, indicating an extremely significant influence. This conclusion was further validated by passing the Brown–Forsythe robust test (p < 0.01), confirming its high statistical reliability.

2.2.4. Effect of AA Dosage on Leachate pH

Figure 8 presents the leachate pH in APAG at curing ages of 3, 7, and 28 days under different AA dosages. As shown, the leachate pH increases with higher AA dosage, indicating a positive relationship. This is can be attributed to the fact that AA, as an alkaline substance, contributes more free alkali ions dissolved in the pore solution at higher dosages, thereby directly elevating the pH of the system. When the AA dosage exceeds 2%, the rising trend of leachate pH moderates. The reason lies in the fact that an appropriate alkali concentration fully reacts with PA, consuming a substantial amount of OH^−^. Simultaneously, as PA is consumed, a certain amount of gel particles is generated, preventing the leachate pH from continuing to increase linearly with the rising AA dosage [38].

Across different curing ages, the leachate pH shows an overall decreasing trend with increasing age. This can be explained by the progressive incorporation over time, water is gradually consumed, and Ca(OH)2 in the APAG is progressively consumed by PA to form gels with high alkali-binding capacity, leading to a decrease in pH [39]. This pattern is similar to that observed under different PA dosages, where leachate pH decreases with extended age. Considering the trend of chemically bound water under different AA dosages, although the alkali-activated reaction primarily occurs in the early stages, it does not cause a sharp decline in internal alkalinity. The synergistic effects maintain a strong alkaline environment throughout the curing period, which is beneficial for long-term stability of the formed reaction products.

ANOVA and Tukey HSD analyses revealed that the effect of AA dosage on leachate pH exhibited significant time dependence: no significant influence was observed at the early stage (3 days), whereas at the mid-to-late stages, F-values of 3031.5 and 2274.68 for 7 and 28 days, respectively, indicated an extremely significant impact. This conclusion was further supported by passing the Brown–Forsythe robust test, confirming its high statistical reliability.

2.3. Compressive Strength of APAG

2.3.1. Effect of PA Dosage on the Compressive Strength of APAG

Figure 9 presents the compressive strength of APAG with varying PA dosages at curing ages of 3, 7, and 28 days. Regarding curing age, the compressive strength progressively increases with longer curing duration, aligning with the general trend observed in conventional permeable concrete. Excluding the reference group (P0), the overall trend at each specific age shows that compressive strength initially increases and subsequently decreases with higher PA dosage. The P20 mix attained the highest compressive strength, reaching 57.8 MPa at 28 days.

This can be attributed to the following factors: PA particles are generally finer than cement particles. At a 20% dosage, these fine particles effectively fill the voids between cement grains, while part of the PA reacts with the alkali activator to form cementitious products. SEM observations show that these cementitious products exhibit needle-like or rod-like morphologies and form a certain three-dimensional interwoven network, thereby enhancing the matrix density and manifesting as higher compressive strength at the macroscopic level [40].

From a compositional standpoint, the dosages of AA and PA in the system likely reached an optimal balance, allowing each component to react fully and exert synergistic positive effects. As previously indicated, the P20 mix exhibited the highest leachate pH, reaching 12.32 as early as 3 days. Such a highly alkaline environment accelerated the dissolution rate of reactive silica and alumina from the PA, generating a substantial amount of cementitious material. Consequently, this mix achieved the highest compressive strength among all groups except the reference group.

To verify the statistical significance of compressive strength differences under varying PA dosages, the data were analyzed using one-way analysis of variance (ANOVA), Tukey HSD post hoc test, and Brown–Forsythe robust test. The results are presented in Table 1, Table 2 and Table 3. Both the ANOVA and Tukey HSD post hoc test confirmed that the PA dosage is an extremely significant factor affecting compressive strength, with p < 0.001 at curing ages of 3, 7, and 28 days. This conclusion is further supported by the Brown–Forsythe robust test (all p < 0.001), indicating that the statistical inference is highly reliable and does not depend on the assumption of homogeneity of variances. Consequently, apart from the reference group (P0), P20 can be regarded as the critical dosage point in this system: when the PA dosage is below this level, the compressive strength shows a gradual declining trend with increasing PA content, whereas exceeding this dosage leads to a notable decrease in strength. This suggests that in practical applications, the PA dosage should be controlled within an appropriate range to achieve an optimal balance between environmental benefits and mechanical performance.

2.3.2. Effect of AA Dosage on the Compressive Strength of APAG

Figure 10 illustrates the effect of different AA dosages on the compressive strength of the APAG at curing ages of 3, 7, and 28 days. As observed from the figure, the overall trend across all ages shows that the compressive strength initially increases and then decreases with rising AA dosage, with the peak consistently occurring at a 4% AA dosage. This pattern closely aligns with the variation in chemically bound water content under different AA dosages, indicating that the alkali-activated reaction is the most complete at the 4% dosage, generating a greater amount of cementitious products.

Within the AA dosage range of 0% to 4%, the leachate pH continuously increases with a higher dosage, providing the optimal reaction driving force and effectively promoting the dissolution and polycondensation of PA. However, when AA exceeds 4%, the excessively high alkalinity accelerates the reaction rate excessively, leading to stagnation in the growth of chemically bound water and hindering further internal chemical reactions [41,42]. Consequently, excessive AA dosage results in a declining trend in compressive strength with increasing dosage.

According to the analysis of ANOVA results, the AA dosage had a highly significant effect on compressive strength at all curing ages (all p < 0.001). Tukey HSD testing further clarified that the compressive strength of the A4 group was significantly higher than that of all other comparison groups (A0, A2, A6, A8), with all differences being statistically highly significant (all p < 0.001). These findings were also supported by the Brown–Forsythe robust test (p < 0.01), confirming high statistical reliability. Therefore, the statistical analysis demonstrates that the AA dosage is a key factor influencing the compressive strength, with 4% identified as the optimal dosage.

2.3.3. Relationship Between Compressive Strength and Chemically Bound Water Content Under Different PA Dosages

Figure 11 illustrates the relationship between the compressive strength and the chemically bound water content of the APAG at curing ages of 3, 7, and 28 days under different PA dosages. A clear positive correlation can be observed between the compressive strength and the chemically bound water content, with a coefficient of determination (R2) of 0.93 within the investigated data range.

The increase in chemically bound water reflects the progressive development of hydration and alkali-activation reactions, which are accompanied by the formation of binding phases such as C-S-H and aluminosilicate gels. As the amount of these reaction products increases, the microstructure becomes denser, leading to the enhanced load-bearing capacity of the matrix.

It should be noted that the observed linear relationship represents an empirical correlation under the specific mix compositions and curing conditions considered in this study. Within this range, chemically bound water can be regarded as an effective indicator of the reaction extent governing the macroscopic mechanical performance of APAG.

The SEM observations further support this interpretation, showing that higher chemically bound water contents are associated with increased gel formation and reduced interparticle voids, which collectively contribute to the observed strength development.

2.3.4. Relationship Between Compressive Strength and Leachate pH Under Different PA Dosages

Figure 12 illustrates the relationship between compressive strength and the leachate pH under different PA dosages. As shown in the figure, the alkalinity values are distributed within a relatively narrow range (12.10–12.69), while the corresponding compressive strengths at 3, 7, and 28 days exhibit distinct stratification.

To better describe this relationship, the data at each curing age were analyzed separately. For all three ages, the variation in compressive strength with leachate pH can be empirically described by a quadratic trend, indicating a non-linear dependence within the investigated alkalinity range. The fitted curves show high coefficients of determination, reflecting a clear correlation between these two parameters under the present experimental conditions.

At a given curing age, higher compressive strength is generally associated with higher leachate pH, suggesting that a sufficiently alkaline environment is favorable for the hydration and alkali-activation reactions at early stages. In contrast, a comparison of different curing ages reveals that the longer the curing age, the lower the leachate pH value, while the compressive strength increases continuously with the extension of curing age, ultimately resulting in an overall leftward shift in the fitted compressive strength–alkalinity curve.

This evolution indicates that although the leachate pH plays an important role in the early reaction stage, the later strength development of APAG is mainly governed by the accumulation and reorganization of reaction products, rather than by maintaining a high alkalinity level. Such behavior is consistent with the progressive consumption of alkaline species during hydration and alkali-activation reactions, accompanied by continuous microstructural densification.

2.3.5. Relationship Between Compressive Strength and Chemically Bound Water Content Under Different AA Dosages

Figure 13 shows the relationship between the chemically bound water content and compressive strength under different AA dosages. As can be seen from the figure, the compressive strength of APAG ranges from 32.8 to 57.8 MPa, and the chemically bound water content varies between 4.2% and 9.1%. With the increase in chemically bound water, the compressive strength exhibits a linear upward trend. This trend aligns with the relationship observed between the compressive strength and chemically bound water under different PA dosages. The goodness of fit reaches 0.9, which fully demonstrates that in the APAG system, despite variations in cementitious materials and precursors, chemically bound water can still serve as an important indicator for characterizing mechanical performance.

2.3.6. Relationship Between Compressive Strength and Leachate pH Under Different AA Dosages

Figure 14 illustrates the relationship between compressive strength and leachate pH under different AA dosages. As can be observed from the figure, both compressive strength and leachate pH exhibit a decreasing trend with the increasing curing age. At any given age, the compressive strength initially increases and then decreases with the rising leachate pH. This pattern is consistent with previous findings, indicating that either excessively high or excessively low alkalinity is unfavorable for strength development. Specifically, the leachate pH value exhibits an optimal range within a relatively narrow interval around 12.3, where the system demonstrates the highest compressive strength. In contrast, significant deviation from this pH range leads to a decline in strength.

The figure also shows that the goodness of fit between the compressive strength and leachate pH under different AA dosages is relatively low. The reason for this may be that with a constant amount of cementitious materials, the continuous addition of alkaline activator directly introduces a large amount of OH^−^ ions into the APAG. The retention of excessive OH^−^ within the gel leads to a sharp, direct increase in the leachate pH as the AA dosage rises. Consequently, under varying AA dosages, the leachate pH cannot reliably be used to predict the mechanical performance of the APAG.

2.4. Phase Composition and Microstructure of APAG

The macroscopic properties of a material are often determined by its microstructure and phase composition. Experimental investigation has confirmed that the P20 (i.e., A4) mix exhibits optimal characteristics in terms of its workability, hydration process, and mechanical strength. However, the underlying micro-mechanisms responsible for the enhancement of its macroscopic properties remain unclear. Therefore, the P20 group was selected as a representative sample, with the P0 group serving as a blank control, for phase composition analysis (XRD) and micro-morphology examination (SEM). The aim is to verify, at the micro-scale, the intrinsic causes of its superior performance—such as gel phase composition, product distribution, and structural densification patterns—thereby providing a theoretical basis for subsequent material optimization.

2.4.1. Phase Composition

Figure 15 presents the X-ray diffraction (XRD) patterns of the P0 and P20 specimens after 28 days of curing. Compared with P0, P20 exhibits a broader diffraction “hump” in the 2θ = 25–35° range, accompanied by an elevated amorphous background. This feature suggests a reduction in crystalline phase content and a corresponding increase in amorphous reaction products [43], implying that the incorporation of PA promotes the formation of amorphous phases.

Further analysis shows that both P0 and P20 contain calcite, dolomite, quartz, portlandite, and brucite as hydration or residual mineral phases. However, the diffraction peak intensity of Ca(OH)2 in P20 exhibits a certain degree of reduction. This change can be attributed to the dilution effect associated with the partial substitution of cement by PA, as well as to the pozzolanic activity of PA under alkaline conditions, in which dissolved reactive Si and Al species react with and consume Ca(OH)2 through secondary reactions, leading to the formation of additional gel phases [44]. This process favors the formation of low-crystallinity calcium silicate or aluminosilicate-type gels, thereby reducing the amount of free Ca(OH)2 [45,46].

The experimental results further demonstrate a good linear correlation between the compressive strength and chemically bound water content under different PA and AA dosages, indicating that the degree of reaction is closely related to the macroscopic mechanical performance. The enhancement of the amorphous diffuse hump in P20 is commonly associated with an increase in gel phases, which is consistent with the observed rise in chemically bound water content, suggesting that PA participates in a sustained reaction under alkaline activation.

Combined with SEM observations, although a greater amount of gel is formed in P20, its microstructure overall appears to be relatively loose. This indicates that the increase in gel content does not directly translate into significant structural densification; instead, it is more likely reflected in the development of a gel network structure and improved interfacial bonding. As a result, the compressive strength of P20 exhibits only a slight difference compared with P0, without a significant increase or decrease.

2.4.2. Microscopic Morphology Analysis

Figure 16 presents the SEM micrographs of P0 and P20 at different magnifications (10 μm and 2 μm). Overall, the two samples exhibit evident differences in microstructural features, indicating that the incorporation of PA significantly influences the reaction process and the morphology of the resulting products.

At a low magnification, the P0 sample predominantly exhibits a relatively continuous blocky or agglomerated matrix structure with a comparatively uniform pore distribution [47], whereas the P20 sample shows more pronounced pore features and a rougher matrix surface. At a higher magnification, a large number of slender fibrous or needle-like products can be clearly observed in P20; such morphologies have been widely reported in Ca–Si–Al systems under alkaline activation or highly alkaline environments [48]. Tambara Júnior et al. [25] reported that needle-like gel structures formed in alkali-activated rice husk ash systems can significantly alter the microstructural configuration of the matrix, while Olivera Bedov et al. [49] further summarized that when biomass ash participates in the reaction, C-A-S-H and N-A-S-H type gels are commonly generated, typically exhibiting fibrous or network-like morphologies. Combined with the enhanced amorphous diffuse hump observed in the XRD patterns in this study, it can be reasonably inferred that the fibrous structures observed in P20 may correspond to low-crystallinity aluminosilicate or calcium silicate-type gel phases.

Further considering the setting time results, the incorporation of PA delays the early-stage reaction process of the system and reduces the rate of ion dissolution and precipitation [26], thereby moderating the structural build-up process and leading to a relatively loose matrix morphology at the microscale [50,51]. Although the overall compactness of P20 is lower than that of P0, its macroscopic compressive strength exhibits only a slight variation. This indicates that strength development is not solely governed by the pore quantity or apparent compactness, but is also closely related to the type of reaction products and their spatial distribution [32,33,52,53]. The increased proportion of low-crystallinity gels in P20, together with the formation of a fibrous network structure, may enhance the internal bonding continuity and stress transfer capacity of the matrix, thereby partially compensating for the adverse effect of increased porosity on mechanical performance [54,55].

3. Materials and Methods

3.1. Experimental Materials

PA is a by-product generated from biomass combustion in thermal power plants, primarily sourced from Maozhuo Building Materials Co., Ltd. in Lingshou County Hebei Province, (Shijiazhuang, China). Prior to testing, it was dried in an oven at 105 °C for 24 h. After complete drying, it was sieved to a particle size below 0.15 mm for subsequent use. The specific surface area of the dried PA was measured as 593.03 m^2^/kg, with a density of 2.43 g/cm^3^. The loss upon ignition of PA is 8.1%. The main chemical composition of the biomass ash, as determined by XRF analysis, is presented in Table 4.

The cement used was Ordinary Portland Cement of grade P·O 42.5, produced by Fengbo Tianrui Cement Co., Ltd. (Ruzhou, China). The loss upon ignition of the cement is 3.2%. Other performance indicators of the cement are shown in Table 5. The fine aggregate was natural river sand supplied by Errong Building Materials Co., Ltd., also located in Henan, (Zhengzhou, China). The alkaline activator employed was sodium hydroxide in pellet form with a purity of 96%, produced by Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China. The mixing water was ordinary tap water from the Zhengzhou region of China.

3.2. Mix Design

According to previous studies [21,56,57], the replacement level of biomass ash in cement-based systems generally does not exceed 25%, in order to avoid significant reductions in mechanical strength and workability at high replacement ratios. Therefore, in this study, PA was used to partially replace cement by mass at levels of 0%, 10%, 15%, 20%, and 25% of the total binder.

To systematically investigate the effects of the PA content and AA level on the performance of APAG, a two-stage mix design strategy with clearly controlled variables was adopted.

In the first stage, a series of mixtures (denoted as the P series) was designed to evaluate the influence of PA replacement under a fixed alkali activation condition. In this series, P20 denotes a mixture in which PA accounts for 20% of the total binder. The NaOH dosage was fixed at 4% of the PA mass and used as the alkaline activator. Previous studies have demonstrated that a moderate alkali concentration can effectively activate the amorphous aluminosilicate phases in biomass ash while avoiding adverse effects associated with excessive alkalinity. In the P series, the PA replacement ratio was the only variable, enabling an independent assessment of its effect on material performance.

Based on the results of the P series, in which the P20 mixture exhibited the most favorable overall performance, a second series was designed to optimize the alkali activation level. In this series, the PA replacement ratio was fixed at 20%, while the alkali activator (NaOH) content was varied from 0% to 8% at intervals of 2% (0%, 2%, 4%, 6% and 8%), and the corresponding mixtures were labeled A0, A2, A4, A6 and A8, respectively. The NaOH dosage was calculated based on the mass of PA per unit volume (m^3^). This dosage range (0–8%) was selected to cover conditions from no alkali activation to moderate alkali activation.

For all mixtures, the water-to-binder ratio (defined as the mass ratio of water to total binder, including cement and PA) was maintained at 0.37, and the sand ratio was fixed at 42%, ensuring strict control of variables so that the observed performance variations could be clearly attributed to changes in the PA or AA content. In the mix design, the total water content consisted of base mixing water and additional water. “Water” refers to the base mixing water, while “additional water” denotes the extra water added to compensate for the water absorption of PA, ensuring comparable initial workability among mixtures. The detailed mix proportions are summarized in Table 6.

3.3. Preparation Procedure

A specific preparation procedure was employed for mixing APAG. Sodium hydroxide was first dissolved in pre-measured mixing water to prepare the alkaline activator. After wetting the mixing blades, the premixed blend of PA and cement was added. The mixing protocol was as follows: low-speed mixing for 30 s, addition of fine aggregate followed by another 30 s of low-speed mixing, then switching to high-speed mixing for 90 s. The alkaline activator was poured in slowly during the high-speed mixing phase.

After mixing, the APAG was placed into molds in three layers, with each layer tamped using a tamping rod and vibrated on a vibration table until a leveled surface was achieved. Subsequently, the specimens were covered with a membrane and cured outdoors for 24 h. This initial outdoor curing stage was adopted to allow for the sufficient setting and early strength development of the specimens prior to demolding, which is commonly required for APAG. During this period, the specimens were sealed with the membrane to minimize moisture loss and to limit exposure to atmospheric CO_2_, thereby reducing drying and early carbonation. After demolding, the specimens were transferred to a standard curing room (curing chamber at (20 ± 2) °C and ≥95% relative humidity) and cured until the designated testing ages (the curing ages were set at 3 days, 7 days, and 28 days, respectively). The preparation process is illustrated in Figure 17.

3.4. Test Methods

3.4.1. Flowability

The flowability test in this experiment was conducted in accordance with the Chinese National Standard GB/T 1346-2011 [58]. First, a glass plate was wiped with a damp cloth to achieve a moist surface without visible water streaks. The plate was then placed on a level table and adjusted to a horizontal position. A mold, which had been covered with a wet cloth, was positioned at the center of the glass plate. Subsequently, the appropriate amounts of cement and PA, sand, water, and mixtures were weighed according to the mix proportion and placed into a mortar mixer for thorough mixing. The freshly mixed APAG was then poured into the mold, and the top surface was struck off level. The mold was lifted vertically, and a stopwatch was started simultaneously. After allowing the mixture to flow freely on the glass plate for the specified duration, a ruler was used to measure the maximum diameter in two perpendicular directions. The average of these two measurements was taken as the final flowability result.

3.4.2. Setting Time

The determination of the setting time in this test was conducted in accordance with the Chinese National Standard GB/T 1346-2011 [58]. The mixture was prepared according to the mix design, placed into molds, and compacted. It was subsequently cured in an environment maintained at 20 ± 2 °C with a relative humidity of ≥90%. Timing commenced immediately upon completion of mixing. The test intervals were controlled in stages. A penetration resistance apparatus was used to penetrate the sample to a depth of 25 mm at a speed of 2–5 mm/s. Three measurement points were tested for each interval, and the average value was recorded. A curve was plotted with time on the x-axis and penetration resistance on the y-axis. The times corresponding to the specified resistance values were identified as the initial and final setting times, respectively. Throughout the test, the temperature was strictly controlled, and measures were taken to prevent moisture evaporation from the samples.

3.4.3. Chemically Bound Water

In this test, specimens cured for the specified ages were selected, crushed, and ground into a fine powder. The powder was then passed through a 0.08 mm square-hole sieve. Approximately 3 g of the sieved sample was weighed and placed in an oven at 65 ± 2 °C for 2 h, until a constant weight was achieved, ensuring complete removal of free water. The sample was then removed and immediately weighed, with the mass recorded as m1. Subsequently, the dried sample was transferred to a muffle furnace and ignited at a high temperature of 1000–1050 °C for 3 h until it was a constant weight. During this process, the chemically bound water was driven off as water vapor. After cooling, the sample was weighed again, and the resulting mass was recorded as m2.

The specific calculation formula for chemically bound water is as follows:

$$[eqn]$$ $$[eqn]$$

where $[eqn]$ refers to the chemically bound water; $[eqn]$ and $[eqn]$ refer to the loss upon ignition of the raw materials of cement and PA, respectively; $[eqn]$ , $[eqn]$ and $[eqn]$ represent the mass fractions of the raw materials, cement and PA, respectively; $[eqn]$ is the mass of the sample after drying in an oven at 65 + 2 °C to a constant weight; and $[eqn]$ is the mass of the sample after ignition at 1000–1050 °C for 3 h.

3.4.4. Leachate pH

The alkalinity of the system was evaluated by measuring the pH of the leachate using a solid–liquid extraction method. Hardened specimens at the specified curing ages were retrieved, crushed, and ground into a fine powder. The powder was sieved through a 0.08 mm sieve, and 3 g of the obtained fine powder was mixed with 30 g of distilled water. The suspension was stirred for 30 min using a magnetic stirrer at a speed of 200 rpm and then allowed to stand for 12 h to reach dissolution equilibrium. Subsequently, the pH value of the supernatant solution (leachate) was measured using a Leici PHS-25 pH meter manufactured by Shanghai INESA Scientific Instrument Co., Ltd. (Leici) (Shanghai, China). It should be noted that the measured pH value represents the alkalinity of the leachate, rather than the in situ chemical composition of the pore solution, and is therefore used as a comparative indicator of the alkaline environment among different mixtures.

3.4.5. Compressive Strength

The compressive strength test of APAG in this study was conducted in accordance with the Chinese National Standard GB/T 17671-2021 [59]. Cubic specimens measuring 40 mm × 40 mm × 40 mm were used, with three replicates prepared for each mix proportion. The tests were performed using a computer-controlled electro-hydraulic servo pressure testing machine manufactured by Shanghai Sansi Zongheng Machinery Manufacturing Co., Ltd., Shanghai, China. The compressive strength was measured at curing ages of 3, 7, and 28 days.

3.4.6. X-Ray Diffraction (XRD)

XRD analysis was carried out using a D8 series X-ray diffractometer manufactured by Bruker-AXS, Karlsruhe, Germany. Representative samples were taken from specimens cured for 28 days, ensuring that contamination was avoided to reliably reflect the composition of APAG. The samples were first broken into small pieces and then ground into a fine powder. During grinding, care was taken to achieve a sufficiently fine particle size. The sieved powder was dried in an oven at 60 °C to eliminate any moisture that might affect the analysis results. After drying, a portion of the sample was sealed for storage. The test conditions were as follows: scanning angle ranging from 5° to 70° (2θ), scanning speed of 0.6 s/step, and a step size of 0.02°/step.

3.4.7. Scanning Electron Microscopy (SEM)

Scanning electron microscopy observations were performed using an FEI QUANTA-650 series microscope (USA) (FEI Company, Hillsboro, OR, USA). The procedure was as follows: first, small blocks of approximately 1 cm × 1 cm in size were cut from the core area of specimens at the specified curing ages. The samples were thoroughly cleaned with distilled water and then dried in an oven at 60 °C until a constant weight was achieved, to ensure no residual moisture interfered with observation. Prior to testing, the sample cross-sections were coated with gold, using a high-resolution sputter coater to prevent the charging effects. Observations were then conducted at accelerating voltages between 5 kV and 20 kV, with the magnification adjusted as needed to obtain clear images of the sample’s microstructure.

4. Conclusions

This study primarily investigated the effects of different PA and AA dosages on the workability, hydration process, and mechanical properties of APAG. The main conclusions are summarized as follows.

Through the rational incorporation of PA combined with the synergistic design of AA, the setting behavior of APAG can be effectively regulated without significantly compromising its flowability, while simultaneously ensuring the normal development of mechanical properties, thereby improving its constructability during the placement stage. This finding provides experimental evidence for optimizing the APAG mix design by balancing workability, early-age reaction control, and mechanical performance development.

Under various PA and AA dosages, a strong linear correlation was observed between the chemically bound water content and compressive strength, indicating that variations in reaction degree can effectively reflect the development of mechanical properties.

It should be noted that this study mainly focused on the relationship between the early-age workability, reaction behavior, and mechanical properties of APAG, while durability issues were not systematically addressed. Nevertheless, the early reaction characteristics and structural evolution patterns revealed under the synergistic effect of PA and AA may provide useful guidance for future durability investigations and the engineering-oriented design of the APAG mixtures.