3.1. Mortar Flowability
Figure 6 illustrates the results of a flow table test to evaluate the influence of FBA as a fine aggregate and partial cement replacement on the flowability of mortar. The results indicate that the mix with no replacement (CR0) exhibited the highest flowability; when FBA was included as a fine aggregate replacement, the mortar flowability decreased. The reduction in mortar flowability increased as the replacement level increased, with reduction percentages of 1.6%, 4.9%, 7.0% and 10.8% corresponding to the inclusion of 25%, 50%, 75%, and 100% of FBA as a fine aggregate replacement, respectively. These reductions are attributable to the high-water absorption of FBA compared to natural fine aggregate, as presented in
Section 2.1. The irregular and rough shape of the FBA particles may have had an effect on increasing intermolecular friction and reducing the kinetic energy of the mortar, thus reducing its flowability [
54,
55]. The results are consistent with observations reported elsewhere [
15,
17], where the workability reduction tended to increase as the level of replacement increased. In the case of the inclusion of GFBA as a partial cement replacement, the F10 mix showed a 4.9% decrease in flow value compared to CR0, identical to the FBA25 mix. The F20 and F30 mixes experienced a 3.8% and 5.4% reduction compared to CR0 and 2.2% and 3.8% compared to the FBA25 mix, respectively. The decrease in flow diameter values compared to the control samples could have attributed to the rough and irregular surface of GFBA (
Figure 1) compared to cement particles. The rough and irregular shape of the GFBA particles may have had an effect of increasing intermolecular friction, unlike the semi-smooth shape of cement particles, thus reducing the mortar flowability. Furthermore, the high fineness of GFBA particles as shown in
Table 2 means they to tend to fill the inner area of the mixture particles, thus reducing the area available for particles dynamics; therefore, an excessive amount of water is needed to lubricate the mixture to give it enough flowability [
56]. These findings are consistent with several studies that have reported decreased flow values in mixes containing GFBA, such as the studies by Abdulmatin et al. [
23], Abbas [
24], and Aydin [
25]. These researchers found that the reduction in mortar flowability gradually increased with increased GFBA content.
3.2. Mortar Densities
Figure 7 presents the densities of the prepared mortar samples measured at 7, 28, and 56 days, with varying percentages of FBA as fine aggregate and cement replacements. From
Figure 7, the density decline is obvious in the case of using FBA as a fine aggregate. For the FBA0 mix, which had no FBA replacement, the densities were 2.34, 2.34, and 2.32 kg/m
3 at 7, 28, and 56 days, respectively. The FBA25 mix has densities of 2.20, 2.19, and 2.17 kg/m
3 at the respective time points, showing a similar gradual decrease. This trend was consistent across the mixes with higher FBA contents; for instance, the FBA100 mix, in which the fine aggregate was fully replaced by FBA, recorded the lowest densities across all the time points—1.81, 1.78, and 1.80 kg/m
3. The reduction in the mortar density with the increase in FBA content might be due to the lower density of FBA compared to the fine aggregate it replaces. In the case of the inclusion of GFBA as partial cement replacement, there was no regular trend in densities. For example, F10 showed a 0.6% higher density compared to FBA25 at 28 days. This could be attributed to the filling effect, leading to a more compact matrix. The reduction extent of densities was lower than that of using FBA as a fine aggregate. However, the F20 mortar showed similar densities to those of FBA25, while F30 showed lower densities by around 1%. Compared to previous studies, Mangi et al. [
57] and Oruji et al. [
27] found that mortar densities decreased with increasing replacement levels of cement by GFBA. They attributed this to the lower specific gravity of GFBA compared to Portland cement. Similarly, in the case of fine aggregate replacement, several studies [
58,
59,
60] indicated a decrease in mortar density when FBA was used as a sand substitute. This is due to the lower specific gravity of FBA, which ranges from 1.3 to 2.2 g/cm
3, as summarized by Onaizi et al. [
7], compared to approximately 2.5 to 2.65 for conventional aggregates. These findings are consistent with the results of this study, which observed a clear reduction in mortar densities with increased fine aggregate replacement due to the lower density of FBA (1.45 g/cm
3) compared to 2.5 g/cm
3 for the natural fine aggregate.
3.3. Compressive Strength
Figure 8 depicts the compressive strength of the mortar samples with various replacement levels of FBA over time, specifically at 7, 28, and 56 days. The control mix (CR0), without any FBA replacement exhibited an initial compressive strength of 25.4 MPa at 7 days, which substantially increased to 34.9 MPa at 28 days and reached 38.9 MPa at 56 days. In contrast, the FBA25 mix, with a 25% replacement of fine aggregate by FBA, started with a higher initial strength of 27.2 MPa at 7 days, surpassed CR0 at 28 days by 7.2%, and achieved the highest compressive strength among all mixes at 56 days with 42.3 MPa. For the FBA50, FBA75, and FBA100 mixes, there was a slight decrease in strength at all time points compared to CR0, and the strength reduction increased with increase in fine aggregate replacement level. However, FBA50 achieved higher strength by 5.9% compared to the control sample without FBA. This could be attributed to the partial pozzolanic reaction between the fine particles of FBA and Ca(OH)
2, resulting in the enhancement of the early strength [
61]. This potentially occurs with all mixes containing fine FBA as sand; however, at high replacement levels, the negative effect of FBA porosity and weak FBA particles likely outweighed that improvement. When 10% and 20% GFBA were used as cement replacements with 25% of FBA as a fine aggregate replacement, there was more enhancement of compressive strength. The compressive strength of both mixes, F10 and F20, surpassed those of CR0 and FBA25 at 28 and 56 days. This indicates the effectiveness of co-incorporation of GFBA as partial replacements for both cement and fine aggregate. This improvement might be correlated to improvement of microstructures, which provides a more compacted and stronger bonding possibility between the binder and the aggregate. Meanwhile, the finer particles and those used as a substitute for cement reacted with the Ca(OH)
2 produced from the initial cement hydration, providing additional hydrates [
62]. The relevant studies [
63,
64] indicate that the improvement in the strength of cementitious mixtures is enhanced by the growth of AFt, which significantly improves the microstructures of the matrices. This was specifically observed in this study for the F10 and F20 samples (as discussed in
Section 3.6). These results align with previous observations reported in the literature. For instance, a study [
65] indicated that compressive strength associated with a 15% replacement level of FBA surpassed that of control specimens by approximately 17% at 28 days. Similarly, another study [
66] reported that 10% GFBA increased compressive strength by around 10% at 28 days. Moreover, another study [
24] noted comparable compressive strength to control specimens at 28 days for specimens made with 10% and 20% FBA. These findings suggest that finely ground coal bottom ash can improve the compressive strength of mortars by up to 20% by mass substitution compared to a control mortar without GFBA. However, these findings contradict some earlier studies [
20,
23,
67] that reported a reduction in mortar strength regardless of the replacement ratios. It was suggested that the lower early-stage strength gain was due to the lower activity of the GFBA, which retards the cement hydration process [
7,
59]. However, it is important to note that these previous studies focused on the use of GFBA as a cement substitute, and there are currently no results available on the combined replacement of fine aggregate and cement.
3.5. Water Absorption
Figure 10 shows absorbed water percentages for mortar samples with FBA as partial alternatives of fine aggregate and cement at 28 days. The water absorption rate of CR0 was 4.61%. When FBA is used as fine aggregate replacement, the percentages of absorbed water increased, and the increment was boosted by increasing the replacement level. Starting with FBA25, the increment was 8.2% compared to that of CR0. This trend reversed slightly with FBA75, where water absorption decreased from 5.98 for FBA50 to 5.54% for FBA75. The higher water absorption rate was observed for the mortar sample with 100% replacement of the fine aggregate with FBA; for this sample, the water absorption was 6.58%, which was higher than that of CR0 by 42.7%. The increase in water absorption with increasing FBA content could be attributed to the intrinsic properties of FBA, particularly its porosity, which might contribute to a greater overall porosity of mortar and thus higher water absorption. However, with the co-incorporation of FBA, 25% as fine aggregate replacement and 10%, 20%, and 30% GFBA as partial cement replacements, the water absorption rates were slightly lower than the FBA25 sample. This improvement in water absorption resistance can be attributed to both the pozzolanic reaction and the filling action by the fine particles. It can be claimed that the high fineness boosted the pozzolanic reactivity of GFBA particles due to the availability of higher surface area per volume, which resulted in the C-S-H gel filling the pores, hence, reducing water absorption [
39]. This finding contradicts that reported by Bheel et al. [
65], who reported that the water absorption regularly decreased with an increase in the FBA content for all mixture groups. The water absorption decreased from 3.8% (for the control sample) to 1.98%, 1.75%, 1.6%, and 1.52% when cement was replaced with 10%, 20%, 30%, and 40% FBA, respectively. However, it is worth noting that the majority of studies [
71,
72] conducted on the use of FBA as an alternative to fine aggregate recorded an increase in water absorption due to the high porosity of FBA particles compared to natural aggregate.
The relationship between the compressive strength and water absorption of the designed mortars at 28 days is illustrated in
Figure 11. The water absorption was found to be inversely proportional to compressive strength, whereby the water absorption dropped from 6.6 to 4.5% and the strength increased from 29.2 to 40.5 MPa. The exponential regression method was applied to correlate the experimental data with the R
2 value of 0.8002, indicating a significantly good confidence in the relationships.
3.6. Microstructure Analysis and Discussion
Figure 12 shows the SEM images of the CR0, FBA25, F10, F20, F30, and FBA100 mortars at 28 days. FBA25 shows noticeable pores in some regions within the matrix, potentially due to the presence of the more porous FBA particles. Overall, FBA25 has a more uniform microstructure than CR0. The number of these pores increased with increasing FBA content as a replacement for fine aggregate, as illustrated in
Figure 12f for the FBA100 sample. When 25% FBA was used as fine aggregate replacement and 10%, 20%, and 30% ground FBA were used as partial cement alternatives, the microstructure became even more compact and uniform. This could be attributed to the finer particles of GFBA filling the pores and cracks in the mortar matrices, positively impacting the mechanical performance of F10, F20, and F30 samples.
The ability of FBA to absorb water may also play a role in reducing the available water-to-cement ratio (w/c) in the mixtures, enhancing compactness of the mortar matrixes and interfacial transition zones. Wang et al. [
73] indicated that in cementitious composites, an excess amount of water beyond what is necessary for cement hydration leads to increased voids in the hardened pastes. This may explain the homogeneity of the microstructures in F10 and F20 compared to CR0. However, with the increased content of FBA as a replacement for fine aggregate, as in the FBA100 sample, or as a substitute for cement, as in the F30 sample, the negative effects of high porosity and a weaker pozzolanic reaction compared to cement gradually negated the positive effects of filling and controlling the water-to-cement ratio available for hydration and early pozzolanic reaction.
The average elemental mapping from scanned 500-micron areas is also shown in
Figure 12. The EDS mapping demonstrated that Ca decreased with increasing level of FBA either as a fine aggregate or cement alternative. In contrast, the concentrations of Si and Al increase with the increased inclusion of FBA.
Figure 13 shows the ratios of Ca/Si, Ca/Al/Si/Al, and modifiers/formers network elements. The highest Ca/Si ratio was recorded for control mortar (CR0), where it was 2.54, which then dropped to 1.87 when 25% of FBA was added as a fine aggregate alternative. The drop in Ca/Si continued with the increasing inclusion of FBA, where it reached 0.94 for F30 and 0.34 for FBA100. When correlating the Ca/Si ratio to strength performance, it can be concluded that the ideal range for the Ca/Si ratio to obtain optimum strength is between 2 and 1, as shown in
Table 6.
Hydration of pure Portland cement results in a C–S–H phase with a Ca/Si ratio of approximately 1.7 [
74], which decreases with increasing replacement by SCMs. As SiO
2 replacement increases, the Ca/Si ratio of C–S–H decreases significantly, ranging from 0.67 to 2.0 [
75,
76]. Kunther et al. [
77] demonstrated that compressive strength increases with decreasing Ca/Si ratio across all samples and ages. Low Ca/Si C–S–H pastes (Ca/Si = 0.83 and 1.00) showed significant strength increases over time, whereas high Ca/Si C–S–H pastes showed only small increases. After 3 months, the compressive strength of the Ca/Si = 0.83 binder was more than double that of the highest Ca/Si ratio binder. However, the range of Ca/Si ratio in this study was between 0.83 and 1.5. Two hypotheses explain this correlation [
77]:
Low Ca/Si C–S–H hydrates have denser microstructures, leading to lower porosities and higher compressive strengths, as lower Ca/Si ratios result in higher surface areas.
The density and molar volume of the C–S–H phase are correlated with the Ca/Si ratio; lower Ca/Si ratios result in lower molar volumes and higher densities, contributing to higher strengths.
Despite the consensus among researchers regarding the impact of both Ca/Si and Al/Si ratios on the mechanical properties of cementitious composites, the nature of this relationship remains debated. Some studies [
78,
79] suggest that the values of the modulus of elasticity and creep increase with the rise in Ca/Si ratio, which might explain the higher flexural strength of CR0 compared to other samples. Conversely, other studies indicate optimal Ca/Si ratios, deviations from which negatively affect mechanical properties. For instance, Wang et al. [
80] found that the typical Ca/Si ratio, when incorporating fly ash, is around 1.4, while Garcia’s results [
81] suggest a value closer to 1.8. Wang [
80] also noted that the optimal Al/Si ratio is approximately 0.25, which is close to the values observed for the mortar samples with optimal performance obtained in this study: 0.24, 0.19, and 0.21 for FBA25, F10, and F20 samples, respectively. This ratio decreased to 0.09 and 0.05 for the F30 and FBA100 samples, respectively, and this decrease falls within the range defined by Wang et al. [
80] as 0.5 to 0.05. Similarly, the Ca/Si ratio also declined to 0.94 and 0.34 for the F30 and FBA100 samples, respectively.
It is also important not to ignore the impact of the ratios of network modifiers to network formers due to their role in modifying pore alkalinity and participating in various complex hydrate chains. In this study, a consistent decrease in the ratio of network modifiers to network formers was observed with the increased incorporation of GFBA. This decrease might have resulted from the reduction of calcium and the increase in amorphous silica in the mixtures containing GFBA. Similar to the cases of Ca/Si and Al/Si ratios, this decrease can be beneficial to mechanical performance up to a certain point, as it involves the consumption of Ca(OH)
2 in forming hydrate chains. However, a sharp decrease might lead to lower pore solution alkalinity, which hinders the dissolution of pozzolan particles. Previous studies indicate that increased alkali content accelerates cement hydration, enhancing strength development [
82,
83,
84]. This acceleration occurs because alkali cations in the fresh cement mixture’s liquid phase accelerate C3A hydration by releasing Ca
2+ cations [
84,
85,
86]. It is also known that increasing glass network modifiers boosts the alkaline medium, enhancing dissolution of glass phase activation [
87,
88]. However, some studies have shown that higher alkali content can reduce concrete compressive strength [
82,
83,
86]. This reduction is attributed to a porous microstructure and the lower strength of alkali-containing C–S–H gel in the hardened mixture [
84,
89,
90]. This highlights the need for further research to optimize the balance between glass network modifiers and formers and the conditions that control them.
Figure 14 shows the types of crystalline phases formed in the formulated mortars. As observed, six crystalline phases were detected: quartz, portlandite, calcite, calcium silicate hydrate, AFt, and monosulfate (AFm). The figure indicates an increase in the intensity of monosulfate peaks with the increase in the level of cement replacement by GFBA (
Supplementary material), especially at a 30% replacement level. This increase was due to the abundance of alumina dissolved from the GFBA, which can react with ettringite in the presence of gypsum to form monosulfate. A slight decrease in AFm peaks was also observed in the outperforming samples such as F10 and FBA25, which may be attributed to its conversion to AFt. This may be due to the chemical instability of AFm, as it may react with calcium carbonate in the presence of appropriate humidity to turn into AFt [
91]. Calcite was also observed, with its peaks primarily distributed between 50° and 27°. Its presence can be attributed to the relative instability of Ca(OH)
2, which undergoes accelerated carbonation under atmospheric conditions. The quartz peaks were nearly identical in the mortar samples CR0, FBA25, F10, F20, and F30, but their intensity was more pronounced in the mortar containing a 100% replacement level of fine aggregate with FBA.