3.1. Flow and Setting Time
shows the flow values for various CNS replacement rates for AASC paste with 5% and 10% activators. Regardless of the activator concentration, the flow value of the paste decreased with increasing CNS replacement ratio. The activation of slag increases when the concentration of activator increases from 5% to 10%. As a result, the hydration reaction rate increases and the amount of hydration reactant increases, so the viscosity of the paste increases and the flow value decreases. The mixing-water weight substitution method applied in this experiment decreases the w/b ratio as the substitution rate of CNS increases. Therefore, the concentration effect of the activator and the w/b reduction effect on the CNS substitution influence the decrease of the flow value.
Gao et al. [19
] performed a paste experiment in which nano-silica slurry (solid content of 50 wt.% and d50
of 0.12 μm) was replaced with 0, 1, 2, and 3% of the binder weight in the binders having GGBS/FA composition ratios of 70/30 and 30/70. Experimental results show that the slump flow decreases steeply as the nano-silica contents increase from 0 to 3%. This fluidity reduction has been explained by the high reactivity of nano-silica particles containing unsaturated Si-O bonds that a certain amount of water in solution can be bound around the nano-silica particles with formation of Si-OH [32
]. Therefore, the presence of nano-silica reduces the amount of mixed water required for hydration of slag and fly ash, which contributes to the decrease in the fluidity of the paste. In addition, Behfarnia and Rostami [21
] and Ramezanianpour and Moeini [34
] have reported a similar tendency of decreasing flow as the amount of nano-silica increased.
However, some researchers have also reported different trends of flow or workability properties depending on the amount of nano-silica.
Ibrahim et al. [35
] reported different flow characteristics in the flow of alkali-activated mortar using colloidal nano-silica (50% solids content, average particle size 35 nm and pH 9.5) and natural pozzolan as a binder. They reported that the flow rates were 164, 170, 172, 158 and 152 mm as the replacement rates of colloidal nano-silica increased to 0, 1, 2.5, 5.0 and 7.5%, respectively. Flow results show a slight increase with increasing colloidal nano-silica up to 2.5%, but a decrease from 5.0%. This fluidity change was noted as a slight increase in flow due to colloidal nature in the range of as low as 2.5% of the replacement rate of colloidal nano-silica. Also, the flow reduction seen over the 5.0% colloidal nano-silica range was explained by the increased demand for nano-silica mixed water with greater specific surface area.
Adak et al. [36
] was tested with colloidal nano-silica (30% solids content, pH 9.0–9.6, average particle size 4–16 μm) and a 12M activator concentration in mortar with fly ash:sand = 1:3 ratio. In the experiments, colloidal nano-silica was added at 4, 6, 8 and 10% of the fly ash weight. As a result, the slump gradually increased with increasing colloidal nano-silica. This is due to the difference in the method of adding colloidal nano-silica, not substitution.
Deb et al. [37
] reported the experimental results of substituting 0, 1, 2 and 3% of powdered nano-silica for three types of binders, fly ash, ground granulated blast furnace slag (GGBFS)andordinary Portland cement (OPC). As a result, the flow value decreased as the content of nano-silica increased in all three types of binder.
In geopolymer studies using various nano-particles, nano-particles seem to have a greater effect of reducing fluidity (or workability) and reducing setting time [27
]. Therefore, the flow or workability of the nano-silica depends on the nature of the nano-silica, the amount used, the method of substitution or addition.
shows the setting time measurement results. As the amount of CNS increased, both initial and final decreased. Also, the setting time of 10% activator samples was faster than 5%. As the substitution rate of the CNS increases, the setting time reduction in Figure 2
and the flow value decrease tendency in Figure 1
Gao et al. [19
] reported a gradual increase in setting time as nano-silica contents increased to 0, 1, 2 and 3% in alkali-activated slag-fly ash blends paste. Adak et al. [36
] increased colloidal nano-silica by 4, 6, 8, and 10% in the fly ash-based geopolymer mortar experiment by adding colloidal nano-silica. As a result, it was reported that the setting time increases with increasing amount of colloidal nano-silica.
Ibrahim et al. [35
] reported setting time results after 0, 1, 2.5, 5.0 and 7.5% substitution of colloidal nano-silica in natural pozzolan experiments activated with sodium silicate and sodium hydroxide. As a result, setting time decreased until colloidal nano-silica replacement rate to 5%. However, 7.5% colloidal nano-silica showed a slightly longer setting time than 5%. In Phoo-ngernkham et al. [39
], an experiment was carried out to add 0, 1, 2 and 3% of powdered nano-silica to FA-based geopolymer. As a result, setting time decreased with increasing amount of nano-silica.
Setting time is influenced by various factors such as nano-silica nature (solid contents ratio, average size), type (powder or colloidal), mixing method (added or substitution), binder type (fly ash, slag, metakaoline), type and concentration of activator, respectively [27
In previous AAC or geopolymer studies using nano-silica, curing time was shortened because nano-silica promotes the production and growth of reaction products [19
]. In addition, the mxing-water substitution method used in this study is an additional factor that reduces the setting time by the w/b reduction effect with increasing CNS substitution rate.
3.2. Compressive Strength
shows the compressive strength values of the samples with 5 and 10% activator. The compressive strength of the samples containing 5 and 10% activator increased at each measurement age as the CNS replacement ratio increased. As the age increased, the compressive strength increased. Figure 4
a shows the measured strength values of the samples with 5% activator. The 1-day strength values of the 0% CNS and 50% CNS samples were 13.3 MPa and 50.1 MPa, respectively. The 3-, 7-, and 28-day strengths of the 0% CNS and 50% CNS samples were 18.6 MPa, 26.2 MPa, and 34.0 MPa and 73.6 MPa, 86.5 MPa, and 94.4 MPa, respectively.
Strength improvement due to CNS substitution was observed at all measurement ages, and the measured compressive strength increased with increasing CNS replacement ratio. The compressive strength of the 10% activator samples for various CNS replacement ratios is shown in Figure 4
b. The strengths of the samples containing 0 and 50% CNS were 16.2 MPa and 77.0 MPa at 1 day, 20.9 MPa and 93.3 MPa at 3 days, 32.8 and 110.8 MPa at 7 days, and 48.0 MPa and 138.24 MPa at 28 days, respectively. Similar to the trend shown by the 5% activator samples, the compressive strength of the 10% activator samples increased with increasing CNS replacement ratio. No unusual trends were found in the variation of compressive strength by measurement age with increasing CNS replacement ratio.
The enhancement of strength due to the use of nano-particle in Portland cement has already been revealed by many previous studies [6
]. The effect of OPC-based nano-particles on the strength enhancement is known to be due to pozzolanic reaction, nucleation effect and filler effect [9
Some researchers have also reported on the effects of nano-particles on the mechanical performance of AAC or geopolymer. Some studies have reported that there is an optimal nano-silica content with the highest strength. Gao et al. [22
] performed an alkali-activated metakaolin experiment using powdered nano-silica. In the experiment, nano-silica added 0, 1, 2, and 3% of the weight of metakaolin as a binder. The flexural strength results show that 1% nano-silica is capable of high strength. The strength results of Gao et al. [22
] suggest that the content of nano-silica with optimal strength is present. Gao et al. [19
] studies showed strength results for nano-silica slurry and alkali-activated slag/fly ash cement. Gao et al. [19
] replaced 0, 1, 2 and 3 of nano-silica with respect to the weight of slag/fly ash as a binder. As a result, 2% nano-silica sample showed the highest compressive strength at all measurement ages. Wang et al. [40
] performed an experiment using powdered nano-silica with two kinds of average particle sizes (30 nm, 15 nm) in AASC. 30 nm nano-silica was 0.5 to 6.0%, and 15 nm nano-silica was 0.5 to 2.5%. As a result, the maximum compressive strength was measured at 3.0% for 30 nm nano-silica and at 2.0% for 15 nm nano-silica. Experimental results of Wang et al. [40
] confirmed that the content of nano-silica, which gives the highest strength depending on the particle size of nano-silica, is affected. Ibrahim et al. [35
] performed an experiment in which the colloidal nano-silica was replaced with 0, 1, 2.5, 5, and 7.5% of the binder weight in the alkali-activated natural pozzolan experiment. As a result, 5% colloidal nano-silica showed the highest compressive strength. Adak et al. [36
] carried out an experiment in which colloidal nano-silica was added at 4%, 6%, 8% and 10% in 12M concentration activator and fly ash-based geopolymer. Experimental results indicated that the highest compressive strength was 6% colloidal nano-silica. In addition, Deb et al. [41
] performed experiments using fly ash-only-based geopolymer, slag blended fly ash-based geopolymer and powdered nano-silica. Nano-silica was substituted for 0, 0.5, 1, 1.5, 2, 2.5, and 3% of binder weight. Fly ash-only-based geopolymer and slag-blended fly ash-based geopolymer showed the highest compressive strength in 2% nano-silica. Deb et al. [37
] reported that fly ash-based geopolymer experiments with 0, 1, 2, and 3% substitution of powdered nano-silica were performed, and the maximum strength was measured on a 2% nano-silica sample. Behfarnia and Rostami [21
] reported that the highest strength was obtained at 3% in AAC experiments where powdered nano-silica was replaced with 0, 0.5, 1, 3, and 5% of slag weight.
However, other researchers reported that as the content of nano-silica increases, the compressive strength increases. Shahrajabian and Behfarnia [42
] reported that compressive strength increased with increasing amount of nano-silica in the AASC experiment using 0, 1, 2, and 3% of powdered nano-silica, and the maximum strength was 3%. In a study of Yang et al. [43
] in which 0.5% of nano-TiO2
was mixed with alkali-activated slag pastes, the mixing of nano-TiO2
exhibited an increase in compressive strength and flexural strength. Long et al. [20
] showed that the 2% replacement sample showed the highest compressive strength and flexural strength in an alkali-activated slag cement with 0 and 2% replacement of colloidal nano-silica. Prakasam et al. [26
] produced a geopolymer of binder with slag:fly ash = 50:50 using powdered nano-silica at 0, 0.5, and 1.0%. strength test showed that 2.0% nano-silica had the highest compressive strength at all measurement times.
The filler effect and nucleation effect have been mentioned as the causes of the strength enhancement of nano-particles in AAC/geopolymer [19
]. In AAC/geopolymer, the pozzolanic reaction was excluded from the strength enhancement effects of OPC-based nano-particles. This is because there is no portlandite required for pozzolanic reaction in the hydration reaction of AAC/geopolymer [30
The strength characteristics of AAC or geopolymer using nano-silica are affected by various factors, and the variation of the strength improvement effect is also great [27
]. However, the experimental factors and the mixing-water weight substitution method considered in this study tend to increase the compressive strength with increasing CNS replacement ratio. This is because the effect of increasing the strength of nano-silica already known in previous studies and the effect of w/b ratio reduction by mixing-water weight substitution method are also affected. It was also found that there was no decrease in compressive strength during replacement of the CNS up to 50%. This is because the mixing-water weight substitution method of the CNS improves the dispersibility by reducing the aggregation or aggregation of the nano-silica particles. Therefore, it was confirmed that the weight substitution method of mixing-water by CNS is an effective mixing method for strength improvement.
calculates the compressive strength increase rate of CNS substituted samples for 0% CNS. Table 3
shows that the increase rate of compressive strength increases with the substitution rate of CNS regardless of the concentration of activator. In addition, the 1-day compressive strength increase rate of the measurement period was the largest increase rate compared with the increase rate in the remaining ages. This means that the strength enhancement effect of CNS appears to be highest within 1 day of the initial stage of hydration reaction.
In the study of OPC, the hydration reaction of nano-silica at the initial hydration stage and the effect of improving the strength at the early ages have already been reported. In particular, colloidal nano-silica has been reported to form C-S-H gel in the initial hydration stage before 12 hours [44
]. The effect of accelerating the initial hydration of CNS in OPC is similar in alkali-activated slag paste using CNS considered in this study. As shown in Table 3
, the highest compressive strength increase rate was 1-day. As the age increases, the rate of increase of the compressive strength according to the substitution rate of CNS gradually decreases. This indicates that the strength improvement effect of CNS is highest at the initial stage of hydration reaction and gradually decreases with time. Therefore, in order to improve the strength of CNS-substituted AASC, the initial hydration reaction step is important. To this end, the mixing method should be adjusted so that a sufficiently even dispersion of the nano-silica particles occurs during the initial hydration stage.
3.3. Hydration Products
shows the XRD analysis of hydration reactants at 1-day and 28-day for 0, 20, and 50% CNS samples. The major reactants were C-S-H gel, calcite, hydratalcite, monosulfate and quartz. Regardless of the concentration of activator, akermanite, which was observed in 0% CNS, decreased with increasing CNS replacement rates of 20 and 50%. Monosulfate and hydrotalcite, which were observed in 0% CNS, also decreased gradually and were not observed in 50% CNS samples. However, C-S-H gels, known as the major hydration reactants of AASC, show complex and irregular changes without a consistent tendency as the substitution rate of the CNS increases. C-A-S-H gels play a central role in improving the mechanical properties and durability of AASC. Therefore, the use of CNS promoted the hydration reaction of slag and expected to increase the amount of C-S-H gel. However, the XRD analysis does not clearly show the tendency to change the C-S-H gel due to the substitution of the CNS. The C-S-H gels are amorphous and cause difficulty in understanding the exact trend because of the overlap of calcite and peak.
Some researchers reported that nano-silica had little effect on the formation of C-S-H gel. Long et al. [20
] mentioned similar trends of C-S-H gel peaks in XRD patterns of 0 and 2% nano-silica samples in AASC. It is reported that the mixing of nano-silica does not contribute to the formation of another polymerization product in AASC. Assaedi et al. [31
] mentioned a small change in crystalline and amorphous contents in geopolymers mixed with 0.5, 1.0, 2.0, and 3.0 wt.% nano-silica.
As seen from the results of Long et al. [20
] and Assaedi et al. [31
], previous studies have not found a clear contribution to the change of C-S-H gel by nano-silica particles in AASC. This is because it is difficult to quantitatively evaluate the effect of nano-silica particles on nucleation and filler effects. Therefore, formation of hydration reactants such as C-S-H gel is insufficient to explain the cause of improvement of strength by nano-silica particles.
From the results of XRD analysis in Figure 4
, the effect of CNS on the formation of C-S-H gel is small or overlapping with the calcite peak, making it difficult to grasp the clear tendency. To evaluate the effect of CNS on the formation of C-S-H gel, it is concluded that XRD and other analyses should be performed in parallel.
3.4. Pore Structure
shows the MIP results for analyzing the void structure of the 0, 20, and 50% CNS samples of the 5 and 10% activator. Comparing the porosity distributions of the 5 and 10% activator samples, the 10% activator showed a decrease in pore size and volume. This is because, as the concentration of the activator increases, the hydration reaction of the slag is promoted and the amount of hydration reaction product is increased to form a dense matrix. Also, no significant change in pore diameter or volume was observed when comparing the pore distribution after 1 day and 28 days, regardless of the concentration of activator.
In the pore structure analysis of Gao et al. [22
], 1% nano-silica sample showed the lowest porosity in the experiments in which 0, 1, 2, and 3% of powdered nano-silica were replaced. The highest compressive strength was also measured on a 1% nano-silica sample. Wang et al. [40
], which reported the results of AASC experiments in which nano-silica was replaced by 2% and 3% of the weight of the binder, also showed a decrease in pore size and volume due to substitution of nano-silica. In a study of Yang et al. [43
] for AAS paste with 0.5% nano-TiO2
reduced the total porosity and the amount of pores larger than 25 nm. Long et al. [20
] reported that pore size and volume decreased in AASC experiments using 0 and 2% colloidal nano-silica.
shows the total porosity of 0, 20, and 50% CNS. As the substitution rate of CNS increased, total porosity decreased. Total porosity at 5 and 10% activator samples was slightly lower than 1-day at 28-day. The total porosity of the 10% activator sample was also lower than that of the 5% activator. The difference in total porosity between 1-day and 28-day was small. This means that the effect of densification of the matrix by the CNS after 1-day becomes insignificant. This supports the result that the above-mentioned maximum compressive strength increase rate is 1-day.
shows the composition ratio of pores of 10,000 nm or less [49
]. The medium capillary pore gradually decreased and the gel pore increased as the CNS replacement ratio increased in the 5% activator sample. In the 28-day sample, the medium capillary pore decreased and the gel pore increased as compared to the 1-day. The 10% activator sample showed similar tendency to 5% activator. As the substitution rate and age of the CNS increase, the medium capillary pore decreases and the gel pore increases. It was also found that 10% activator samples had more gel pores than 5% activator. Therefore, the substitution of the CNS with increasing concentration of the activator decreases the size and amount of pores by making the hydration reaction matrix dense. This is due to nucleation and filler effects [24
]. In Table 4
, no significant difference in the ratio of pore sizes between 1-day and 28-day was observed, regardless of the concentration of activator. This means that the effect on the pore structure by the hydration reaction after 1-day is relatively small. Therefore, the compressive strength increase rate described in Table 3
was highest at 1-day and lowest at 28-day.
3.5. Thermal Analysis
shows the thermal analysis results for 1-day and 28-day hydration reactants. Figure 8
a shows the DTG results for 1-day and 28-day hydration reactants of 0, 20, and 50% CNS samples of 5% activator. The weight loss rate seen at 50–200 °C is evaporation of water or C-S-H gel [50
]. In addition, the weight loss rate observed at 130–200 °C is monosulfate [49
]. The weight loss rate at 330–400 °C represents hydrotalcite-like phase [50
], and the weight change between 665 and 800 °C means calcite [50
]. Figure 8
b shows an enlarged view of the temperature range of less than 300 °C to confirm the formation of AASC’s representative hydration reactants, C-S-H gel. The weight loss rate of the C-S-H gel at 28-day was slightly increased compared with that at 1-day, and the difference in the weight loss rates was not significant. This means that the CNS reactions mostly to the initial hydration stage of the AASC. As a result, the influence of CNS on the formation of hydration products after 1-day is estimated to be insignificant. Also, as the substitution ratio of CNS increases, the weight loss rate of C-S-H gel increases. Therefore, it shows the largest weight loss rate of 50% CNS.
c shows the thermal analysis results of 10% activator. The results are similar to the thermal analysis results of the 5% activator samples. That is, as the substitution ratio of CNS increases, the weight loss rate of C-S-H gel in the range of 50–200 °C becomes larger. The comparison of the weight loss rate of C-S-H gel shown in Figure 8
d shows that the formation amount of C-S-H gel of 28-day is larger than that of 1-day. This tendency is clear as the substitution rate of the CNS increases. Thermal analysis clearly shows the relationship between the formation of C-S-H gel and the rate of CNS replacement, which was not evident in the XRD analysis. This confirms that the substitution of the CNS affects the formation of C-S-H gel. In Figure 8
, the difference in weight loss between 1-day and 28-day C-S-H gel was not significant in either the 5 and 10% activator samples. As a result, the effect of substitution of CNS on the formation of C-S-H gel is decreased after 1-day. Therefore, the XRD analysis described above shows a calcite-C-S-H gel peak with little change and tendency.
In the 5% and 10% activator, the weight loss in the 50–200 °C region increases as the substitution rate of the CNS increases. This is consistent with the tendency of the weight loss in the 50–200 °C region to increase with increasing contents of the nano-silica particles reported in previous studies [19
]. However, a clear analysis of the thermal properties of AAC/gepolymer with nano-silica particles is still being attempted. In particular, the cause of the weight loss shown in the thermal analysis results in a temperature range lower than 200 °C is unclear. This is because the formation of hydration products by nano-silica particles and the effect of simple evaporation of water act simultaneously [19
]. In Figure 5
, the change in the C-S-H peak was negligible despite the increasing rate of substitution of CNS. Nevertheless, the pore size decreased and the compressive strength increased. This is due to the effect of promoting the formation of C-S-H gel by nano-silica and the filler effect. The increase in nano-silica contents requires a large amount of bound water [8
]. Bound water by nano-silica particles evaporates at temperatures below 200 °C, causing weight loss. Therefore, weight loss in the temperature range of less than 200 °C shown in Figure 8
is due to the formation of C-S-H gel and increase of bound water content with increasing nano-silica particles. As a result, weight loss at a temperature of less than 200 °C shows a clear increase as the substitution rate of the CNS increases. However, the clear distinction and mechanism of these two effects still need further research.
The loss observed in the region of 350 to 550 °C in Figure 8
d is estimated to be due to the hydrotalcite phase shown in the XRD analysis of Figure 4
. In the XRD results of Figure 5
, the 10% activator sample has the hydrotalcite peak that is relatively larger than the 5% activator. As a result, in Figure 8
a of the 5% activator, the change in weight loss in the 350 to 550 °C region is negligible. In Figure 5
c,d, the hydrotalcite peak decreased as the substitution ratio of CNS increased to 0, 20, and 50%. In addition, the peak intensity of hydratalcite at 28-day is lower than that at 1-day. In Figure 8
c, 1-day samples of 0% CNS and 20% CNS clearly show weight loss in the 350 to 550 °C range. In addition, 28-day samples of 0% CNS and 20% CNS show small weight loss. However, 50% CNS samples showed little weight loss of hydrotalcite. Therefore, the tendency of hydrotalcite in XRD results was confirmed by thermal analysis (TG/DTG).
The weight loss in the region of 665–800 °C shown in Figure 8
a,c shows complex characteristics. In Figure 8
a, 1-day samples of 0% CNS and 20% CNS show the greatest weight loss. Greater weight loss than 28-day samples may be due to unexpected exposure to CO2
during sample storage, powder manufacture, and measurement. In Figure 8
c, the weight loss rate decreases as the substitution ratio of CNS increases. The definite cause of this is difficult to confirm in the scope of this study. The reason is that unlike OPC, AAC has no portlandite [30
], so the carbonate process is affected by various factors. To date, the carbonation of AAC has been associated with the pH reduction of the pore solution, the precipitation of Na-rich carbonates and the decalcification of Ca-rich phases (C-S-H gel) [56
]. The CaCO3
, which is formed as a result of carbonation, is classified into crystalline phases and an amorphous phase [57
]. Therefore, CaCO3
, which is distinguished by calcite, vaterite and aragonite, may be different from XRD and TG/DTG [57
]. The weight loss in the region of 665–800 °C observed in Figure 8
is estimated to be the weight loss due to the carbonate, but the analysis of causes and trends is not clear in the scope of this study.