Influence of Na2O Content and Ms (SiO2/Na2O) of Alkaline Activator on Workability and Setting of Alkali-Activated Slag Paste

The performance of alkali-activated slag (AAS) paste using activators of strong alkali components is affected by the type, composition, and dosage of the alkaline activators. Promoting the reaction of ground granulated blast furnace slag (GGBFS) by alkaline activators can produce high-strength AAS concrete, but the workability might be drastically reduced. This study is aimed to experimentally investigate the heat release, workability, and setting time of AAS pastes and the compressive strength of AAS mortars considering the Na2O content and the ratio of Na2O to SiO2 (Ms) of binary alkaline activators blended with sodium hydroxide and sodium silicate. The test results indicated that the AAS mortars exhibited a high strength of 25 MPa at 24 h, even at ambient temperature, even though the pastes with an Na2O content of ≥6% and an Ms of ≥1.0 exhibited an abrupt decrease in flowability and rapid setting.


Introduction
Environmental imperatives such as the reduction of CO 2 and conservation of natural resources are becoming issues worldwide. However, because the cement industry emits large amounts of CO 2 and consumes much energy in the supply of raw materials and production processes, efforts are needed to reduce CO 2 emission and energy consumption by replacing cement with supplementary cementitious materials (SCMs) such as ground granulated blast furnace slag (GGBFS) and fly ash (FA). Thus, in order to totally replace cement with GGBFS and FA attempts have been made to create zero-cement concrete by the alkali activation of GGBFS and FA [1][2][3][4][5].
Alkali-activated slag (AAS) concrete, which is produced by using an alkali activator on GGBFS, is a typical zero cement concrete. As GGBFS is calcium aluminosilicate vitreous, it has strong latent hydraulic properties. However, alkaline (pH = 12) activators are needed to stimulate GGBFS. After the GGBFS is stimulated, Si 4+ , Al 3+ , Ca 2+ , Mg 2+ , and Na + ions are eluted from GGBFS and the reaction of GGBFS proceeds. It should be noted that GGBFS activated in this way has very fast reaction rates due to chemical ionic reactions [6]. Alkaline activators largely impact the properties of AAS concrete and thus, their characteristics such as the alkali concentration, the dosage, and the water-to-solid ratio should be examined prior to the mixture design of AAS concrete [7,8]. In general, soluble alkali or alkali salts can be used as alkaline activators. When considering the strength development and economic efficiency of AAS concrete, the most commonly used activators are Na 2 SiO 3 (sodium silicate) and NaOH (sodium hydroxide). A better understanding of the effects of such alkaline activators on the reaction mechanisms of alkali-activated slag could indicate ways to optimize the use of alkaline activators.
Recently, research has been conducted to improve the performance of AAS concrete by blending two types of alkaline activators [9]. When such a binary alkaline activator is used, the characteristics possessed by each alkaline activator can be complementarily utilized [10,11]. NaOH promoted initial reaction of AAS, and Na 2 SiO 3 was effective in increasing the strength of AAS concrete. Chang [12] evaluated the setting characteristics of silica-activated pastes depending on the type of alkaline activator used and concluded that setting time significantly reduced as the amount of alkaline activator increased. Zuo et al. [13] showed that though the reaction degree increased as the amount of alkaline activator increased, the reaction degree no longer increased when the amount of alkaline activator was greater than 6 wt.% of cement. Krizan and Zivanovic [14] reported that high strength could be expected when the Ms (SiO 2 /Na 2 O) were in the range of 0.6-1.5. As described above, as the amount of alkaline activator increases, the reaction of GGBFS increases and the strength of AAS concrete also increases. Therefore, it is necessary to determine an appropriate range for the usage amount of the activator for the mix design of AAS concrete. In particular, the effects of Na 2 O and SiO 2 concentrations in alkaline activators on the strength development of AAS concrete should be examined in order to effectively use the alkaline activators.
AAS concrete, which stimulates GGBFS using alkaline activators, rapidly increases in strength in the early ages and exhibits high strength of 40 MPa or more at 28 days [15,16]. However, because the initial reaction rate of GGBFS by alkaline activators is extremely high, several problems with AAS, such as the flowability being lost initially or rapid setting occurring, must be improved for it to be used in a practical setting [17][18][19]. To address this, the effects of the alkaline activator on the fresh properties of AAS concrete such as workability and setting should be examined, even though the AAS activation mechanism at early ages has not been clearly identified.
In this study, in order to identify the characteristics of AAS pastes, a binary alkaline activator that was a blend of NaOH and Na 2 SiO 3 was used. Nine types of AAS paste mixture, varying by their Na 2 O content and the ratio of Na 2 O to SiO 2 (Ms), were tested. First, the reactivity of the AAS was evaluated for 24 h through the calorimetric measurement of the AAS pastes, and the workability and setting time were measured. Furthermore, to quantitatively analyze the instant at which flowability is lost and setting occurs in the AAS pastes, the viscosity and ultrasonic pulse velocity of the pastes were measured. Lastly, the compressive strength of the AAS mortars was measured up to 24 h. Consequently, it was found that the workability and setting of AAS pastes and the initial strength development of the AAS mortars were significantly affected by Na 2 O content and Ms.

Materials and Alkaline Activator
GGBFS obtained from POSCO (Pohang, Korea) was used as a binder. The chemical composition of the GGBFS determined with X-ray fluorescence (XRF, BRUKER S8 Tiger, Billerica, MA, USA) is given in Table 1. The GGBFS was composed of 44.0% CaO, 33.7% SiO 2 , 13.8% Al 2 O 3 , and 5.2% MgO and thus, its basicity coefficient (Kb = (CaO + MgO)/(SiO 2 + Al 2 O 3 )) was 1.04, which is similar to the neutral value of 1.0 for the ideal alkali activation [12]. The hydration modulus (HM = (CaO + MgO + Al 2 O 3 )/SiO 2 ) was 1.87, which was 33.5% higher than the required value of 1.4 for sound hydration properties of GGBFS [12,19]. Table 2 shows the physical properties of GGBFS. The density and Blaine air permeability of the GGBFS were 2.90 g/cm 3 and 4253 cm 2 /g, respectively. Figure 1 shows the particle size distribution of GGBFS measured with a laser particle size analyzer (PSA, BECKMAN COULTER LS 13 320, Brea, CA, USA). The medium particle size (d 50 ), which represents the particle size of a cluster of particles, was 10.2 µm, and d 10 and d 90 were 1.4 µm and 33.2 µm, respectively. The dissolution of GGBFS is dominated by small particles. That is, particles >20 µm react slowly, while particles <2 µm react completely after 24 h in alkali-activated binders [20,21].
The physical properties of the fine aggregates that were used and the mortar production were measured complying with ASTM C 128-15 [22] and ASTM C 33-13 [23]. The fine aggregates were the crushed sands of granite, and their density and absorption rate were 2.62 g/cm 3 and 1.04%, respectively. Figure 2 shows the particle size distribution of fine aggregates. The max. size of the fine aggregate was 4.76 mm, and the fineness modulus was 2.74.
Two types of alkaline activator, sodium hydroxide (98% purity, Duksan, Ansan, Korea) and sodium silicate (Type 3 industrial water glass, Na 2 SiO 3 , Ganachem, Ulsan, Korea) were used to activate the GGBFS. The chemical composition of sodium silicate solution was SiO 2 = 28.3 wt.%, Na 2 O = 9.3 wt.%, and H 2 O = 62.4 wt.%. A 2 M sodium hydroxide solution was prepared by mixing with deionized water. Furthermore, because NaOH would release heat when mixed with water, in order to cool down to ambient temperature, the alkaline activator was stored in a chamber at 20 • C and the relative humidity (RH) of 60% for 24 h prior to casting AAS pastes. The pH of the activators was ranged 13.2-13.6, as measured using a pH meter (HORIBA, LAQUA 74bW, Kyoto, Japan). respectively. Figure 2 shows the particle size distribution of fine aggregates. The max. size of the fine aggregate was 4.76 mm, and the fineness modulus was 2.74. Two types of alkaline activator, sodium hydroxide (98% purity, Duksan, Ansan, Korea) and sodium silicate (Type 3 industrial water glass, Na2SiO3, Ganachem, Ulsan, Korea) were used to activate the GGBFS. The chemical composition of sodium silicate solution was SiO2 = 28.3 wt.%, Na2O = 9.3 wt.%, and H2O = 62.4 wt.%. A 2 M sodium hydroxide solution was prepared by mixing with deionized water. Furthermore, because NaOH would release heat when mixed with water, in order to cool down to ambient temperature, the alkaline activator was stored in a chamber at 20 °C and the relative humidity (RH) of 60% for 24 h prior to casting AAS pastes. The pH of the activators was ranged 13.2-13.6, as measured using a pH meter (HORIBA, LAQUA 74bW, Kyoto, Japan).    respectively. Figure 2 shows the particle size distribution of fine aggregates. The max. size of the fine aggregate was 4.76 mm, and the fineness modulus was 2.74. Two types of alkaline activator, sodium hydroxide (98% purity, Duksan, Ansan, Korea) and sodium silicate (Type 3 industrial water glass, Na2SiO3, Ganachem, Ulsan, Korea) were used to activate the GGBFS. The chemical composition of sodium silicate solution was SiO2 = 28.3 wt.%, Na2O = 9.3 wt.%, and H2O = 62.4 wt.%. A 2 M sodium hydroxide solution was prepared by mixing with deionized water. Furthermore, because NaOH would release heat when mixed with water, in order to cool down to ambient temperature, the alkaline activator was stored in a chamber at 20 °C and the relative humidity (RH) of 60% for 24 h prior to casting AAS pastes. The pH of the activators was ranged 13.2-13.6, as measured using a pH meter (HORIBA, LAQUA 74bW, Kyoto, Japan).    Table 3 shows the mixture design of the AAS pastes used to evaluate the workability and setting of the AAS pastes. In this study, a liquid-to-binder ratio (l/b, binder = GGBFS content) was fixed as 0.42 and the Na 2 O content and Ms of alkaline activator were varied. It should be noted that the water content of Na 2 SiO 3 was included in the required water content. Also, NaOH was added as an alkali activator in order to adjust Ms according to the dosage of Na 2 SiO 3 . The Na 2 O content was set to 4%, 6%, and 8% according to the weight of the GGBFS, and the three levels of Ms were considered at 0.75, 1.00, and 1.25. Thus, nine AAS paste mixtures were designed, and the amount of alkaline activators for each mixture was determined, as shown as Table 3. In the mixture section in the table, the first and second components represent Na 2 O content and Ms, respectively. To produce the AAS mortar mixture, twice as much fine aggregate as GGBFS was added to the AAS paste mixture. Table 3. Mixture design of alkali-activated slag (AAS) pastes with respect to 100 g of binder.

Calorimetric Measurement
GGBFS generates heat during reacting with alkaline activator. To measure the heat, GGBFS and water were stored for 24 h in the calorimeter equipment at 20 • C. The AAS paste mixtures were prepared by mixing the GGBFS, water, and alkaline activators using an internal propeller for 5 min. The mixture was placed into the calorimeter (Tokyo-Riko, Three-point multi-purpose conduction calorimeter, Tokyo, Japan). The temperature range of the equipment was 5-60 • C, and the temperature stability was 5 × 10 −3 • C/5 • C. The amount of heat release was measured immediately after mixing the AAS paste up to 24 h.

Workability
The workability of AAS paste mixtures was tested using the mini slump test method. Figure 3 shows a mini slump setup and flow measurement of AAS paste. The mold for the mini slump test was made of brass, and its height was 50 mm, the top opening was 70 mm, and the bottom opening was 100 mm. After filling the mold with the paste mixture, the mold was lifted and the spread length of the mixture was measured. The mini slump of the AAS paste was tested a total 13 times for each paste, from 0 min through 60 min at 5 min intervals after casting the paste complying with ASTM C 1437-15 [24]. In order to eliminate the possibility of measurement errors in the test due to moisture evaporation from the surface of the paste, the paste was remixed for 30 s before mini slump test.

Setting Time
The Vicat test (per ASTM C 191-18a [25]) test was carried out to measure the setting time of the AAS pastes inside a room chamber at 23 °C and 50% (RH). The Vicat initial time of setting is defined as the elapsed time between the initial contact of cement and water and the penetration of 25 mm. The Vicat final time of setting end point is defined as the first penetration measurement that did not mark the specimen surface with a complete circular impression.

Rheology and Ultrasonic Pulse Velocity (UPV)
The rheology and UPV tests were adopted to evaluate the timing at which the AAS paste lost its flowability, by measuring the viscosity and ultrasonic pulse velocity of the AAS paste according to the elapsed time. The rheology test measured viscosity through a paste viscometer using the friction resistance between the spindle and the paste rotating at a speed of 5 rpm. To determine the moment that flowability was lost in the AAS paste, the time to reach maximum measurable viscosity was recorded. The UPV test was carried out by placing the oscillator and receiver on acrylic boxes (thickness: 10 mm; ultrasonic velocity of acryl: 2740 m/s) placed 20 mm apart, filling the container with paste, and measuring the transmission time of ultrasonic waves through the oscillator and receiver at intervals of 30 s with the measurement frequency of 5 kHz.

Compressive Strength
To investigate the influence of Na2O content and Ms on the compressive strength development of AAS mortars up to 24 h, uniaxial compression test was conducted. A mortar mixer was used for mixing 2-liter batches. After premixing GGBFS and fine aggregates for 30 s, the alkaline activator and water were added at low speed. Mixing continued at low speed for 90 s and at high speed for 1 min. Cube mortar specimens (50 × 50 × 50 mm 3 ) were prepared complying with ASTM C109-16a [26] and cured in a chamber at 20 ± 2 °C and the relative humidity (RH) of 60 ± 2%, for curing at the environmental condition similar to general ambient condition. Compressive strength of the mortar specimens was measured at 12 h, 18 h, and 24 h after casting the AAS paste. The compressive strength was obtained as an average value of three AAS mortar specimens. Figure 4 shows the calorimetric curves for the rate of heat release according to Na2O content. The AAS pastes, excluding the paste with 4% Na2O, had two peaks of heat [13]. When Na2O content was greater than 6% of binder weight, the first peak of heat occurred at 30-60 min, which resulted mainly from the wetting and partial dissolution of GGBFS by NaOH and partly from the reaction of GGBFS by Na2SiO3. The second peak was observed between 6 h and 12 h, which corresponds mainly

Setting Time
The Vicat test (per ASTM C 191-18a [25]) test was carried out to measure the setting time of the AAS pastes inside a room chamber at 23 • C and 50% (RH). The Vicat initial time of setting is defined as the elapsed time between the initial contact of cement and water and the penetration of 25 mm. The Vicat final time of setting end point is defined as the first penetration measurement that did not mark the specimen surface with a complete circular impression.

Rheology and Ultrasonic Pulse Velocity (UPV)
The rheology and UPV tests were adopted to evaluate the timing at which the AAS paste lost its flowability, by measuring the viscosity and ultrasonic pulse velocity of the AAS paste according to the elapsed time. The rheology test measured viscosity through a paste viscometer using the friction resistance between the spindle and the paste rotating at a speed of 5 rpm. To determine the moment that flowability was lost in the AAS paste, the time to reach maximum measurable viscosity was recorded. The UPV test was carried out by placing the oscillator and receiver on acrylic boxes (thickness: 10 mm; ultrasonic velocity of acryl: 2740 m/s) placed 20 mm apart, filling the container with paste, and measuring the transmission time of ultrasonic waves through the oscillator and receiver at intervals of 30 s with the measurement frequency of 5 kHz.

Compressive Strength
To investigate the influence of Na 2 O content and Ms on the compressive strength development of AAS mortars up to 24 h, uniaxial compression test was conducted. A mortar mixer was used for mixing 2-liter batches. After premixing GGBFS and fine aggregates for 30 s, the alkaline activator and water were added at low speed. Mixing continued at low speed for 90 s and at high speed for 1 min. Cube mortar specimens (50 × 50 × 50 mm 3 ) were prepared complying with ASTM C109-16a [26] and cured in a chamber at 20 ± 2 • C and the relative humidity (RH) of 60 ± 2%, for curing at the environmental condition similar to general ambient condition. Compressive strength of the mortar specimens was measured at 12 h, 18 h, and 24 h after casting the AAS paste. The compressive strength was obtained as an average value of three AAS mortar specimens. Figure 4 shows the calorimetric curves for the rate of heat release according to Na 2 O content. The AAS pastes, excluding the paste with 4% Na 2 O, had two peaks of heat [13]. When Na 2 O content was greater than 6% of binder weight, the first peak of heat occurred at 30-60 min, which resulted mainly from the wetting and partial dissolution of GGBFS by NaOH and partly from the reaction of GGBFS by Na 2 SiO 3 . The second peak was observed between 6 h and 12 h, which corresponds mainly to the formation of reaction products [27]. That is, when Na 2 O content was fixed to be 6% or 8%, the first peak of heat increased as Ms increased, in other words, SiO 2 content increased. In particular, the paste with 6% Na 2 O content showed a relatively high first peak and a clear secondary peak. Ravikumar and Neithalath [28] reported a similar result, stating that at constant Na 2 O content, the Ms was a main factor in increasing both the speed and the magnitude of the first peak of heat. For the pastes with alkaline activators of 4% Na 2 O content, the first peak of heat was relatively low and the secondary peak occurred after the induction since the amount of alkaline activators was less. This is a similar tendency to the rate of heat release curve when only Na 2 SiO 3 was used [27]. to the formation of reaction products [27]. That is, when Na2O content was fixed to be 6% or 8%, the first peak of heat increased as Ms increased, in other words, SiO2 content increased. In particular, the paste with 6% Na2O content showed a relatively high first peak and a clear secondary peak. Ravikumar and Neithalath [28] reported a similar result, stating that at constant Na2O content, the Ms was a main factor in increasing both the speed and the magnitude of the first peak of heat. For the pastes with alkaline activators of 4% Na2O content, the first peak of heat was relatively low and the secondary peak occurred after the induction since the amount of alkaline activators was less. This is a similar tendency to the rate of heat release curve when only Na2SiO3 was used [27].  Figure 5 shows the mini slump test results for AAS pastes with a l/b of 0.42. The paste without an alkaline activator demonstrated an initial flow of 182 mm, while the pastes with alkaline activators showed an initial flow of 205-350 mm. All of the AAS pastes with alkaline activators of 6% and 8% Na2O content lost their flowability within 15 min. The flowability decreased more rapidly as Ms increased; specifically, for the 8-1.25 paste, with the highest Na2O content and Ms, hardening occurred just after 5 min after mixing. As mentioned in Section 3.1, this is because the activation and reaction of GGBFS by the alkali activators occurred rapidly at the early ages. Once the AAS paste lost its flowability, the pastes began to harden quickly. These tendencies appeared as Ms increased because more reaction products of the GGBFS were produced by the sufficient supply of SiO2 [14].

Workability
The AAS paste with alkaline activators of 4% Na2O content maintained a flow of >200 mm for 1 h after mixing. Similarly, Puertas et al. [29] reported that the analysis of the flow performance of AAS pastes with alkaline activators of less than 5% Na2O content showed that when Ms was lower than 0.8, the flowability of the paste was maintained for 60 min.  Figure 5 shows the mini slump test results for AAS pastes with a l/b of 0.42. The paste without an alkaline activator demonstrated an initial flow of 182 mm, while the pastes with alkaline activators showed an initial flow of 205-350 mm. All of the AAS pastes with alkaline activators of 6% and 8% Na 2 O content lost their flowability within 15 min. The flowability decreased more rapidly as Ms increased; specifically, for the 8-1.25 paste, with the highest Na 2 O content and Ms, hardening occurred just after 5 min after mixing. As mentioned in Section 3.1, this is because the activation and reaction of GGBFS by the alkali activators occurred rapidly at the early ages. Once the AAS paste lost its flowability, the pastes began to harden quickly. These tendencies appeared as Ms increased because more reaction products of the GGBFS were produced by the sufficient supply of SiO 2 [14].

Workability
The AAS paste with alkaline activators of 4% Na 2 O content maintained a flow of >200 mm for 1 h after mixing. Similarly, Puertas et al. [29] reported that the analysis of the flow performance of AAS pastes with alkaline activators of less than 5% Na 2 O content showed that when Ms was lower than 0.8, the flowability of the paste was maintained for 60 min.  It was also observed that the higher the Ms, the lower the initial flow of the AAS paste. This is because as the Ms increased, workability was reduced due to the increase of the reaction products of the GGBFS [14,30,31]. However, the times at which the AAS pastes lost their workability were nearly identical regardless of the Ms.  It was also observed that the higher the Ms, the lower the initial flow of the AAS paste. This is because as the Ms increased, workability was reduced due to the increase of the reaction products of the GGBFS [14,30,31]. However, the times at which the AAS pastes lost their workability were nearly identical regardless of the Ms. It was also observed that the higher the Ms, the lower the initial flow of the AAS paste. This is because as the Ms increased, workability was reduced due to the increase of the reaction products of the GGBFS [14,30,31]. However, the times at which the AAS pastes lost their workability were nearly identical regardless of the Ms. Table 4 shows the Vicat test results of the AAS pastes on the initial and final setting time. The setting of the AAS paste was accelerated by the increase of activation and reaction of the GGBFS when the dosage of alkaline increased. Similarly, Chang [12] reported that the setting time of the AAS pastes correlated with the sum of the Na 2 O (from NaOH and Na 2 SiO 3 ) and SiO 2 (from Na 2 SiO 3 ) concentrations of the alkali activator. For AAS pastes with less than 8 g of alkaline (per 100 g of GGBFS), setting time significantly increased as the dosage of alkaline was low. For AAS pastes with more than 10 g of dosage of alkaline, setting was accelerated; initial setting occurred within 1 h, as shown in Figure 7. When the dosage of alkaline increased to more than 14 g, final setting occurred within 30 min.  Table 4 shows the Vicat test results of the AAS pastes on the initial and final setting time. The setting of the AAS paste was accelerated by the increase of activation and reaction of the GGBFS when the dosage of alkaline increased. Similarly, Chang [12] reported that the setting time of the AAS pastes correlated with the sum of the Na2O (from NaOH and Na2SiO3) and SiO2 (from Na2SiO3) concentrations of the alkali activator. For AAS pastes with less than 8 g of alkaline (per 100 g of GGBFS), setting time significantly increased as the dosage of alkaline was low. For AAS pastes with more than 10 g of dosage of alkaline, setting was accelerated; initial setting occurred within 1 h, as shown in Figure 7. When the dosage of alkaline increased to more than 14 g, final setting occurred within 30 min.   Figure 8 shows the rheology test results of the AAS pastes with Na2O content and Ms. The viscosity measurement limit (VML) refers to the point at which flowability is lost in the fluid state. As described in Section 3.2, the flow of the paste showed a tendency for the flowability to decrease abruptly immediately before the loss of flowability, and the shear stress rapidly increased 3 to 5 min before reaching VML. As shown in Table 4, the min. and max. VML times of the AAS paste with ≥6% Na2O content were 540 s for 8-1.25 and 1040 s for 6-0.75, respectively. These VMLs were much correlated with the time at which AAS pastes lost their flowability, as shown in Figure 5b,c, since the GGBFS reaction was quickened by tens of minutes after the mixing.

Viscosity Measurement Limit and Ultrasonic Pulse Velocity
The AAS paste with 4% Na2O retained its flowability for up to 1 h in the mini slump test, but its VML was only 15-20 min. This was because the AAS paste was remixed for 30 s before the measurement of the flow, while in the rheology test, the shear stress of the AAS paste was measured continuously for 30 min without remixing.  Figure 8 shows the rheology test results of the AAS pastes with Na 2 O content and Ms. The viscosity measurement limit (VML) refers to the point at which flowability is lost in the fluid state. As described in Section 3.2, the flow of the paste showed a tendency for the flowability to decrease abruptly immediately before the loss of flowability, and the shear stress rapidly increased 3 to 5 min before reaching VML. As shown in Table 4, the min. and max. VML times of the AAS paste with ≥6% Na 2 O content were 540 s for 8-1.25 and 1040 s for 6-0.75, respectively. These VMLs were much correlated with the time at which AAS pastes lost their flowability, as shown in Figure 5b,c, since the GGBFS reaction was quickened by tens of minutes after the mixing.

Viscosity Measurement Limit and Ultrasonic Pulse Velocity
The AAS paste with 4% Na 2 O retained its flowability for up to 1 h in the mini slump test, but its VML was only 15-20 min. This was because the AAS paste was remixed for 30 s before the measurement of the flow, while in the rheology test, the shear stress of the AAS paste was measured continuously for 30 min without remixing.  Figure 9 shows the ultrasonic pulse velocity (UVP) measurements of the AAS pastes; the initial ultrasonic velocity of the paste was 302.5-405.0 m/s regardless of the mixture type [32,33]. This indicates that the paste was in a fluid state. Afterwards, as the paste lost its flowability, the UPV increased sharply, and during certain periods of time before VML was reached, the UPV increased intermittently rather than continuously. Table 5 summarizes the UPVs of the time at which the VML was reached. The UPVs of the AAS pastes increased gradually during the reaction of the GGBFS [34]. The rheology test results showed that when the AAS paste reached the VML, the UPVs were confirmed to be 900-1000 m/s. After the VML, the UPVs increased steadily, and then, the increase rate of UPVs differed in accordance with the Na2O content and Ms. It is observed in Figure 9 that after the VML, UPVs of the AAS pastes with an Na2O content of ≥6% and Ms of ≥1.00 rapidly increased, but UPVs of the AAS pastes with 4% Na2O content or Ms of 0.75 modestly increased.   Figure 9 shows the ultrasonic pulse velocity (UVP) measurements of the AAS pastes; the initial ultrasonic velocity of the paste was 302.5-405.0 m/s regardless of the mixture type [32,33]. This indicates that the paste was in a fluid state. Afterwards, as the paste lost its flowability, the UPV increased sharply, and during certain periods of time before VML was reached, the UPV increased intermittently rather than continuously. Table 5 summarizes the UPVs of the time at which the VML was reached. The UPVs of the AAS pastes increased gradually during the reaction of the GGBFS [34]. The rheology test results showed that when the AAS paste reached the VML, the UPVs were confirmed to be 900-1000 m/s. After the VML, the UPVs increased steadily, and then, the increase rate of UPVs differed in accordance with the Na 2 O content and Ms. It is observed in Figure 9 that after the VML, UPVs of the AAS pastes with an Na 2 O content of ≥6% and Ms of ≥1.00 rapidly increased, but UPVs of the AAS pastes with 4% Na 2 O content or Ms of 0.75 modestly increased.   Figure 10 shows the compressive strength development of the AAS mortars. The AAS mortars with a 4% Na2O content had a compressive strength of <1 MPa up to 24 h regardless of Ms, implying that the increase of SiO2 content did not significantly affect the compressive strength development. This is consistent with the very low heat release for the initial 24 h, as shown in Figure 4a. On the other hand, although the mortar with ≥6% Na2O showed an increase in strength as time elapsed, the compressive strength differed in accordance with the Ms. When Ms remained constant, the compressive strength development of the AAS mortars with 8% Na2O was very similar to that of the AAS mortars with a 6% Na2O. This was in accordance with the results of Zuo et al. [13], who reported that when Na2O was greater than 6%, the reaction degree did not increase significantly with time, which in turn kept compressive strength relatively constant.  Figure 10 shows the compressive strength development of the AAS mortars. The AAS mortars with a 4% Na 2 O content had a compressive strength of <1 MPa up to 24 h regardless of Ms, implying that the increase of SiO 2 content did not significantly affect the compressive strength development. This is consistent with the very low heat release for the initial 24 h, as shown in Figure 4a. On the other hand, although the mortar with ≥6% Na 2 O showed an increase in strength as time elapsed, the compressive strength differed in accordance with the Ms. When Ms remained constant, the compressive strength development of the AAS mortars with 8% Na 2 O was very similar to that of the AAS mortars with a 6% Na 2 O. This was in accordance with the results of Zuo et al. [13], who reported that when Na 2 O was greater than 6%, the reaction degree did not increase significantly with time, which in turn kept compressive strength relatively constant.  Figure 10 shows the compressive strength development of the AAS mortars. The AAS mortars with a 4% Na2O content had a compressive strength of <1 MPa up to 24 h regardless of Ms, implying that the increase of SiO2 content did not significantly affect the compressive strength development. This is consistent with the very low heat release for the initial 24 h, as shown in Figure 4a. On the other hand, although the mortar with ≥6% Na2O showed an increase in strength as time elapsed, the compressive strength differed in accordance with the Ms. When Ms remained constant, the compressive strength development of the AAS mortars with 8% Na2O was very similar to that of the AAS mortars with a 6% Na2O. This was in accordance with the results of Zuo et al. [13], who reported that when Na2O was greater than 6%, the reaction degree did not increase significantly with time, which in turn kept compressive strength relatively constant.

Conclusions
The objective of this study was to investigate the effects of Na2O content and Ms of the alkali activator on the workability and setting of AAS pastes. To evaluate the fresh properties of the pastes, mini slump, Vicat test, and rheology/UPV test were applied. Based on the experimental results, the following conclusions can be drawn: 1. The Na2O content and Ms strongly affected the rate of heat release, workability, and setting time of alkali-activated pastes. In particular, when alkaline activators with Na2O content of ≥6% (Ms = 0.75-1.25) were used, the activation and reaction of the GGBFS were enhanced because of the effects of high dosage of alkaline (Na2O + SiO2), which resulted in workability rapidly decreasing and setting quickly occurring. However, the Na2O content and Ms did not significantly affect the compressive strength development of the AAS mortars when the Na2O content increased from 6% to 8%. On the other hand, for the pastes with Na2O content = 4%, the rate of heat release was quite low for the initial 24 h and setting was delayed up to 4 h and 40 min, and even the workability was maintained for more than 1 h. 2. In AAS pastes with high Na2O content, setting occurred within a few minutes after flowability was lost, and thus, it was not easy to measure the flowability with the existing mini slump test. However, since the changes in viscosity of the paste can be continuously monitored over time using the rheology/UPV test, the measurement of VML and UPV could allow the quantitative assessment of the time at which flowability was lost and consequently, to predict the initial setting time of the paste. 3. For the production of AAS concrete with high early strength of 25 MPa at 24 h, it is suggested from the test results that the Na2O content and Ms of the alkaline activator should be as much as 6% and 1.0, respectively. However, to secure the workability of AAS paste, chemical admixtures such as superplasticizers should be used to adequately control the workability and setting in highly alkaline environments.

Conclusions
The objective of this study was to investigate the effects of Na 2 O content and Ms of the alkali activator on the workability and setting of AAS pastes. To evaluate the fresh properties of the pastes, mini slump, Vicat test, and rheology/UPV test were applied. Based on the experimental results, the following conclusions can be drawn: 1.
The Na 2 O content and Ms strongly affected the rate of heat release, workability, and setting time of alkali-activated pastes. In particular, when alkaline activators with Na 2 O content of ≥6% (Ms = 0.75-1.25) were used, the activation and reaction of the GGBFS were enhanced because of the effects of high dosage of alkaline (Na 2 O + SiO 2 ), which resulted in workability rapidly decreasing and setting quickly occurring. However, the Na 2 O content and Ms did not significantly affect the compressive strength development of the AAS mortars when the Na 2 O content increased from 6% to 8%. On the other hand, for the pastes with Na 2 O content = 4%, the rate of heat release was quite low for the initial 24 h and setting was delayed up to 4 h and 40 min, and even the workability was maintained for more than 1 h.

2.
In AAS pastes with high Na 2 O content, setting occurred within a few minutes after flowability was lost, and thus, it was not easy to measure the flowability with the existing mini slump test. However, since the changes in viscosity of the paste can be continuously monitored over time using the rheology/UPV test, the measurement of VML and UPV could allow the quantitative assessment of the time at which flowability was lost and consequently, to predict the initial setting time of the paste.

3.
For the production of AAS concrete with high early strength of 25 MPa at 24 h, it is suggested from the test results that the Na 2 O content and Ms of the alkaline activator should be as much as 6% and 1.0, respectively. However, to secure the workability of AAS paste, chemical admixtures such as superplasticizers should be used to adequately control the workability and setting in highly alkaline environments.

Conflicts of Interest:
The authors declare no conflict of interest.